A Corner of Mathematical

Olympiad and Competition

Book I
1. An Introduction to Mathematical Olympiad Problems
2. Alberta High School Mathematics Competition
3. Bulgarian Mathematical Olympiad and Competition
4. Annual Vojtĕch Jarník International Mathematical
Competition

Problems with Solutions

Phnom Penh, October 24, 2014

Prepare by: Keo Sodara
Contact: [email protected]
[email protected]
www.facebook.com/aprissigi
(+855)0967308278
 An Introduction of
Mathematical Olympiad
Problems
1. Auckland Mathematical Olympiad
1997, 1998 and 2001 - 2004
2. Irish Mathematical Olympiad
2007 - 2009
3. Mexican Mathematical Olympiad
2001 - 2004

Problems with Solution

Auckland Mathematical Olympiad 1997
DIVISION 1
Questions
1. A piece of alloy of copper and silver contains 5% of silver and has mass 4
kg. What mass of copper should be added to get a piece of alloy which
contains 2% of silver?

2. Find and draw the set of points (x, y) in the plane which satisfy the equa-
tion
|2x + y| = |x + y + 1|.

3. Can you draw 6 line segments so that each segment intersects exactly four
other segments?

4. Two squares are placed one inside the other so that their sides are parallel
as shown in Figure 1. The parts A and C are coloured in red and the
parts B and D are coloured in green.

A C

Figure 1:

Prove that the amount of red colour used is the same as the amount of
green colour.

5. Given a number written on the blackboard in one move you can either
replace it by this number squared or else by this number doubled. How
to get 245 exactly in 10 moves starting from 1?

1
 Auckland Mathematical Olympiad

DIVISION 2
Questions
6. When Alice entered the Forest of Forgetfulness, she did not forget every-
thing; only certain things. She often forgot her name, and the one thing
she was most likely to forget was the day of the week. The Lion and the
Unicorn were frequent visitors to the forest. These two are strange crea-
tures. The Lion lies on Mondays, Tuesdays, and Wednesdays and tells the
truth on the other days of the week. The Unicorn, on the other hand, lies
on Thursdays, Fridays, and Saturdays, but tells the truth on the other
days of the week.
One day Alice met the Lion and the Unicorn resting under a tree. They
made the following statements:

Lion: Yesterday was one of my lying days.

Unicorn: Yesterday was one of my lying days too.

From these statements, Alice (who was a very bright girl) was able to
deduce the day of the week. What day was it?

7. In chess, is it possible for a knight to travel (by legal moves only) from
the lower left-hand corner of the board to the upper right-hand corner of
it visiting on the way each cell exactly once?

8. Prove that




√ √ √
17 − 12 2 + 3−2 2+ 3 + 2 2 = 3.

9. Let r be the inradius and R be the circumradius of a right-angled

√ triangle
ABC. Let S be the area of ABC. Prove that r + R > 2S.

10. We are given 80 coins of the same denomination. We know that one of
them is a counterfeit and that it is lighter than the others. Determine the
counterfeit coin by using four weighings on a pan balance.

2
 Solutions
1. The piece contains 4000/20 = 200g of silver. For having 2% of silver the
total mass must be 200 × 50 = 10000g = 10kg. Therefore 6kg of copper
should be added.
2. This equation is equivalent to
(2x + y) = ±(x + y + 1).
Therefore point (x, y) belongs to the given set if either (2x+y) = (x+y+1)
or (2x + y) = −(x + y + 1). The first equation represents the line x = 1,
the second the line 3x + 2y = −1. Therefore the given set is the union of
these two lines.
3. Yes, one can do this. The idea is to take three pairs of parallel lines as
shown on Figure 2. Each line will intersect all lines which are not parallel
to it, i.e., four lines out of five. Now these lines can be shortened to
segments and if these segments are sufficiently long they will also satisfy
the property.

Figure 2:

4. Let us prove that the combined area of parts A and C is equal to the
combined area of parts B and D. Let x be the sidelength of the large
square and y be the sidelength of the small one. Let us denote hA , hB ,
hC , hD the altitudes of the trapeziums A, B, C, D, respectively (see
Figure 3).
As hA + hC = hB + hD = x − y, the combined area of A and C is
x+y x+y x+y x+y x2 − y 2
hA + hC = (hA + hC ) = (x − y) = .
2 2 2 2 2
Similarly, the combined area of B and D is
x+y x+y x+y x+y x2 − y 2
hB + hD = (hB + hD ) = (x − y) = .
2 2 2 2 2

3
 hD

hA hC

hB

Figure 3:

These combined areas A + C and B + D are the same.

5. We will start from 245 and move backwards as fast as we can. Since
245 is not the square of a whole number the last move could be only
doubling: 245 → 244 . Then we can extract square root three times in a
row: 244 → 222 → 211 . This is again not a square of the whole number
and we can get it only from 210 . Eventually we get

245 → 244 → 222 → 211 → 210 → 25 → 24 → 22 → 2 → 1

This steepest descent gives us 9 moves. The 10th move we can get at the
very beginning by squaring 1. Thus the ten moves will be

1 → 1 → 2 → 22 → 24 → 25 → 210 → 211 → 222 → 244 → 245 .

6. The Lion could make such a statement either on Monday or on Thurs-

day. The Unicorn could make such a statement either on Thursday or on
Sunday. Therefore it was Thursday.

7. No, it is impossible. Since the knight has to visit every cell exactly once,
it has to make exactly 63 moves. At each move the knight changes the
colour of the cell. Therefore the initial cell and the cell where the knight
stops have to be of different colours. But the lower left-hand corner of the
board and the upper right-hand corner are black.
√ √ √ √ √ √
8. Since 17−12 2 = (3−2 2)2 , 3−2 2 = ( 2−1)2 , and 3+2 2 = (1+ 2)2




√ √ √
17 − 12 2 + 3 − 2 2 + 3 + 2 2 =
√ √ √
(3 − 2 2) + ( 2 − 1) + (1 + 2) = 3.

4
 9. Suppose that the incircle touches the sides of ABC at points K, L, M , as
shown in Figure 4. We note that AK = AM = r, BK = BL, CM = CL,
and as the angle
A = 90◦ we have BL + LC = 2R. Therefore r + R =
AK + BL + LC = p, the semiperimeter. Now, since r < R, we have

2S = 2pr = 2(r + R)r = 2r2 + 2rR < r2 + 2rR + R2 = (r + R)2 ,

√
or 2S < r + R.

L
K

A M C

Figure 4:

10. Put arbitrary 27 coins on the left pan of the balance and another 27
coins on the right pan of the balance. If their weights are equal then
the counterfeit coin is among those 26 coins which are set aside. If one
pan is lighter than the other then we know that the counterfeit coin is
one of the coins on this pan. In any case we can locate 27 coins with
the counterfeit coin among them. Now we deal with these 27 coins and
divide them in three groups with 9 coins in each group. Setting one group
aside we compare the two others. As before, as a result of this second
weighing we will find a group of 9 coins with the counterfeit coin among
them. By the third weighing such a group can be reduced to a group of 3
coins. Setting one of them aside and weighing the two others we find the
counterfeit coin as a result of the fourth and final weighing.

5
 Auckland Mathematical Olympiad 1998
Division 1
Questions
1. During her journey in Wonderland Alice changed her size four times.
She successively: took a sip from a bottle marked “Drink me!” and
became 25% taller; made a bite of a pie marked “Eat me!” and became
10% shorter; took a sip from a bottle marked “Drink me!” and became
10% taller; made a bite of a pie marked “Eat me!” and became 20%
shorter. Was she taller or shorter after the journey than she was before?

2. The first number in a sequence is 7. The next one we obtain as follows:

we compute the square of the previous number 72 = 49, then compute
the sum of the digits of this square and increase it by 1, i.e., the second
number is 4 + 9 + 1 = 14. We repeat this procedure to get 142 = 196
and the third number of the sequence, which is 1 + 9 + 6 + 1 = 17, and
so on. What would be the 1999th number in this sequence?

3. A square BCDE is constructed on the hypotenuse BC of a right-angled

triangle ABC as shown in the Figure.

C E
O
n

A m B

Let O be the centre of the square. Given that the legs AB and AC
have lengths m and n, respectively, find the distance AO.

1
4. The parabola whose equation is 8y = x2 meets the parabola whose
equation is x = y 2 at two points. What is the distance between these
two points?

5. Towns A and B are situated on a river at a distance 10 km apart. What

will take more time for a ship: to steam from A to B and back to A or
to steam 20 km in a lake with no current?

2
 Auckland Mathematical Olympiad 1998
Division 2
Questions
6. Find all real solutions of the system of equations


x + y + xy = 11
x2 + xy + y 2 = 19

7. Some cells of an infinite square grid are coloured black and the rest
are coloured white so that each rectangle consisting of 6 cells (2 × 3
or 3 × 2) contains exactly 2 black cells. How many black cells might a
9 × 11 rectangle contain?
8. Two circles C1 and C2 of radii r1 and r2 touch a line
at points A1 and
A2 , as shown in the figure below.

N C2

C1

A1 A2 l

The circles intersect at points M , N . Prove that the circumradius of

the triangle A1 M A2 does not depend on the length of A1 A2 and is
√
equal to r1 r2 .
9. Let α and β be two acute angles such that sin2 α + sin2 β = sin (α + β).
Prove that α + β = π/2.
10. Find all prime numbers p for which the number p2 + 11 has exactly 6
different divisors (including 1 and the number itself).

3
 Solutions
1. Let h be Alice’s initial height. Then at the end of her journey her
height was h × 1.25 × 0.9 × 1.1 × 0.8 = 0.99h, i.e., after her trip she
was slightly shorter than she originally was.

2. The first numbers of the sequence are:

7, 14, 17, 20, 5, 8, 11, 5, . . .
and we see that apart from the first four numbers the sequence is pe-
riodic with the period 3. Since 1999 has remainder 1 on dividing by 3,
the number sought will coincide with the number in the 7th place, i.e.,
it is 11.

3. Let us rotate the square BCDE clockwise about its centre O through
90◦ , 180◦ and 270◦ . Vertex A will be then mapped onto A1 , A2 and A3 ,
respectively (see Figure 3).

A1 D A2

E
C O

A B A3

Since ∠ACB = ∠A1 DC we have ∠ACB + ∠BCD + ∠DCA1 = 180◦ ,

thus ACA1 is a straight segment. Hence AA1 A2 A3 is a square with
side m + n. Since OA rotated 180◦ becomes OA2 , the segment AA2 is
a diagonal of this new square with OA being half its length. Therefore
OA = √12 (m + n).

4
4. One point is clearly (0, 0). Solving the system of equations

2
x = 8y
x = y2

we get y 3 = 8 and x3 = 64, i.e.,

√ another √
point of intersection
√ is (x, y) =
(4, 2). Thus the distance is 42 + 22 = 20 = 2 5.

5. Let u be the speed of the ship and v be the speed of the current. Then
the time which is needed for the ship to get from A to B and back to
A will be
10 10 20u 20
+ = 2 = 2 .
u+v u−v u −v 2
u − vu
20
This is clearly greater than , which is the time required to steam
u
20km in the lake.

6. Adding these two equations we get (x + y)2 + (x + y) = 30, and hence

either x+y = 5 or x+y = −6. In the first case xy = 6 and (x, y) = (2, 3)
or (x, y) = (3, 2). In the second case xy = 17 and the corresponding
quadratic equation z 2 + 6z + 17 = 0 does not have solutions.

7. By a m × n rectangle we understand one with m rows and n columns.

We will prove that there are 33 black cells in a 9 × 11 rectangle. This
will be clear if we can prove that in every 3×1 rectangle there is exactly
one black cell. (Certainly a similar proof will be applicable to any 1 × 3
rectangle.) Obviously, it cannot contain three black cells; suppose that
exactly two of them are black (see the Figure 4).

5
 Then the two neighboring 3 × 1 rectangles contain only white cells,
otherwise we can find a 3 × 2 rectangle with more than two black cells
in it. In the emerging 3 × 3 square, shown in the Figure, at least one
2 × 3 rectangle contains only one black cell. Hence, it is not possible
that two cells in a 3 × 1 rectangles are black. It is also impossible that
all of them are white as the neighboring 3 × 1 rectangle must in this
case have two black cells which, as we know, is not possible. Therefore
every 3 × 1 rectangle contains exactly one black cell.

8. Let us draw the perpendicular to
passing through A1 which intersects

the circle C1 a second time at K.

A1 A2 l

Let ρ be the circumradius of A1 M A2 . Then

A1 M = 2ρ sin ∠A1 A2 M = 2r1 sin ∠M KA1

and since ∠M KA1 = ∠M A1 A2 we have

ρ sin ∠A1 A2 M = r1 sin ∠M A1 A2 .

Similarly
ρ sin ∠M A1 A2 = r2 sin ∠A1 A2 M.
Multiplying these two equations we get ρ2 = r1 r2 .

6
 9. We prove this by contradiction. The equation sin2 α + sin2 β = sin(α +
β) is equivalent to

sin α(sin α − cos β) + sin β(sin β − cos α) = 0. (1)

Now both cases lead to a contradiction. If α+β < π/2, then α < π/2−β
and β < π/2 − α, whence
π  π 
sin α < sin − β = cos β, sin β < sin − α = cos α,
2 2
hence (1) cannot be fulfilled as the left-hand-side of (1) is negative.
If α + β > π/2, then α > π/2 − β and β > π/2 − α, whence
π  π 
sin α > sin − β = cos β, sin β > sin − α = cos α,
2 2
hence (1) also cannot be fulfilled as the left-hand-side of (1) is now
positive.

10. For p = 2, we have p2 + 11 = 15, which has 4 divisors. For p = 3,

we have p2 + 11 = 20, which has 6 divisors as required. No more such
primes exist. We will prove that for p > 3 the number p2 +11 is divisible
by 12. Since p2 + 11 > 12 and 12 has 6 divisors already all such primes
will not satisfy the condition.
Since p2 + 11 = (p2 − 1) + 12, it is enough to prove that p2 − 1 is
divisible by 12. In fact, it is divisible even by 24. We know that
p2 − 1 = (p − 1)(p + 1). Among the three numbers p − 1, p, p + 1 one is
divisible by 3 and it is not p. Since p is odd, then p − 1 and p + 1 are
both even and one of them is divisible by 4. Thus p2 − 1 is divisible by
24.

7
 Auckland Mathematical Olympiad 2001
Division 1
Questions
1. Kate walks to school every morning. Once, when she had walked 1/5
of the way, Kate realised that she had left her homework behind. If
she continued walking, she would get to school five minutes before the
bell. If she returns home to collect her homework, she would arrive to
school one minute late. How long does her usual trip to school take?

2. Prove that one of the two solutions of the quadratic equation

2000x2 − (2000 + 2001 + 0.12000 )x + 2001 = 0

is smaller than 1 and the other is greater than 1.

3. Two circles intersect at points A and B. Let AC be a diameter of the

first circle and AD be the diameter of the second. Prove that the points
B, C, D lie on the same straight line.

4. Find all values a for which the equation

|x − a| + |x + 3a − 8| = 4

has infinetely many solutions.

5. Among four given coins three coins are genuine and have the same
weight which is unknown. The remaining one is false and its weight
is different. There is a spring balance, which shows the correct weight
but only when two or more coins are measured. The weight of a single
coin cannot be measured correctly. Using four measurements how is it
possible to determine which coin is false and whether it is heavier or
lighter than a genuine one?

1
 Auckland Mathematical Olympiad 2001
Division 2
Questions
6. How many solutions in positive integers has the equation
√ √ √
x + y = 2001.

7. A tourist wants to tour a country starting from the capital A. From

the capital he flies to the city B which is the farthest city from the
capital. Then he flies to the city which is the farthest city from B and
which happens to be different from A. Prove that, if he follows this
rule further, he will never return to the capital.

8. A circle with the centre O is circumscribed about triangle ABC. An-

other circle which passes through points A, B, O touches the side AC
at point A. Prove that ABC is an isosceles triangle.

9. Two players are playing a game using two heaps of beans. A legal move
is to take one bean from a single heap (either of the two), or else to
take one bean from each of the two heaps. The loser is the player who
cannot make a legal move. Determine all positions in which the first
player has a winning strategy.

10. There is an island in the Pacific where only honest people and liars live.
Honest people always tell the truth, liars always lie. When presidential
elections were announced, there were n ≥ 3 candidates, all of them
locals. During a TV debate all the candidates, each in turn, made the
following statements. The kth candidate said:

Among the candidates, except me, there are k more liars than
honest people.

How many candidates were there?

2
 Solutions

1. Returning home takes additional 5 + 1 = 6 minutes. The additional

distance is 1/5 × 2 = 2/5 part of Kate’s usual trip. So 2/5 of her usual
trip takes 6 minutes, thus the whole trip takes 15 minutes.

2. let us consider the graph of the quadratic

y(x) = 2000x2 − (2000 + 2001 + 0.12000 )x + 2001.

The value y(1) = 2000−(2000+2001+0.12000 )+2001 = −0.12000 which

is negative. As the branches of this parabola are directed upwards they
intersect the x-axis once on the left and once on the right of the point
1. So one solution of our quadratic equation is smaller than 1 and the
other solution is greater than 1.

3. Let us draw the chord AB. As AC and AD are diameters of their

respective circles, the angles ∠ABC and ∠ABD are equal to 90◦ , hence
CBD is a straight segment.

C B D

4. If x takes values in such interval where x − a and x + 3a − 8 are both

positive or both negative then we can get rid of absolute values and
we will see that the left hand side of the equation is a linear function
in x. Hence the equation can have no more than one solution on that
interval. Therefore we can have infinitely many solutions only on those
intervals where x − a and x + 3a − 8 have opposite signs. If x + 3a − 8

3
 is negative, then the equation takes form (x − a) − (x + 3a − 8) = 4 or
a = 1. Substituting a = 1 we get the equation

|x − 1| + |x − 5| = 4,

which indeed has infinitely many solutions, namely the whole interval
1 ≤ x ≤ 5.
If x − a is negative, then then the equation takes form −(x − a) + (x +
3a − 8) = 4 or a = 3. Let a = 3, then the equation will be

|x − 3| + |x + 1| = 4.

It indeed has infinitely many solutions, namely the whole interval −1 ≤

x ≤ 3.

5. Let us weigh every combination of three coins from among the four. We
will have four measurements. Three of these measurements will show
the weight of two genuine coins and one false coin, so they are the
same. One measurement will show the weight of three genuine coins.
This measurement will be different. Immediately we can find a false
coin - this is the coin which did not participate in the weighing which
appeared to be different. If on one hand this triple of coins is heavier
than any other triple then a genuine coin is heavier than the false one.
If on the other hand this triple of coins is lighter than any other triple
then a genuine coin is lighter than the false one.
√ √ √
6. Suppose that x + y = 2001. Then x < 2001 and y < 2001. Then
√ √ √ √
x = 2001 − y and x = 2001 + y − 2 2001y. Hence

4 × 2001y = (2001 + y − x)2

Since 2001 = 3 × 23 × 29 we get

4 × 3 × 23 × 29 × y = n2 .

Therefore y must be divisible by 3, 23, and 29, hence divisible by

2001, the contradiction. Hence this equation does not have solutions
in positive integers.

4
7. Suppose that this tourist will eventually return to the capital. Let us
denote B as B1 and let B2 , . . . , Bn be all other cities that the tourist
had visited before returning to the capital. Then we have

AB1 < B1 B2 < B2 B3 < . . . < Bn−1 Bn < Bn A.

This cannot be the case since it means that Bn not B1 was the farthest
city from A, contrary to the initial assumption.

8. Let O1 be the centre of the second circle. Then AO1 = O1 O and the
triangle O1 AO is isosceles. As O1 A ⊥ AC and AB ⊥ O1 O we see that
∠BAC = ∠AO1 O. Also ∠ACB = 12 ∠AOB = ∠AOO1 . Hence the
triangles ABC and O1 OA are similar and ABC is also isosceles.

O1
B

A C

9. The players must obey (if they can) the following Golden Rule:

Leave an even number of beans in every heap after each move.

If in her first move the first player can leave an even number of beans
in each heap, then she wins. Indeed, the second player will not be able
to fulfil the Golden Rule and the first player again makes the numbers
of beans in each heap by simply repeating the second player’s move.
The second player will always leave an odd number of beans at least in
one heap, thus she cannot win. Hence the winning positions are those
for which at least one heap contains an odd number of beans.

5
10. If all candidates were liars then the statement of the next to last can-
didate would have been truthful. If there were at least two honest
candidates they would have made identical statements. Therefore only
one candidate was honest and she was the (n−1)th candidate to speak.
If n > 3 then the statement of the (n − 3)th candidate would have been
also truthful. Hence only the case n = 3 does not contradict the con-
ditions of the problem.

6
 Auckland Mathematical Olympiad 2002
Division 1

Questions

1. Nine identical books cost less than $10 in total and ten such books
cost more than $11. How much does one book cost?

2. Two squares, whose areas are A and B, are placed in a semicircle as

shown below. Find the ratio A/B.

A
B

3. The Land of Oz is blessed by many things but not efficient structure of

their military forces. Their armored brigade consists of two armored
batallions and one infantry batallion. Their infantry brigade consists
of two infantry batallions and one armored batallion. Any armored
batallion consists of two armored squadrons and one infantry squadron
and any infantry batallion consists of two infantry squadrons and one
armored squadron. Any armored squadron consists of two armored
squads and one infantry squad and any infantry squadron consists
of two infantry squads and one armored squad. Any armored squad
consists of two armored soldiers and one infantry soldier and any in-
fantry squad consists of two infantry soldiers and one armored soldier.
What is the least number of soldiers that must be retrained in order
to transform an infantry brigade into an armored brigade?

4. Find all integer solutions of the equation xy + 3x − 5y = +1.

5. Given that there exists a triangle with lengths of its sides√a, b, c. prove
√ √
that there exists a triangle with lengths of its sides a, b, c.

1
 Auckland Mathematical Olympiad 2002
Division 2
Questions
6. Solve the system of equations
 2

 x + y2 + z2 = 14
 2
x + y 2 + t2 = 21

 x2 + t2 + z 2 = 26
 2
t + y2 + z2 = 29

7. The number 11 

. . . 1 is divisible by 37. Prove that the number of units
m
m is divisible by 3.
8. Cut a rectangle 1.5 × 4 into two pieces with which it is possible to
cover the surface of a unit cube.
9. How many negative roots does the following equation have
−2x5 + x4 − 5x3 + 4x2 − 13x + 4 = 0?

10. Three circles c1 , c2 , c3 pass through the common point I. Let P , Q,

R be the points of intersection of c1 and c2 , c2 and c3 , c3 and c1 ,
respectively, different from I. Let A be a point on c1 lying outside c2
and c3 as shown in the picture below.

Q c3

B
I
R

c2 c1
P
A

Let B be the point of intersection of AP with c2 , C be the point of

intersection of BQ with c3 . Prove that CR will pass through A.

2
 Solutions

1. Let x be the price of one book. Then 9x < 10 and 10x > 11 from
which we have 1.1 < x < 1.1111... The only possibility is x = 1.11, i.e.
one dollar and 11 cents.
2. Let us denote the lengths of the sides of these squares as a and b,
respectively.

M
a

O b

We have (a/2)2 + a2 = ON 2 = OM 2 = (a/2 + b)2 + b2 from which for

the ratio x = a/b we find x2 − x − 2 = 0. Hence a/b = 2 and A/B = 4.
3. Answer: one soldier. Indeed, one soldier should be retrained to trans-
form an infantry squad into an armored squad, hence one soldier should
be retrained to convert an infantru squadron into an armored squadron
and an infantry batallion into an armored batallion, and finally an in-
fantry brigade into an armored brigade.
4. Since we cannot have any solutions with y = −3, we can write our
equation as
5y + 1 14
x= =5− .
y+3 y+3
A solution is possible only when y + 3 is a divisor of 14, i.e y + 3 =
±1, ±2, ±7, ±14. Respectively, we have eight solutions: (−9, −2);
(19, −4); (−2, −1); (12, −5); (3, 4); (7, −10); (4, 11); (6, −17).
5. Suppose that a ≤ b ≤ c. Then the existence of a triangle with √ sides
√ 2
+ b ≥ c. Then, since
a, b, c is√equivalent to a √ √ a + b = ( a) + ( b)2 ≤
√ √
( a + b) , we obtain a + b ≤ a + b
2

√ √ √ √
c ≤ a + b ≤ a + b,
√ √ √
which, together with √a ≤ b ≤ c, gives us the existence of a
√ √
triangle with sides a, b, c.

3
 6. Adding all equations together we get 3(x2 + y 2 + z 2 + t2 ) = 90 or
(x2 + y 2 + z 2 + t2 ) = 30. Subtracting each equation in turn from this
total, we get x2 = 1, y 2 = 4, z 2 = 9, t2 = 16. Thus x = ±1, y = ±2,
z = ±9, t = ±4

7. Let N = 11 
. . . 1 be a multiple of 37. Suppose m = 3k + 2, then
m

N = 11 
. . . 1 = 11 
. . . 1 00 + 11
m 3k

Since 111 = 37 × 3, we see that 11 . . . 1 00 is divisible by 111 and hence

 
3k
by 37. But then 11 = N − 11 . . . 1 00 must be divisible by 37, which is
 
3k
impossible. Similarly we consider the case m = 3k + 1.

8. We cut it as shown below on the left picture and join the two parts as
shown on the right one.

9. The equation

−2x5 + x4 − 5x3 + 4x2 − 13x + 4 = 0,

may be written as x4 + 4x2 + 4 = 2x5 + 5x3 + 13x or (x2 + 2)2 =

2x5 + 5x3 + 13x. This equation cannot have negative roots because
when x is negative, the right-hand-side is negative whereas the left-
hand-side is always positive.

10. Instead of proving that CR passes through A, we will draw the seg-
ments CR and AR and prove that ∠CRA = 180◦ . Let us draw the
segments IP, IQ, IR. Since the quadrilateral AP IR is cyclic, we ob-
tain that ∠ARI = 180◦ − ∠AP I and therefore ∠ARI = ∠BP I.
Similarly ∠IRC = ∠BQI. But ∠BP I + ∠BQI = 180◦ . Hence
∠ARC = ∠ARI + ∠IRC = ∠BP I + ∠BQI = 180◦ .

4
 Auckland Mathematical Olympiad 2003
Division 1

Questions

1. Meteorologists of the Land of Oz during one week of observation no-

ticed that on any day if the morning was fine, then the afternoon was
rainy and if the afternoon was rainy, then the morning was fine. Dur-
ing this week 7 mornings and 6 afternoons were fine and there were 9
rainy days. How long is the week in the Land of Oz?

2. Show all points (x, y) on the coordinate plane which satisfy the con-
dition


x2 − 4x + 4 + y ≤ x.

3. The vertex A of square ABCD is in the centre of square EF GK. Both

squares have area 1. When the area of the common part of the squares
is minimal? Find this area.

G K B

C
F E

4. Prove that the last digit (of units) of n5 is the same as the last digit
of n.

5. 1,000,000 points are marked on the plane. Prove that there exists a
circle which contains exactly 2003 marked points in its interior.

1
 Auckland Mathematical Olympiad 2003
Division 2

Questions
6. A number 2n + 1 is a square of a positive integer. Prove that n is a
multiple of 4.

7. Given that



(x + x2 + 1)(y + y 2 + 1) = 1.

prove that x + y = 0.

8. A starfish often seen at Piha beach looks like the star on the picture

C
B

A
E

Calculate the sum of angles ∠A + ∠B + ∠C + ∠D + ∠E.

9. Among 4n numbers any four are consequtive members of an arithmetic

progression. Prove that at least n of these numbers are equal.

10. Let A, B be two distinct points of intersection of two unequal coplanar

circles C1 and C2 . One of the common tangents to the circles touches
C1 at P1 and C2 at P2 while the other touches C1 at Q1 and C2 at
Q2 . Prove that the line AB passes through the midpoints of P1 P2 and
Q1 Q2 .

2
 Solutions

1. On each rainy day, exactly half of the day (morning or afternoon) was
rainy. Therefore we are able to count the total number of mornings
and afternoons in one week: 6+7+9=22. Therefore a week lasts 11
days.
√

2. Since x2 − 4x + 4 = (x − 2)2 = |x − 2| we obtain an equivalent
inequality

|x − 2| + y ≤ x.

For x ≥ 2 all points (x, y) with y ≤ 2 satisfy this inequality and for
x < 2 only those which lie under the line y = 2x − 2.

3. Since AKM = AEM , the area is constant and is equal to 1/4.

G K B

C
F N E

4. Let n = 10a + b, where 0 ≤ b ≤ 9, i.e. b is the last decimal digit of N .

Then

n5 = 105 a5 + 5 · 104 a4 b + 10 · 103 a3 b2 + 10 · 102 a2 b3 + 5 · 10ab4 + b5 ,

which is of the form n5 = 10c + b5 . Hence the last digit of n5 is the

last digit of b5 . Therefore the statement has to be checked only for
one-digit numbers, which is easy.

5. For any two of the marked points let us draw the perpendicular bisec-
tor. They will intersect at a finite number of points. We may always

3
 choose a point O where no two perpendicular bisectors intersect. Let
us draw a very small circle centred at O, so small that it does not con-
tain any of the marked points. Then let us start gradually increasing
its radius. The marked points will get into the interior of the circle
one at a time. Eventually there will be 2003 of them.
6. Suppose that 2n + 1 = a2 , then a is odd. We have 2n = a2 − 1 =
(a − 1)(a + 1). Since A is odd, one of the two numbers a − 1, a + 1
is divisible by 2 and another by 4. Hence 2n is divisible by 8 and n is
divisible by 4.
√
7. Multiplying by x − x2 + 1 we get



y + y 2 + 1 = x − x2 + 1


and multiplying by y − y 2 + 1, we get



y − y 2 + 1 = x + x2 + 1

Adding them together we get 2(x + y) = 0, whence x + y√= 0. Al-

ternatively, one might prove that the function f (x) = x + x2 + 1 is
monotone increasing and



(x + x2 + 1)(x − x2 + 1) = 1.

8. Counting angles of the triangle BDE we find

∠B + ∠D + ∠E + ∠F DE + ∠F ED = 180◦

C
B

D
F

A
E

Since ∠F DE + ∠F ED = 180◦ − ∠EF D = 180◦ − ∠AF C = ∠A + ∠C,

we get ∠A + ∠B + ∠C + ∠D + ∠E = 180◦ .

4
 9. Suppose that there are no n equal numbers but still the condition
is satisfied. Then there will be at least five distinct numbers, say
a, b, c, d, e. By the assumption, a, b, c, d are four consequtive members
of an arithmetic progression, i.e

b = a + d, c = a + 2d, d = a + 3d

for some d = 0. Let us consider now the numbers a, c, d, e. They also

must be four consequtive members of an arithmetic progression and
the only way they can is that e fills the gap left by the absence of b,
i.e. e = b, which is a contradiction.

10. We simply notice is that P2 S 2 = SA · SB = P1 S 2 , whence P2 S = P1 S.

P2
S
P1

B
Q1

Q2

5
 Auckland Mathematical Olympiad 2004
Division 1

Questions

1. Let a and b be two odd integers such that a > b. Prove that a2 − b2 is
a multiple of 8.

2. Find the smallest integer root of the equation

 7

3 sin x 5
x −3
2
(|x| − 1) x+ = 0,
2

where |x| is the absolute value of x.

3. The square ABCD has sides of length a. The side AB is extended

beyond B to point E by a line segment of length x. Similarly the
sides BC, CD, DA are extended beyond C, D, A, to points F , G, H,
respectively, by line segments of the same length x.

(a) Prove that the quadrilateral EF GH is a square.

(b) Compute x, given that the area of EF GH is twice the area of
ABCD.

4. The sum of three positive integers is 2005. What is the minimal value
of their least common multiple?

5. Is it possible to assign numbers 1, 2, . . . , 12 to the edges of a cube (one

number to each edge) so that, for each vertex, the sum of the numbers
assigned to the three edges adjacent to this vertex is the same.

1
 Auckland Mathematical Olympiad 2003
Division 2

Questions

4 cos x
6. Find two solutions to the equation 3 = sin x belonging to the
interval [0, 2π].

7. Find all integer solutions to the equation x! + y! + z! = t!.

8. Write down a polynomial P (x) with real coefficients such that P (1) = 1,
P (2) = 2 and P (n) is irrational for all integers n
= 1, 2.

9. F is the point on the side BC of triangle ABC such that AF intersects

the median line BD at point E such that AE = BC. Prove that the
triangle BEF is isosceles.

10. 21 vertices of a regular 50-gon are coloured. Prove that there exist
three coloured vertices that are the vertices of an isosceles triangle.

2
 Solutions

1. Suppose a = 2k + 1 and b = 2m + 1, where k, m are integers. Then

a2 − b2 = (a + b)(a − b) = 4(k − m)(k + m + 1).

If k and m are of the same parity, k − m is even, and, if one of these

numbers is odd and another is even, then k + m + 1 is even. In both
cases the number (k − m)(k + m + 1) is even.
2. If x is an integer, then
 7

3 5
x −3
2
x+
= 0.
2
Hence all integer roots are the roots of |x| − 1 = 0, which are ±1. The
smallest integer root is −1.
3. The right-angled triangles EBF , F CG, GHD, HEA are equal, hence
their hypothenuses are also equal. Thus EF = F G = GH = HE.
Also ∠HEF = ∠BEF + ∠HEB = ∠BEF + ∠EF B = 90◦ . Hence
EF GH is a square. The area of each of the triangles equals 12 x(a + x).

E
F
C

A
H
G
√
Thus we have the equation 4 × 12 x(a + x) = a2 from which x = 3−1
2 a.

4. Let a, b, c be the numbers such that a + b + c = 2005. We may assume

that a ≤ b ≤ c. Since 2005 is not divisible by 3, they are not all equal.
Hence 2a ≤ lcm (a, b, c). Also b ≤ lcm (a, b, c) and c ≤ lcm (a, b, c). We
obtain the following inequality

4010 = 2(a + b + c) ≤ 5 × lcm (a, b, c)

3
 Hence lcm (a, b, c) ≥ 802. At the same time lcm (a, b, c) = 802 is
realisable, when a = 401, b = c = 802.

5. Suppose that the desired assignment is achieved and, for each vertex,
the sum of the labels of its adjacent edges is N . Then the sum of all
labels will be 12 × 8N = 4N since every label is counted twice. But on
the other hand, it is 1 + 2 + · · · + 12 = 78, which is not divisible by
four. Contradiction.

6. x = π/2 and x = 2π/3.

7. We may assume that x ≤ y ≤ z. Then t! > z!, hence t! ≥ (z + 1)z!.

At the saame time t! = x! + y! + z! ≤ 3z!. Thus z + 1 ≤ 3 and z ≤ 2.
By inspection we find that the only solution is x = y = z = 2, t = 3.
√
8. P (x) = 2(x − 1)(x − 2) + x.

9. Let us continue BD beyond D to the point G such that BD = DG.

Then ABCG is a parallelogram with ∠AGB = ∠GBC. Since AG =
BC = AE, the triangle AEG is isosceles and ∠AGB = ∠AEG. Thus
∠BEF = ∠AEG = ∠EBF and the triangle BEF is isosceles.

10. Let A0 A1 . . . A49 be the 50-gon. Let us consider 10 pentagons:

Ai A1i A2i A3i A4i , i = 0, 1, 2, . . . , 9.

Since 21 points are coloured, at least one of the above regular pen-
tagons will have three coloured vertices. It is now easy to check that
any three vertices of a regular pentagon will be the vertices of an
isosceles triangle.

4
 Mathematical Olympiad Training
Solutions to the Dublin Area First Qualification Test
10th February 2007

1. Three people A, B and C play a tournament of table-tennis. After each match, the winner
plays the person who didn’t play last time. At the end of the tournament, A has played
34 matches, B has played 47 matches and C has played 57 matches.

(a) Determine, with proof, who lost the fourth match.

(b) Show that we do not have enough information to determine who won the fourth
match.

Solution:

(a) The total number of matches played is (34 + 47 + 57)/2 = 69. Since C played 57
matches we deduce that A and B played each other in exactly 69 − 57 = 12 matches.
Similarly we find that A played C in exactly 69 − 47 = 22 matches and B played C
in exactly 69 − 34 = 35 matches.
Since matches between B and C cannot be consecutive, we see that B and C must
have played each other in matches number 1,3,5,. . .,69. Therefore A played the 4th
match (against B or C) and lost.
(b) Either of B or C could have played A in the fourth match. A played all of the 34
even numbered matches and lost each one. Since A played B 12 times, A played C
22 times, and 12 + 22 = 34, there is no conditioning on which order these happened
in. It is easy to write down an example of a tournament in which B played A in the
fourth match and won, and an example where C played A and won.

2. Let H be the set {1, 5, 9, 13, . . .} of all positive integers which leave remainder 1 on division
by 4. Say that a number in H, other than 1, is a H-prime if it cannot be expressed as a
product of two smaller numbers belonging to H. Thus 5 and 9 are the first two H-primes.
Find four H-primes p1 , p2 , p3 , p4 , no two equal, with the property that p1 p2 = p3 p4 .
Solution: Clearly if p is an ordinary prime number and if p belongs to H, then p is a
H-prime. However, it is a fundamental property of ordinary primes that if q1 ≤ q2 and
q3 ≤ q4 are primes with q1 q2 = q3 q4 then q1 = q2 and q3 = q4 . So we can’t obtain examples
in this way.
On the other hand, a product of two numbers each of which leave remainder 3 on division
by 4 is a number leaving remainder 1:

(4k + 3)(4m + 3) = 4(4km + 3m + 3k + 2) + 1.

In particular, if q1 and q2 are both primes of the form 4k + 3, then q1 q2 is a H-prime (since
the nontrivial factors of this number do not belong to H).
Thus, taking any 4 primes of the form 4k + 3 - eg. q1 = 3, q2 = 7, q3 = 11 and q4 = 19
- we can obtain the four distinct H-primes p1 = q1 q2 , p2 = q3 q4 , p3 = q1 q3 , p4 = q2 q4 ,
satisfying p1 p2 = q1 q2 q3 q4 = p3 p4 .

1
3. Prove that the equation
x2 + y 2 = 2007 + z 2
has infinitely many integer solutions (i.e. with x, y, z all integers).
Solution: Re-write the equation as
x2 − z 2 = 2007 − y 2 .

Let y be any even integer. Then 2007−y 2 is odd, and so has the form 2m+1 for some integer
m. On the other hand, if we let x = z + 1, then x2 − z 2 = (x − z)(x + z) = x + z = 2z + 1.
So we can let z = m (and x = z + 1) to get a solution.
With this method we get the following infinite set of integer solutions:
For each n = 1, 2, . . . let
(2007 − y 2 ) − 1
y = 2n, z = = 1003 − 2n2 , x = z + 1 = 1004 − 2n2 ;
2
i.e., we have the family of integer solutions
(x, y, z) = (1004 − 2n2 , 2n, 1003 − 2n2 ), n = 1, 2, 3, . . .

4. (a) Find, with proof, the greatest possible value of xyzw subject to the condition that
x, y, z, w are positive real numbers with x + 2y + 3z + 4w = 16.
(b) Find, with proof, the smallest possible value of xyzw subject to the condition that
x, y, z, w are positive integers with x + 2y + 3z + 4w = 16.
Solution:
(a) Recall the Arithmetic-Geometric Mean Inequality (AGM). This states that if x1 , x2 , ... , xn
are positive real numbers, then the product
x1 + x2 + ... + xn n
 
x1 · x2 · ... · xn ≤
n
with equality if and only if x1 = x2 = ... = xn .
Here, take n = 4, x1 = x, x2 = 2y, x3 = 3z and x4 = 4w to get:
x + 2y + 3z + 4w 4
   4
16
(x)(2y)(3z)(4w) ≤ = = 44 ,
4 4
so

44 32
xyzw ≤ = .
2·3·4 3
So the maximum of xyzw is 32/3 , with equality when x = 2y = 3z = 4w = 4, that
is, when x = 4, y = 2, z = 4/3, w = 1.
(b) The popular choice w = 2, z = 1, y = 2, x = 1 gives xyzw = 4. The better choice
w = 1, z = 3, y = 1, x = 1 gives xyzw = 3. We now show this is the minimum possible
under the given hypotheses.
For suppose xyzw < 3. Since x = y = z = w = 1 does not satisfy the hypotheses,
xyzw = 2 and one of the x, y, z, w must be equal to 2 and the rest all equal to 1.
But then x + 2y + 3z + 4w = 16 is not satisfied.
Hence the minimum is 3.

2
5. A triangle has side lengths a, b, c. Prove that:

√ √ √ √ √ √
b+c−a+ c+a−b+ a+b−c≤ a+ b+ c.

Solution: The fact that a, b, c are the side lengths of a triangle implies that a + b ≥
c, b + c ≥ a, c + a ≥ b so none of the terms under the square root signs are negative.
Recall that for real numbers x, y the inequality 2(x2 + y 2 ) ≥ (x + y)2 holds, since

2(x2 + y 2 ) − (x + y)2 = x2 − 2xy + y 2 = (x − y)2 ≥ 0.

√ √
Take x = a + b − c, y = c + a − b to get

√ √
2(a + b − c + c + a − b) ≥ ( a + b − c + c + a − b)2

that is √ √
4a ≥ ( a + b − c + c + a − b)2 .

Hence √ √ √
2 a≥ a+b−c+ c+a−b .... (1).
Similarly √ √ √
2 b≥ a+b−c+ b+c−a .... (2).
and √ √ √
2 c≥ a+c−b+ c+b−a .... (3).
Adding (1),(2),and (3) gives
√ √ √ √ √ √
2( a + b + c) ≥ 2( a + b − c + b + c − a + c + b − a).

Dividing by 2 gives the required result.

6. Find, with proof, the greatest positive integer n which is divisible by all positive integers
k < n1/3 .
Solution: Note that the problem requires n to be divisible by all the positive integers
k < n1/3 but it does not require n to be divisible by the product of all such k (for
example, 4 is divisible by 2 and 4 but not by 2 × 4).
Let h be the greatest integer < n1/3 . If h = 7, then n must be divisible by 1, 2, 3, 4, 5, 6, 7
and hence n must be divisible by 4 × 3 × 5 × 7 = 420. Since 73 = 343 < 420 < 512 = 83 ,
n = 420 satisfies the conditions.
We now show that h cannot be greater than 7.
For suppose that h > 7. Then n must be divisible by 8 and hence by 8 × 3 × 5 × 7 = 840.
However, 840 > 729 = 93 , so n must also be divisible by 9 and hence by 8×9×5×7 = 2520.
So n ≥ 2520.
Observe that 2(x − 1)3 > x3 for all real numbers x > 4. So, since (n1/3 − 1) ≤ h < n1/3 ,
we deduce that h3 > n/2. Let r, s, t be the integers satisfying

3
 2r ≤ h < 2r+1 , 3s ≤ h < 3s+1 , 5t ≤ h < 5t+1 .

Then h is divisible by 2r , 3s and 5t . Also, n > 113 , so by hypothesis, 7 and 11 also divide
n.
Hence n is divisible by 2r 3s 5t × 7 × 11. But then

2r+1 3s+1 5t+1 × 7 × 11 77h3 77n

n ≥ 2r 3s 5t × 7 × 11 = > > > n.
2×3×5 30 60
This is a contradiction.
Hence h ≤ 7 and thus n = 420 is the number sought.

7. Let F0 = a, F1 = b and F2 = c where a, b and c are integers. For all n ≥ 3, let

Fn = Fn−1 + Fn−2 + Fn−3 . Determine, with proof, the conditions on a, b and c which
ensure the sequence F0 , F1 , F2 , F3 , F4 , . . . contains an infinite number of odd integers and
an infinite number of even integers.
Solution:
For any m ≥ 0, let us condition on the parity of {F3m , F3m+1 , F3m+2 } and see how this
dictates the parity of {F3m+3 , F3m+4 , F3m+5 }. Notice that there are eight distinct cases to
consider:

Case F3m F3m+1 F3m+2 F3m+3 F3m+4 F3m+5

(1) odd odd odd odd odd odd
(2) odd odd even even odd odd
(3) odd even odd even odd even
(4) odd even even odd odd even
(5) even odd odd even even odd
(6) even odd even odd even odd
(7) even even odd odd even even
(8) even even even even even even.

If F0 = a, F1 = b and F2 = c are all odd, then we find, by using the above table ‘Case (1)’,
that F3 , F4 and F5 will all be odd also. It follows that all subsequent entries will be odd.
We denote this by Case (1) ⇒ case (1). Similarly case (8) ⇒ case (8), producing an infinite
sequence of only even entries.
We find case (2) ⇒ case (5) ⇒ case (7) ⇒ case (4) ⇒ case (2), therefore containing an
infinite number of odd and an infinite number of even entries.
Also, case (3) ⇒ case (6) ⇒ case (3), therefore containing an infinite number of even and
an infinite number of odd entries.
From this, we may deduce that {Fn }n≥0 will contain an infinite number of odd entries and
an infinite number of even entries if and only if a, b, c are not all odd and not all even.
(Equivalently one could say there must be at least one odd and at least one even number
in {a, b, c}.)

4
8. 25 points are placed randomly inside a regular hexagon with side-length 2 units.

(a) Prove there are two points that are no farther than 1 unit apart.
(b) Show that with 24 points instead, the points may be placed so that no two are less
than 1 unit apart.

Solution:

(a) Divide the hexagon into 24 equilateral triangles of side length 1 as shown:

If we distribute 25 points amongst the 24 equilateral triangles, one triangle will contain
at least d25/24e = 2 points. If two points are placed in, or on the boundary of, an
equilateral triangle of side length 1, they are no farther than 1 unit apart.
(b) If we are dealing with 24 points instead, then the points may be placed as in the
diagram:

The black points in adjacent equilateral triangles are exactly 1 unit apart.

9. Determine the number of north-east lattice paths from the origin (0, 0) to the point (9, 9)
which never pass through the point (5, 5). By a north-east lattice path, we mean a path
whose steps are formed by going 1 unit north (0, +1) and 1 unit east (+1, 0).
Solution:
The number of such paths is the number of paths from (0, 0) to (9, 9) less the number of
paths from (0, 0) to (9, 9) which always pass through the point (5, 5). Every path that
goes from (0, 0) to (9, 9) and passes through (5, 5) corresponds to a path from
(0, 0)
to
(5, 5) and another path from (5, 5) to (9, 9). The number of such paths is 5+55
4+4
4 =
252 × 70 = 17640.
The number of paths from (0, 0) to (9, 9) is 9+9


9 = 48620.
Thus the answer is 48620 − 17640 = 30980.

5
10. Consider the sequence of numbers a1 , a2 , . . . , an , . . . defined in the following way:
a1 = a2 = 1, a3 = 251 and for all n ≥ 3
2007 + an an−1
an+1 = .
an−2

Prove that all of the numbers in this sequence are integers.

Solution: We have
an+1 an−2 = 2007 + an an−1
for all n ≥ 3. Replacing n by n − 1 gives us

an an−3 = 2007 + an−1 an−2

for n ≥ 4.
Therefore
an+1 an−2 − an an−1 = 2007 = an an−3 − an−1 an−1
for n ≥ 4.
Rearranging gives:
an−2 (an+1 + an−1 ) = an (an−1 + an−2 )
or
an+1 + an−1 an−1 + an−3
=
an an−2
for all n ≥ 4.
If n is even, we obtain
an+1 + an−1 an−1 + an−3 a3 + a1
= = ··· = = 252
an an−2 a2

and hence an+1 = 252an − an−1 .

If n is odd, we obtain
an+1 + an−1 an−1 + an−3 a4 + a2 2259
= = ··· = = =9
an an−2 a3 251

and hence an+1 = 9an − an−1 .

Either way, an argument by induction now shows that all of the an are integers.

6
 TWENTIETH IRISH MATHEMATICAL OLYMPIAD

Saturday, 12 May 2007

10 a.m.– 1 p.m.

First Paper

1. Find all prime numbers p and q such that p divides q + 6 and q divides p + 7.

2. Prove that a triangle ABC is right-angled if and only if

sin2 A + sin2 B + sin2 C = 2.

3. The point P is a fixed point on a circle and Q is a fixed point on a line. The
point R is a variable point on the circle such that P , Q and R are not collinear.
The circle through P , Q and R meets the line again at V . Show that the line
V R passes through a fixed point.

4. Air Michael and Air Patrick operate direct flights connecting Belfast, Cork,
Dublin, Galway, Limerick and Waterford. For each pair of cities exactly one
of the airlines operates the route (in both directions) connecting the cities.
Prove that there are four cities for which one of the airlines operates a round
trip. (Note that a round trip of four cities P , Q, R and S, is a journey that
follows the path P → Q → R → S → P .)

5. Let r and n be nonnegative integers such that r ≤ n.

(a) Prove that  

n + 1 − 2r n
n+1−r r
is an integer.
(b) Prove that
X n + 1 − 2r n
⌊n/2⌋

< 2n−2
r=0
n+1−r r
for all n ≥ 9.
 
n n!
(Note that = . Also, if x is a real number then ⌊x⌋ is the
r r!(n − r)!
unique integer such that ⌊x⌋ ≤ x < ⌊x⌋ + 1.)
 Solutions

1. p = 2 and p divides q + 6 together imply 2 divides q. Hence q = 2. But then

q divides p + 7 is not satisfied. Thus p is odd.
q = 2 and p divides q + 6 together imply p = 2 but p = 2 = q contradicts q
divides p + 7. Thus p and q are both odd.
p + 7 is even and so q ≤ (p + 7)/2 ≤ (q + 6 + 7)/2. Hence q ≤ 13.
Now examine each of the cases q = 3, 5, 7, 11, 13. Only one case q = 13 pro-
vides a solution in which p = 19.

Second solution: By hypothesis, there are a, b ∈ N such that q + 6 =

pa, p + 7 = bq. Hence

13 = p(a − 1) + q(b − 1) ≡ pm + qn,

where m, n are nonnegative integers. Suppose m, n ∈ N. Then

4 ≤ p + q ≤ pm + qn = 13,

and so p, q ∈ {2, 3, 5, 7, 11}. An inspection of the possible pairs (p, q) that

can be formed from this set shows that no such pair satisfies the hypotheses.
Hence, one of m, n is zero. Suppose n = 0. Then b = 1, a = 2, p = 13 which
means that q = 20, which isn’t a prime number. It follows that m = 0, i.e.,
a = 1, b = 2, q = 13 and hence p = 26 − 7 = 19. Thus p = 19, q = 13 is the
only solution pair.

2. Suppose one of ∠A, ∠B, ∠C is a right angle. Say ∠A is. Then ∠B +∠C = π/2
and so

sin2 A + sin2 B + sin2 C = 1 + sin2 B + sin2 (π/2 − B) = 1 + sin2 B + cos2 B = 2.

Suppose the identity holds. By the Sine Rule,

sin A sin B sin C 1
= = = ≡ λ.
a b c 2R
Hence
2
2 = λ2 (a2 + b2 + c2 ), λ2 = ,
a2 + b2 + c2
so that
2a2 b2 + c2 − a2 2bc cos A
sin2 A = 2 2 2
, cos2
A = 2 2 2
= 2 .
a +b +c a +b +c a + b2 + c2
Similarly,
2ca cos B 2ab cos C
cos2 B = , cos 2
C = .
a2 + b2 + c2 a2 + b2 + c2
Hence cos A, cos B, cos C are nonnegative. Suppose all are positive, then
2bc 2ca 2ab
cos A = , cos B = 2 , cos C = 2 .
a2 2
+b +c 2 2
a +b +c 2 a + b2 + c2
Hence, for instance,

(b2 + c2 − a2 )(b2 + c2 + a2 ) = (2bc)2 , (b2 + c2 )2 − a4 = 4b2 c2 ,

or
(b2 − c2 )2 − a4 = 0, (b2 − c2 − a2 )(b2 − c2 + a2 ) = 0,
i.e., cos B cos C = 0, a contradiction. Hence, one of cos A, cos B, cos C is zero.

Second solution: Let S be the circumcircle of ABC and suppose that S has
diameter d. Using the sine rule we have
1 2
sin2 A + sin2 B + sin2 C = (a + b2 + c2 ).
d2
Suppose that ABC is not a right angled triangle. There are two cases to
consider
Case I: One of the angles is obtuse. Without loss of generality, suppose that
A > π2 . It is easy to see from a diagram that a < d and that b2 + c2 < d2 .
Thus
1 2
(a + b2 + c2 ) < 2
d2
in this case.
Case II: ABC is an acute angled triangle (i.e. all angles < π2 ). Then Let X
be the point on S that is diametrically opposite to A (so |AX| = d). Let
e = |BX| and f = |CX|. Pythagoras’ Theorem implies that

b2 + e2 + c2 + f 2 = 2d2

However ∠BXC > π2 , so e2 + f 2 < a2 (a simple application of the cosine rule

to the triangle BXC). Therefore, in this case we have
1 2
(a + b2 + c2 ) > 2.
d2

Third solution: Assume

4∆2 4∆2 4∆2
2 = sin2 A + sin2 B + sin2 C = + + ,
b2 c2 c2 a2 a2 b2
 where ∆ is the area of the triangle, i.e.,
8a2 b2 c2 = 16∆2 (a2 +b2 +c2 ) = (2(a2 +b2 +b2 c2 +a2 b2 )−a4 −b4 −c4 )(a2 +b2 +c2 ).
Let x = a2 , y = b2 , z = c2 . Then
0 = 8xyz − (2(xy + yz + zx) − x2 − y 2 − z 2 )(x + y + z)
= 8xyz − (xy + yz + zx)(x + y + z) + (x2 + y 2 + z 2 − xy − yz − zx)(x + y + z)
= 8xyz − (xy + yz + zx)(x + y + z) + x3 + y 3 + z 3 − 3xyz
= 2xyz − xy(x + y) − yz(y + z) − zx(z + x) + x3 + y 3 + z 3
= x3 − x2 (y + z) − x(−2yz + y 2 + z 2 ) + y 3 + z 3 − zy(y + z)
= x3 − x2 (y + z) − x(y − z)2 + (y + z)(y − z)2
= (x + y − z)(x − y − z)(z + x − y).
It follows that one of
a2 + b2 − c2 , b2 + c2 − a2 , c2 + a2 − b2
is zero. Hence the triangle is right-angled.

Fourth solution (due to Stephen Dolan): Suppose the identity holds.

Then
0 = 2 − sin2 A − sin2 B − sin2 C
= cos2 A + cos2 B − sin2 (A + B)
= cos2 A + cos2 B − sin2 A cos2 B − 2 sin A sin B cos A cos B − cos2 A sin2 B)
= 2 cos2 A cos2 B − 2 sin A sin B cos A cos B
= 2 cos A cos B(cos A cos B − sin A sin B)
= 2 cos A cos B cos(A + B)
= −2 cos A cos B cos C,
whence one of cos A, cos B, cos C is zero. Hence the triangle is right-angled.
3. There are different diagrams possible depending on relative positions of the
circle, line and Point P . In one case, ∠P RV and ∠P QV are equal and in the
other are complementary.
∠P RV = 180−∠P QV as P QV R is a cyclic quadrilateral. S is the intersection
of V R and the circle. T is the intersection of the line P Q and the circle.
∠P RV = ∠ST P . Thus S is a fixed point on the circle and S is on V R. Hence
V R always passes through S.
The other case is similar.
Second solution: Let X be the other point where P Q meets the circle and
let W be the other point where the line through X parallel to QV meets the
circle - clearly W is a fixed point on the circle.
 •P
•R

•
W
•
X •
V
•
Q

Now P QV R is a cyclic quadrilateral. Therefore ∠P RV = 180◦ − ∠P QV .

Similarly P XW R is a cyclic quadrilateral, so ∠P RW = 180◦ − ∠P XW . But
∠P XW = ∠P QV . Therefore ∠P RW = ∠P RV . Hence R, W and V are
collinear, as required. (Note that different configurations are possible depend-
ing on which side of the line P Q the point R lies. However, the arguments are
essentially the same in all cases.)
4. The problem can be reinterpreted in graph theory terminology as the follow-
ing: If K6 is edge-coloured with two colours then there exists a monochromatic
C4 (cycle of length 4). As K6 has 15 edges one of the colours, say red, must
appear at least 8 times. Suppose the other colour is blue. Consider all the
induced graphs on 4 vertices which have 6 edges. They cannot all have the
same number of edges of the two colours as then the same would be true of
K6 . Thus there are four vertices P, Q, R, S defining a K4 which has 4,5,or 6
red edges. In the case that there are 5 or 6 red edges it is easy to find a red C4 .
If the four red edges form a C4 we are done. If not, we have P Q, QR, RP, P S
red and QS, RS blue. There are at least 4 more red edges in our K6 . Thus one
of the remaining vertices must be connected to P, Q, R, S by at least two red
edges. In all cases a red C4 appears except in the case that T P and T S are
red and T Q and T R are blue. But now T RSQ is a blue C4 . This completes
the proof.

Alternative solution with less graph theory: There are 62 = 15 routes
between the 6 cities. Thus one airline, say Air Michael, must run at least 8 of
the routes.

Choose any four cities. There are 42 = 6 routes between these four cities. If
for all choices of four cities the routes were divided 3-3 between Air Michael
and Air Patrick then they would also have the same number of routes overall
which is impossible since 15 is odd. Thus for one choice of four cities, Air
Michael must run 4 or more of the routes. If they run five or six of the routes,
there will be a round trip. If they run four and these form a round trip we are
also done. The other possibility is they run four routes which do not form a
 round trip. Calling these four cities P, Q, R, S we can assume that Air Michael
runs routes P Q, P R, P S and QR. Air Michael runs at least four more routes,
at least 3 of the routes from P, Q, R, S to the other two cities T and U. One
of these cities, say T must have two or more routes to P, Q, R, S. If there are
three routes then a round trip occurs. The only two Air Michael routes which
will not produce a round trip are T P and T S. But then QSRT will be a
round trip run by Air Patrick.

n+1−2r n


5. Define S(n, r) := n+1− r r
, for all pairs r, n with −1 ≤ r and 2r ≤ n + 1. In
particular S(n, r) = 0 if r = −1 or if 2r = n + 1. Note that S(1, 0) = 1. We
prove that if n ≥ 2 and 0 ≤ r ≤ n/2 then

S(n, r) = S(n − 1, r − 1) + S(n − 1, r) .

The first statement will then follow by induction on n. For n ≥ 2 we have

S(n − 1, r − 1) + S(n − 1, r)
(n + 2 − 2r) (n − 1)! (n − 2r) (n − 1)!
= +
(n + 1 − r) (r − 1)!(n − r)! (n − r) r!(n − 1 − r)!
[r(n + 2 − 2r) + (n − 2r)(n + 1 − r)](n − 1)!
=
(n + 1 − r)r!(n − r)!
[(n + 1 − 2r)n](n − 1)!
=
(n + 1 − r)r!(n − r)!
= S(n, r) .

P
The final statement is that ⌊n/2⌋
r=0 S(n, r) < 2
n−2
, for n ≥ 9. The base case
here is n = 9, where the values of S(9, r) for r = 0, 1, 2, 3, 4 are 1, 8, 27, 48, 42,
respectively. These sum to 126, which is less than 29−2 = 128. The inductive
step follows from the fact that

X
⌊n/2⌋
X
⌊n/2⌋
S(n, r) = [S(n − 1, r) + S(n − 1, r − 1)]
r=0 r=0

X
⌊(n−1)/2⌋
≤2 S(n − 1, r) < 2 × 2(n−1)−2 = 2n−2 .
r=0

(each S(n − 1, r − 1) with r < n/2 contributes twice to the sum: to S(n, r − 1)
and S(n, r + 1). When r = n/2, S(n − 1, n/2 − 1) contributes only once: to
S(n, n/2 − 1).)

Note: This problem is a simple case of the ‘hook formula’ for the dimension
of certain irreducible complex representations of the symmetric group Sn . The
number S(n, r) is the dimension of the Specht module corresponding to the
two part partition n = (n − r) + r. The recursion formula follows from the
‘Branching rule’ that gives the decomposition of the Specht module into Specht
modules for Sn−1 . In this case there are only two potential Specht modules for
Sn−1 , corresponding to the partitions n − 1 = (n − r) + (r − 1) and n − 1 =
(n − r − 1) + r.

The estimate for the sum of the S(n, r)’s is probably not very good asymp-
totically. It arose out of the observation that the S(n, r)’s are a truncation of
Pascal’s triangle, where the row sum is 2n .
Second Solution: Let
 
n + 1 − 2r n
S(n, r) = .
n+1−r r

(a) Then
   
n+1−r n r n
S(n, r) = −
n+1−r r n+1−r r
 
n rn!
= −
r r!(n + 1 − r)(n − r)!
 
n n!
= −
r (r − 1)!(n + 1 − r)!
   
n n
= − ,
r r−1

a difference between two successive binomial coefficients. Hence S(n, r) is an

integer, which is nonnegative if 2r ≤ n + 1. Alternatively,
   
n+1 n r n
S(n, r) = −2
n+1−r r n+1−r r
(n + 1)n! rn!
= −2
r!(n + 1 − r)(n − r)! r!(n + 1 − r)(n − r)!
(n + 1)! n!
= −2
r!(n + 1 − r)! (r − 1)!(n − (r − 1))!
   
n+1 n
= −2 ,
r r−1

a difference between multiples of two binomial coefficients.

(b) Using the first representation of S(n, r), we see that

X n + 1 − 2r n
⌊n/2⌋
X
⌊n/2⌋

= S(n, r)
r=0
n + 1 − r r r=0
⌊n/2⌋    
X n n
= [ − ]
r=0
r r−1
 
n
= .
⌊n/2⌋

It must now be confirmed that

 
n
< 2n−2 , ∀n ≥ 9.
⌊n/2⌋

We’ll use induction to establish this considering separately when n is even and
when n is odd.
Case (i): Suppose n = 2m, m ≥ 5. Since
 
n (2m)!
= ,
⌊n/2⌋ (m!)2

and  
10 10.9.8.7.6
= = 4.63 < 28 = 210−2 ,
5 5.4.3.2.1
we can start the induction at m = 5. So, assuming that

(2m)!
< 22m−2 , ∀m ≥ 5,
(m!)2

we see that
(2(m + 1))! 2(m + 1)(2m + 1) (2m)! 2m + 1 2m−1
2
= 2 2
< 2 < 22m .
((m + 1)!) (m + 1) (m!) m+1

Hence the inequality holds if n is even.

Case (ii): Suppose n = 2m + 1, m ≥ 4. Since
   
n 2m + 1 (2m + 1)!
= = ,
⌊n/2⌋ m m! (m + 1)!

and  
9 9.8.7.6
= = 2.63 < 27 = 29−2 ,
4 4.3.2.1
we can start the induction at m = 4. So, assuming that

(2m + 1)!
< 22m−1 , ∀m ≥ 4,
m! (m + 1)!

we see that
(2m + 3)! (2m + 3)(2m + 2) (2m + 1)! 2m + 3 2m
= < 2 < 22m+1 .
(m + 2)!(m + 1)! (m + 1)(m + 2) m! (m + 1)! m+2

Hence the inequality holds if n is odd.

 TWENTIETH IRISH MATHEMATICAL OLYMPIAD

Saturday, 12 May 2007

2 p.m.– 5 p.m.

Second Paper

1. Let r, s and t be the roots of the cubic polynomial

p(x) = x3 − 2007x + 2002.

Determine the value of

r−1 s−1 t−1
+ + .
r+1 s+1 t+1

2. Suppose a, b and c are positive real numbers. Prove that

r ab
a+b+c a 2 + b2 + c2 + bca + cab
≤ ≤ c .
3 3 3
For each of the inequalities, find conditions on a, b and c such that equality
holds.
3. Let ABC be a triangle the lengths of whose sides BC, CA, AB, respectively,
are denoted by a, b, c, respectively. Let the internal bisectors of the angles
∠BAC, ∠ABC, ∠BCA, respectively, meet the sides BC, CA, AB, respectively,
at D, E, F , respectively. Denote the lengths of the line segments AD, BE, CF ,
respectively, by d, e, f , respectively. Prove that
4abc(a + b + c)∆
def = ,
(a + b)(b + c)(c + a)
where ∆ stands for the area of the triangle ABC.
4. Find the number of zeros in which the decimal expansion of the integer 2007!
ends. Also find its last non-zero digit.
5. Suppose a and b are real numbers such that the quadratic polynomial

f (x) = x2 + ax + b

has no nonnegative real roots. Prove that there exist two polynomials g, h,
whose coefficients are nonnegative real numbers, such that
g(x)
f (x) = ,
h(x)
for all real numbers x.
 Solutions
1. Proposed by Stephen Buckley.
Solution
First Solution.
Call the indicated sum S. Note that S = 3 − 2R, where
1 1 1
R= + + .
r+1 s+1 t+1
Moreover r + 1, s + 1, t + 1 are the roots of the polynomial q, where
q(x) = p(x − 1) = x3 − 3x2 − 2004x + 4008.
Using theP usual formulae for the sums of products of roots of a polynomial
f (x) = ni=0 ci xi , we see that the sum of the reciprocals of the roots of f is
−c1 /c0 . In particular, R = −(−2004)/4008 = 1/2 and S = 3 − 1 = 2.
Second Solution.
Since −1 is not a root of p, we see that (r +1)(s+1)(t+1) 6= 0. The expression
in question is equal to
(r − 1)(s + 1)(t + 1) + (s − 1)(r + 1)(t + 1) + (t − 1)(r + 1)(s + 1) A(r, s, t)
≡ .
(r + 1)(s + 1)(t + 1) B(r, s, t)
Now
(r − 1)(s + 1)(t + 1) = (r − 1)(st + s + t + 1) = rst + rs + rt − st + r − s − t − 1.
Hence, permuting r, s, t, we see that
A = 3rst + (rs + st + tr) − (r + s + t) − 3.

Also,
B = (r+1)(s+1)(t+1) = (r+1)(st+s+t+1) = rst+(rs+st+tr)+(r+s+t)+1.

But
x3 −2007x+2002 = (x−r)(x−s)(x−t) = x3 −(r+s+t)x2 +(rs+st+tr)x−rst,
so that
r + s + t = 0, rs + st + tr = −2007, rst = −2002.
It follows that
A = −6006 − 2007 − 3 = −8016,
and
B = −2002 − 2007 + 1 = −4008.
Hence the value of the expression is A/B = 2.
2. Proposed by Finbarr Holland.

Solution
We first show that r
a+b+c a2 + b2 + c2
≤ ,
3 3
with equality iff a = b = c. The inequality is equivalent to saying that

(a + b + c)2 ≤ 3(a2 + b2 + c2 ) ⇔ a2 + b2 + c2 + 2(ab + bc + ca) ≤ 3(a2 + b2 + c2 ),

i.e.,

2(ab + bc + ca) ≤ 2(a2 + b2 + c2 ) ⇔ 0 ≤ (a − b)2 + (b − c)2 + (c − a)2 .

The latter clearly holds, with equality only when a = b = c.

OR use
a2 + b2 b2 + c2 c2 + a2
ab ≤ , bc ≤ , ca ≤ .
2 2 2
OR apply the CS inequality at this point or at the start.
Next, r ab bc ca
a 2 + b2 + c2 c
+ a
+ b
≤
3 3
holds iff
ab bc ca 2 a2 b2 b2 c2 c2 a2
3(a2 + b2 + c2 ) ≤ ( + + ) = 2 + 2 + 2 + 2(a2 + b2 + c2 ),
c a b c a b
i.e., iff
a2 b2 b2 c2 c2 a2
a 2 + b2 + c2 ≤ + 2 + 2 .
c2 a b
Since,
1
x+
≥ 2, ∀x > 0,
x
with equality holding iff x = 1, we see that

2 a2 b2 c2 a2 2 a2 b2 b2 c2 2 b2 c2 c2 a2
2a ≤ 2 + 2 , 2b ≤ 2 + 2 , 2c ≤ 2 + 2 ,
c b c a a b
whence
2 2 a2 b2
2 b2 c2 c2 a2
2a + 2b + 2c ≤ 2 2 + 2 2 + 2 2 ,
c a b
the claimed result. Further, equality holds if and only if a2 = b2 = c2 , i.e.,
a = b = c.
 OR setting
bc ca ab
x= , y= z= ,
a b c
the inequality becomes

xy + yz + zx ≤ x2 + y 2 + z 2

which can be treated as above. OR use CS directly.

3. Proposed by Jim Leahy.

Solution
There are various formulae for the lengths of the bisectors. The one appropri-
ate for the solution is given by
√ p
2 bc s(s − a)
d= ,
b+c
where 2s = a + b + c. One way to derive this formula is as follows. Since
1
∆(ABD) + ∆(DAC) = ∆(ABC) = ∆ = bc sin A,
2
we deduce that
A A A
d(b + c) sin = 2∆ = bc sin A = 2bc sin cos ,
2 2 2
whence
2∆ 2bc cos A2
d= = .
(b + c) sin A2 b+c
By the Cosine Rule
A 2bc + b2 + c2 − a2 (b + c)2 − a2 2s(s − a)
2 cos2 = 1 + cos A = = = ,
2 2bc 2bc bc

and so r
A s(s − a)
cos = .
2 bc
Hence √ p
2 bc s(s − a)
d= .
b+c
Alternatively, we can obtain this formula by appealing to the following facts,
which are propositions in Euclid’s Geometry, viz.,
|BD| |AB|
|AB| |AC| = |BD| |DC| + |AD|2 , and = .
|DC| |AC|
(Here’s a trigonometric derivation of these formulae. By the Sine Rule,
|BD| sin ∠BAD sin ∠CAD |DC|
= = = ,
|BA| sin ∠BDA sin ∠CDA |CA|
which is equivalent to the second formula. Next
|AB| |AC| |BD| |DC| sin ∠BDA sin ∠CDA sin ∠BAD sin ∠CAD
2
− 2
= −
|AD| |AD| sin B sin C sin B sin C
2 2 A
sin ∠BDA − sin 2
=
sin B sin C
sin2 ∠(B + A2 ) − sin2 A2
=
sin B sin C
(sin ∠(B + A2 ) − sin A2 )(sin ∠(B + A2 ) + sin A2 )
=
sin B sin C
4 cos( 2 ) sin 2 sin A+b
A+B B
2
cos B2
=
sin B sin C
sin(A + B) sin B
=
sin B sin C
= 1.
This is equivalent to the first statement.)
To continue, we have
|BD| c
bc = |BD| |DC| + d2 , and = ,
|DC| b
and since |BD| + |DC| = a, we deduce that
ca ab
|BD| = , |DC| = .
b+c b+c
Hence
a2 bc bc[(b + c)2 − a2 ] 4bcs(s − a)
d2 = bc − = = ,
(b + c)2 (b + c)2 (b + c)2
whence the same result ensues.
OR, applying the Cosine Rule to the angles ∠BAD, ∠DAC, we see that
c2 + d2 − |BD|2 b2 + d2 − |DC|2
= ,
2cd 2bd
whence
b(c2 + d2 ) − c(b2 + d2 ) = b|BD|2 − c|DC|2
ba2 c2 ca2 b2
= −
(b + c)2 (b + c)2
a2 bc(c − b)
= .
(b + c)2
 Hence
2 a2 bc(c − b)
bc(c − b) − d (c − b) = ,
(b + c)2
and so, if b 6= c,

a2 bc bc((b + c)2 − a2 ) 4bcs(s − a)

d2 = bc − 2
= 2
= ,
(b + c) (b + c) (b + c)2
as before. But this holds also if b = c, because, then, ∠ADC is a right-angle,
so that
2 2 a2
d =b − .
4
OR, perhaps more directly, the same result can be obtained by applying the
Cosine Rule to the angles ∠BDA, ∠ADC which yields
µ 2 ¶
d2 + |DC|2 − b2 d + |BD|2 − c2
=− ,
2d|DC| 2d|BD|

etc.. In any event, any one of these methods tells us that

√ p
2 bc s(s − a)
d= .
b+c

Similarly √ p √ p
2 ca s(s − b) 2 ab s(s − c)
e= , f= .
c+a a+b
Hence
p
8abc s3 (s − a)(s − b)(s − c) 8abcs∆
def = = ,
(a + b)(b + c)(c + a) (a + b)(b + c)(c + a)
by Heron’s formula.

4. Proposed by Bernd Kreussler.

Solution
For each positive integer k, we let F (k) denote the number of positive integers
less than or equal to 2007 which are divisible by k. Clearly, F (k) = b 2007k
c,
2007
the integer part of the rational number k .
Because 55 > 2007, the exponent of 5 in the decimal expansion of 2007! is
equal to
F (5) + F (52 ) + F (53 ) + F (54 ).
An easy calculation gives F (5) = 401, F (25) = 80, F (125) = 16 and F (625) =
3, hence the exponent of 5 in the decimal expansion of 2007! is equal to 500.
On the other hand, there are 1003 even numbers among the positive integers
less than 2007 which implies that the exponent of 2 in the decimal expansion
of 2007! is at least 1003. This proves that 10500 divides 2007!, but 10501 does
not divide 2007!. The decimal expansion of 2007! ends in 500 zeros.
The idea for the calculation of the digit directly in front of these 500 zeros
is the following. We calculate modulo 10 the product of all odd factors in
the product 2007!, discarding those which are divisible by 5. The product
of all even factors is equal to 1003! × 21003 and we calculate modulo 10 the
product of all those factors in 1003! which are co-prime to 5. From the factors
not considered so far, we split off the factor 5 and proceed as before. After
repeating this procedure a few times we obtain 2007! = n × 21003 × 5500 . The
calculation gives the value of n modulo 10.
The calculations modulo 10 simplify if we take into account that for each in-
teger k we have

(10k + 1)(10k + 3)(10k + 7)(10k + 9) ≡ −1 mod 10 (1)

and
9
Y
(10k + i) ≡ −6 mod 10. (2)
i=1
i6=5

Moreover, (−6)m ≡ (−1)m 6 mod 10.

The calculation is the following.
The product of all odd numbers 1, 3, . . . , 1999, co-prime to 5, consists of 200
products as in (1), so this contributes (−1)200 ≡ 1 mod 10. We also have
2001 × 2003 × 2007 ≡ 1 mod 10. As seen above, there are F (5) = 401 positive
integers below 2007 which are divisible by 5. Exactly 201 of them are odd.
Hence, the product of all positive odd numbers below 2007 is equal to

M1 × 5201 × 401!!,

where M1 ≡ 1 mod 10 and 401!! denotes the product of all positive odd
integers less or equal to 401.
Similarly, 401!! = M2 × 540 × 79!! with M2 ≡ (−1)40 ≡ 1 mod 10. In the next
step we get 79!! = M3 × 58 × 15!! with M3 ≡ (−1)8 ≡ 1 mod 10. Finally,

15!! = 1 × 3 × 5 × 7 × 9 × 11 × 13 × 15 = M4 × 52

with M4 ≡ 1 mod 10. Altogether we obtained so far

2007!! = M × 5251

with M ≡ M1 M2 M3 M4 ≡ 1 mod 10.

 To deal with the product of all even factors in 2007!, which is equal to 1003! ×
21003 , we study 1003!. This product contains 100 products as in (2), which
contributes (−1)100 6 ≡ 6 mod 10. In addition, we have to consider 1001 ×
1002 × 1003 ≡ 6 mod 10. The number of factors in 1003! which are divisible
by 5 is equal to 200 so that

1003! = N1 × 5200 × 200!

with N1 ≡ 6 mod 10.

Similarly, 200! = N2 × 540 × 40! with N2 ≡ (−1)20 6 ≡ 6 mod 10. Next, we
obtain 40! = N3 × 58 × 8! with N3 ≡ (−1)4 6 ≡ 6 mod 10. Finally, 8! = N4 × 5
with N4 = 2 × 3 × 4 × (−4) × (−3) × (−2) ≡ −6 mod 10. Therefore,

1003! = N × 5249

with N ≡ N1 N2 N3 N4 ≡ −6 mod 10. We obtain

2007! = M × N × 5500 × 21003 = M × N × 2503 × 10500

and we know M × N ≡ −6 ≡ 4 mod 10. It remains to find 2503 ≡ 8 mod 10,

which is easily seen from 24 ≡ 6 mod 10. We conclude that

M × N × 2503 ≡ 4 × 8 ≡ 2 mod 10.

Therefore, the last non-zero digit of 2007! is equal to 2.

5. Proposed by Tom Laffey.

Solution
First Solution. Since f (x) = 0 has no nonnegative real solutions, b = f (0) > 0.
If a ≥ 0, then take g(x) = f (x), h(x) = 1. Suppose that a < 0. Since
f (x) =√0 has no nonnegative roots, we must have a2 < 4b. Replacing f (x) by
b−1 f (x b), we may assume that b = 1.
Put c1 = −a and f1 (x) = f (x) = x2 − c1 x + 1. Note that c21 < 4. Put
f2 (x) = f1 (x)(x2 + c1 x + 1) = x4 − c2 x2 + 1, where c2 = c21 − 2. If c2 ≤ 0, take
g(x) = f2 (x), h(x) = g(x)/f (x) to satisfy the conclusion. If c2 > 0, then put

f3 (x) = f2 (x)(x4 + c2 x2 + 1) = (x8 − c3 x4 + 1),

k k−1
so c3 = c22 −2. Proceeding in this way, suppose that fk (x) = (x2 −ck x2 +1),
where ck = c2k−1 −2, and that ck ≤ 0, then take g(x) = fk (x), h(x) = g(x)/f (x)
to satisfy the conclusion.
If ck > 0, then put
k k−1 k+1 k
fk+1 (x) = fk (x)(x2 + ck x2 + 1) = x2 − ck+1 x2 + 1.
Suppose that ct > 0 for all positive integers t. p
√ 2 √
Then, for a positive integer n, c2n > 2, c2n−1 > 2+ r 2, cn−2 > 2+ 2 + 2, c2n−3 >
q p q p
√ 2
√
2+ 2 + 2 + 2, ...., and in general cn−s > 2+ 2 + 2 + 2 + ... + 2(s
√ p √
terms 2 under the radical signs). By the AGM inequality, 2+ 2 > 2 2 2 =
1 1
21+ 2 + 4 , and, in general,
s r q
√ 1 1 1
2 + 2 + 2 + 2 + ... + 2 > 21+ 2 + 22 + ... + 2h+1 ,

1 1 1
where there are h terms 2 under the radical sign. Hence c21 > 21+ 2 + 22 + ... +
2m+1 ,
for all positive integers m.
Since the sum of the geometric progression
1 1 1
1+ + 2+ ... + + ..
2 2 2n
is 2, while c21 < 4, this is a contradiction.
Hence ct ≤ 0 for some positive integer t, and the proof is complete.

Second Solution.
This is based on the following observation, viz., if

p(x, θ) = x2 − 2 cos θx + 1,

then

p(x, θ)p(x, θ + π) = (x2 − 2 cos θx + 1)(x2 + 2 cos θx + 1)

= x4 − 2(2 cos2 θ − 1)x2 + 1
= x4 − 2 cos 2θx2 + 1
= p(x2 , 2θ).

To continue, as in the First Solution, it suffices to consider the case f (x) =

x2 − cx + 1, where 0 < c < 2. Select θ ∈ (0, π/2), so that c = 2 cos θ. Then
f (x) = p(x, θ) and so

f (x)p(x, θ + π) = p(x2 , 2θ) = x4 − 2 cos 2θx2 + 1.

If cos 2θ ≤ 0, we stop and take g(x) = p(x2 , 2θ) and h(x) = p(x, θ + π).
Otherwise, we repeat the process and use the fact that

p(x2 , 2θ)p(x2 , 2θ + π) = p(x4 , 4θ).

We stop if cos 4θ ≤ 0, and put

g(x) = p(x4 , 4θ), h(x) = p(x, θ + π)p(x2 , 2θ + π).

Otherwise we continue. We stop at the nth stage where cos 2n θ ≤ 0, and

cos 2n−1 θ > 0, at which point we let

n−1
2Y
2n n k
g(x) = p(x , 2 θ), h(x) = p(x, θ + π) p(x2 , 2k θ + π).
k=1

It’s clear that g, h are polynomials whose coefficients are nonnegative, and that
f = g/h.
 TWENTY FIRST IRISH MATHEMATICAL OLYMPIAD

Saturday, 10 May 2008

First Paper

1. Let p1 , p2 , p3 and p4 be four different prime numbers satisfying the

equations
2p1 + 3p2 + 5p3 + 7p4 = 162,
11p1 + 7p2 + 5p3 + 4p4 = 162.
Find all possible values of the product p1 p2 p3 p4 .

2. For positive real numbers a, b, c and d such that a2 + b2 + c2 + d2 = 1

prove that
3
a2 b2 cd + ab2 c2 d + abc2 d2 + a2 bcd2 + a2 bc2 d + ab2 cd2 ≤ ,
32
and determine the cases of equality.

3. Determine, with proof, all integers x for which x(x + 1)(x + 7)(x + 8)
is a perfect square.

4. How many sequences a1 , a2 , . . . , a2008 are there such that each of the
numbers 1, 2, . . . , 2008 occurs once in the sequence, and i ∈ {a1 , a2 , . . . , ai }
for each i such that 2 ≤ i ≤ 2008?

5. A triangle ABC has an obtuse angle at B. The perpendicular at B to

AB meets AC at D, and |CD| = |AB|. Prove that

|AD|2 = |AB|.|BC| if and only if ∠CBD = 30◦ .

 Solutions And Marking Schemes

1. Solution: There are many ways to go about this. One can get upper
bounds on each of p1 , p2 , p3 , p4 and work from there. Alternatively, you
can use simple number theory ideas.
As p1 , p2 , p3 , p4 are all different there can be at most one of them even.
If they were all odd, the LHS of the first equation would be odd. Thus
one of p2 , p3 , p4 is an even prime and hence equal to 2. Now look at the
second equation and we see that p2 or p3 is 2.
Case 1. p2 = 2.

2p1 + 5p3 + 7p4 = 152

11p1 + 5p3 + 4p4 = 148
Subtracting gives 9p1 − 3p4 = −8 which is impossible.
Case 2. p3 = 2.

2p1 + 3p2 + 7p4 = 152

11p1 + 7p2 + 4p4 = 152
Subtracting 9p1 + 4p2 − 3p4 = 0. Hence p2 = 3. That leaves two linear
equations for p1 and p4 with solution p1 = 5 and p4 = 19. Thus the
only solution of the system of equations is (p1 , p2 , p3 , p4 ) = (5, 3, 2, 19).
Hence the only possible value of p1 p2 p3 p4 = 570.
Marking scheme: Complete solution 10 marks.Miscalculation or non-
calculation of product at the end of otherwise complete solution earns 9
marks. For partial solutions, 2 marks for obtaining the value of p1 , p2 , p3
or p4 and 4 marks for obtaining two values of those. For solutions that
do not find actual values of any of p1 , p2 , p3 , p4 award two marks for
useful bounds on the sizes of these.

2. Solution 1: We have

(a) (6, 0, 0, 0) Â (2, 2, 1, 1)

(b) (2, 2, 2, 0) Â (2, 2, 1, 1)
(c) (4, 2, 0, 0) Â (2, 2, 1, 1)

By Muirhead and the above majorizations we have the following

1 6(a6 + b6 + c6 + d6 ) ≥ 4(a2 b2 cd + ab2 c2 d + abc2 d2 + a2 bc2 d + a2 bcd2 +
ab2 cd2 )
2 6(a2 b2 c2 + a2 b2 d2 + a2 c2 d2 + b2 c2 d2 ) ≥ 4(a2 b2 cd + ab2 c2 d + abc2 d2 +
a2 bc2 d + a2 bcd2 + ab2 cd2 )
3 2(a4 d2 + a2 b4 + a2 d4 + b4 d2 + b2 d4 + a4 c2 + a2 c4 + b4 c2 + b2 c4 + c4 d2 +
c2 d4 + b2 a4 )
≥ 4(a2 b2 cd + ab2 c2 d + abc2 d2 + a2 bc2 d + a2 bcd2 + ab2 cd2 )
And so we have
1’ (a6 + b6 + c6 + d6 ) ≥ 46 (a2 b2 cd + ab2 c2 d + abc2 d2 + a2 bc2 d + a2 bcd2 +
ab2 cd2 )
2’ (a2 b2 c2 + a2 b2 d2 + a2 c2 d2 + b2 c2 d2 ) ≥ 64 (a2 b2 cd + ab2 c2 d + abc2 d2 +
a2 bc2 d + a2 bcd2 + ab2 cd2 )
3’ (a4 d2 + a2 b4 + a2 d4 + b4 d2 + b2 d4 + a4 c2 + a2 c4 + b4 c2 + b2 c4 + c4 d2 +
c2 d4 + b2 a4 )
≥ 2(a2 b2 cd + ab2 c2 d + abc2 d2 + a2 bc2 d + a2 bcd2 + ab2 cd2 )
Taking the linear combination (10 .) + 6(2.0 ) + 3(3.0 ) we have it that l.h.s.
of the resulting inequality becomes (a2 + b2 + c2 + d2 )3 and thus
(a2 + b2 + c2 + d2 )3 ≥ 32/3(a2 b2 cd + ab2 c2 d + abc2 d2 + a2 bc2 d + a2 bcd2 + ab2 cd2 ).
As a2 + b2 + c2 + d2 = 1 the result follows.

Solution 2: Writing the expression a2 b2 cd + ab2 c2 d + abc2 d2 + a2 bc2 d +

a2 bcd2 + ab2 cd2 as abcd(ab + bc + cd + ac + ad + bd) and given the
condition a2 + b2 + c2 + d2 = 1 we note that abcd can not exceed 1/16.
Furthermore, using non-negativity of square of (a − b), (b − c), (c − d),
(d − a), (c − a), (d − b), and the condition a2 + b2 + c2 + d2 = 1 yields
that ab + bc + cd + ac + ad + bd can not exceed 3/2, and result follows.
Solution 3: As above, we have
a2 b2 cd+ab2 c2 d+abc2 d2 +a2 bc2 d+a2 bcd2 +ab2 cd2 = abcd(ab+bc+cd+ac+ad+bd)
1
and abcd ≤ 16
(by AM-GM and a2 + b2 + c2 + d2 = 1). Now
a2 +b2 b2 +c2 2 2
ab + bc + cd + ac + ad + bd ≤ 2
+ 2
+ c +d
2
+
a2 +c2 a2 +d2 b2 +d2
2
+ 2
+ 2
 (apply AM-GM to each term of the left hand side). Now the right
hand side of the last inequality is equal to 32 (a2 + b2 + c2 + d2 ) which,
by assumption is 23 . The required inequality follows immediately from
these observations.
Marking Scheme: 10 marks for a complete solution to this question
with 2 of those marks to be awarded for outlining an inequality that
leads to a successful outcome.
3. We would like to find all pairs of integers (x, y) which satisfy
x(x + 1)(x + 7)(x + 8) = y 2 .
With z = x + 4 this equation translates into the equivalent equations
(z − 4)(z − 3)(z + 3)(z + 4) = y 2
(z 2 − 9)(z 2 − 16) = y 2
z 4 − 25z 2 + 122 = y 2 now multiply by 4
4z 4 − 100z 2 + 242 = 4y 2
¡ 2 ¢2
2z − 25 − 252 + 242 = 4y 2
¡ 2 ¢2
2z − 25 − (2y)2 = 49
¡ 2 ¢¡ ¢
2z − 25 − 2y 2z 2 − 25 + 2y = 49.
Because (x, y) is a solution if and only if (x, −y) is so, we may assume
y ≥ 0. This implies A := 2z 2 − 25 − 2y ≤ B := 2z 2 − 25 + 2y. Because
AB = 49 and A, B are integers, the table below contains all possibilities
for them. The values of y, z, x are obtained from
A+B
B − A = 4y, 2z 2 = 25 + A + 2y = 25 + and x = z − 4.
2
A B y z x
−49 −1 12 0 −4
−7 −7 0 ±3 −1, −7
7 7 0 ±4 0, −8
1 49 12 ±5 1, −9
This gives at most seven possible integers x for which x(x + 1)(x +
7)(x + 8) is a perfect square, namely
x ∈ {−9, −4, 1} with y 2 = 144 and
x ∈ {−8, −7, −1, 0} with y 2 = 0.
 and a quick check reveals that these in fact solve the given equation.
Marking scheme: Give 3 marks if all seven solutions are found, no
matter how. Give 1 mark for checking that the x found are in fact
solutions. If nothing but the four solutions with y = 0 are found,
award 1 mark only. The remaining 6 marks should be given for a
complete argument that there is no other solution. If the student does
not mention a check that the pairs (x, y) he/she found are indeed so-
lutions, at most 9 marks should be given.

4. There are several different solutions:

Solution 1: The function f : {1, . . . , 2008} → {1, . . . , 2008} given by
f (i) = ai is a bijection. Let us consider its inverse f −1 : {1, . . . , 2008} →
{1, . . . , 2008}. The condition that i ∈ {a1 , . . . , ai } for i ≥ 2 is equiv-
alent to f −1 (i) ∈ {1, . . . i} for all i ≥ 2. Therefore, there are two
choices for f −1 (2), two choices for f −1 (3) (since one of {1, 2, 3} has
been ’used already’ for f −1 (2)), two choices for f −1 (4), . . . , two choices
for f −1 (2008). Thus there are 22007 possibilities for the function f −1 .
Therefore there are 22007 such sequences.

Solution 2: We will say that a 2008-tuple is good if it has the required

properties. Let A = (a1 , . . . , a2008 ) be a good 2008-tuple. Let SA =
{i : 2 ≤ i ≤ 2008 and ai 6= i}. We will show that A 7→ SA is a
bijection between the set of good 2008-tuples and the set of subsets of
{2, . . . , 2008}.
First we observe that the function i 7→ ai is a bijection from {1, . . . , 2008}
to itself.
Lemma: If A is good then for each i = 1, . . . 2008, either ai = 1 or
ai ≥ i.
Proof: This is obvious for i = 1 and i = 2. Suppose that i ≥ 3 and
ai ∈ {2, . . . i − 1}. Property (a) implies in particular that ai 6= aj for
j = 1, 2, . . . , ai − 1. That is ,

ai ∈
/ {a1 , a2 , . . . , aai −1 }.

Therefore we must have aai = ai . Therefore i = ai (bijectivity again)

which contradicts our assumption that ai < i. 2
Now suppose that A is a good 2008-tuple and that SA = {n1 , n2 , . . . nk }
where n1 < n2 < · · · < nk . Clearly ai = i for all i ∈/ SA ∪ {1}. So for
i ∈ S ∪ {1} we must have ai ∈ S ∪ {1} also. That is to say, i 7→ ai is a
bijection SA ∪ {1} → SA ∪ {1}. Now the lemma above clearly implies
that a1 = an1 , anj = nj+1 for j = 1, . . . , k − 1 and ank = 1. In other
words A is determined by the set SA . So A 7→ SA is an injection.
Now suppose that S = {n1 , . . . , nk } ⊂ {2, . . . , 2008}. Let ai = i for all
i∈
/ S. Define a1 = n1 , anj = nj+1 for j = 1, . . . k − 1 and let ank = 1.
Then is easily seen that A = (a1 , . . . , a2008 ) is good and clearly SA = S.
Therefore A 7→ SA is a surjection. 2
Solution 3: (By induction.) We claim that, for integers n ≥ 2,
the number of bijections f : {1, . . . , n} → {1, . . . , n} such that i ∈
{1, . . . f (i)} for all i ≥ 2, is 2n−1 . Let us call such functions good.
This easily checked for n = 2. Suppose that the statement is true
for some n = 2, . . . , k where k ≥ 2. Now consider a good function
f : {1, . . . , k + 1} → {1, . . . , k + 1}. There are 2008 possible choices
for f (1). If f (1) = 1 then we must have f (i) = i for all i, so there is
one such function. If f (1) > 1, then f (i) = i for i = 2, . . . f (1) − 1.
Now it is easy to see that the number of possible such functions f is
the same as the number of good functions from {1, . . . , k − f (1)} to it-
self. The induction hypothesis implies that there are 2k−f (1)−1 of these.
Therefore, the number of good functions on {1, . . . , k + 1} is

1 + 2k−1 + 2k−2 + · · · + 2 + 1 = 2k .

This completes the induction step.

Marking Scheme: Note that + m indicates partial marks that are
additive (within reason).

• +1 for observing that i 7→ ai must be a bijection.

• +2 checking some small relevant examples (e.g. same problem

with 2008 replaced by 2 or 3).
• + 2 for observing that if a1 > 1, then a2 = 2, a3 = 3, . . . aa1 −1 =
a1 − 1 and + 2 for observing that ai = 1 or ai ≥ i for all i
• ≥ 2 for attempting to prove the general result by induction.
5. Let |BD| = a, |AC| = b, |AB| = |CD| = c. We are given b2 = ac and
need to prove ∠CBD = 30◦ .
a2 +c2 −(b+c)2
2ac
= cos(90◦ +∠CBD) = − sin(∠CBD). Hence sin(∠BCD) =
b2 +2bc−a2
2ac
.
Also sin(∠CBD) = c/a sin(∠BCD) = c/a sin(∠BCA) = c/a.c/b.
Hence
b2 + 2bc − a2 c2
=
2ac ab
Thus b +2b c−a b−2c = 0. Using b = ac we obtain (ab−2c2 )(c−a) =
3 2 2 3 2

0.
If a = c then |AD| = b + c > c + c which contradicts the triangle in-
equality. Hence ab = 2c2 and from previous calculations sin(∠CBD) =
c2 /(ab) = 1/2. Thus ∠CBD = 30◦ .
2 2
Conversely, sin(∠CBD) = b +2bc−a2ac
= 1/2 implies b2 +2bc−a2 −ac = 0.
Also sin(∠CBD) = c /ab = 1/2 implies ab = 2c2 . Substituting for
2

a we obtain b2 + 2bc − 4c4 /b2 − 2c3 /b = 0. Factorising we obtain

(b3 − 2c3 )(b + 2c) = 0. As b + 2c > 0, b3 = 2c3 = 2abc. Hence b2 = ac.
Marking scheme: 5 marks for each implication. For partial solutions
award 1 mark for knowing Cosine Rule and 1 mark for knowing Sine
Rule. Partial marks can also be given for translating the problem in to
a coordinate geometry problem.
 TWENTY FIRST IRISH MATHEMATICAL OLYMPIAD
Saturday, 10 May 2008
Second Paper

1. Find, with proof, all triples of integers (a, b, c) such that a, b and c are
the lengths of the sides of a right angled triangle whose area is a + b + c.
2. Circles S and T intersect at P and Q, with S passing through the centre
of T . Distinct points A and B lie on S, inside T , and are equidistant
from the centre of T . The line P A meets T again at D. Prove that
|AD| = |P B|.
3. Find a3 , a4 , . . . , a2008 , such that ai = ±1 for i = 3, . . . , 2008 and
2008
X
ai 2i = 2008,
i=3

and show that the numbers a3 , a4 , . . . , a2008 are uniquely determined by

these conditions.
4. Given k ∈ {0, 1, 2, 3} and a positive integer n, let fk (n) be the number
of sequences x1 , . . . , xn , where xi ∈ {−1, 0, 1} for i = 1, . . . , n, and
x 1 + · · · + xn ≡ k mod 4.
(a) Prove that f1 (n) = f3 (n) for all positive integers n.
(b) Prove that
3n + 2 + (−1)n
f0 (n) =
4
for all positive integers n.
5. Suppose that x, y and z are positive real numbers such that xyz ≥ 1.
(a) Prove that
27 ≤ (1 + x + y)2 + (1 + y + z)2 + (1 + z + x)2 ,
with equality if and only if x = y = z = 1.
(b) Prove that
(1 + x + y)2 + (1 + y + z)2 + (1 + z + x)2 ≤ 3(x + y + z)2 ,
with equality if and only if x = y = z = 1.
 Solutions And Marking Schemes

1. Assume for the moment that with a ≤ b ≤ c. So

a 2 + b2 = c2 . (1)

The condition on the area entails ab = 2(a + b + c) which implies that

4c2 = (ab)2 − 4ab(a + b) + 4(a2 + b2 + 2ab). (2)

Equations (1) and (2) imply that

ab − 4(a + b) + 8 = 0

or
ab − 4(a + b) + 16 = 8.
Thus
(a − 4)(b − 4) = 8. (3)
Now, there are only two ways to factorise 8 in positive integers. Thus,
either (a, b) = (5, 12) or (a, b) = (6, 8). Thus, there are (up to permu-
tations) two triples that satisfy the given conditions namely, (5, 12, 13)
and (6, 8, 10). 2

Marking Scheme:

• +2 for writing down the triples (5, 12, 13) and (6, 8, 10) (1 mark
each) without any proof that there aren’t any others.
• +1 for correctly formulating the problem in terms of simultaneous
diophantine equations.
• +2 for arriving at the equation ab − 4(a + b) + 8 = 0.

2. Let O be the centre of the circle T . Extend P B to meet the circle at

C. Let ∠QP A = α and let ∠BP O = β. Since |AO| = |OB| the join of
the two centres is perpendicular to AB. The join of the two centres is
also perpendicular to P Q. Thus P QkAB. Hence ∠BAP = α. Hence
|P B| = |QA|.
Since |AO| = |OB|, ∠AP O = ∠BP O = β. Hence |P C| = |P D|.
∠P CD = π/2 − β, ∠DQP = π/2 + β. Also ∠P DC = ∠P CD =
 π/2 − β.
As QDCP is a cyclic quadrilateral

∠QDP = π = (α + 2β + π/2 − β) = π/2 − (α + β)

Also ∠AQP = ∠BP Q = α + 2β. Then since ∠DQP + ∠DCP = π we

get

∠DQA = π − (π/2 − β + α + 2β = π/2 − (α + β) = ∠QDP

Hence |AD| = |QA| = |P B|.

Marking scheme: Complete solution 10 marks. For incomplete solu-
tions award up to six marks for useful observations about angles/lengths/parallel
lines/cyclic quadrilaterals.

3. Existence: Dividing both sides by 8, we require

2005
X
251 = ai+3 2i .
i=0

Now 251 = 20 − 2 + 22 + 23 + 24 + 25 + 26 + 27 . Also

−(1 + 2 + . . . + 2m−1 ) + 2m = 1, for m = 1, 2, . . .. So

251 = 20 − 2 + 22 + 23 + 24 + 25 + 26 + 27 (−1 − 2 − . . . − 21997 + 21998 ).

So a3 = a5 = a6 = a7 = a8 = a9 = a2008 = +1 and a4 = a10 = a11 =

a12 = . . . = a2007 = −1 gives a solution.
Uniqueness: More generally, for n ≥ 1, each odd integer m with −2n <
m < 2n has a unique expression as
n−1
X
m= ai 2i , where ai = ±1, for each i. (4)
i=0

We prove this by induction on n. The base case n = 1 is just the

statement that 1 = 20 and −1 = −20 .
Let n > 1 and assume the result for n−1. Then there is a unique integer
a0 = ±1 such that m − a0 ≡ 2 (mod 4) Clearly −2n < m − a0 < 2n .
 So −2n−1 < (m − a0 )/2 < 2n−1 and (m − a0 )/2 is odd. The inductive
hypothesis implies that
n−1
X
(m − a0 )/2 = ai 2i−1 , for unique ai = ±1.
i=1

This gives the expression for m. The uniqueness of the expression is

apparent from the construction. It is also a consequence of the fact
that there are 2n odd integers m with −2n < m < 2n , but only 2n
Pn−1
expressions i=0 ai 2i .
Marking Scheme: I would give 1 mark for getting the expression
for 251, and an additional couple of marks for finding the ai ’s. The
difficulty in this problem is that 2008 is a red herring. So formulating
(4) should attract most marks, say 4 or 5. Proving the existence and
uniqueness would attract the remaining marks (that bit is not difficult).

4. (a) Let Fk (n) be the collection of sequences/vectors x corresponding

to fk (n). Suppose x = (x1 , . . . , xn ) ∈ F1 (n). Then at least one of
the entries xi 6= 0. For the first such nonzero value,
P change its sign
and call
P the resulting sequence α(x). Since x ≡ 1 mod 4, we
have α(x) ≡ 3 mod 4. The map α is also an involution. Thus

f1 (n) = f3 (n). (5)

(b) Applying the same map α to the set F0 (n) − ~0 (where ~0 is the
sequence of all zero’s) gives a bijection between F0 (n) − ~0 and
F2 (n). Thus

f0 (n) − 1 = f2 (n). (6)

Since these sum residue classes partition all such sequences we

have

f0 (n) + f1 (n) + f2 (n) + f3 (n) = 3n . (7)

One more equation is required. Let An = F0 (n) ∪ F2 (n), the set of

all sequences that have an even sum, and let Bn = F1 (n) ∪ F3 (n),
the set of all sequences that have an odd sum.
 For a sequence x = (x1 , . . . , xn ), let β(x) be the sequence obtained
by finding the first element of the sequence x that is not 1, and
changing it with the third entry (i.e. −1 7→ 0 or 0 7→ −1). The
operation β(x) is well defined for x 6= (1, . . . , 1). It also changes
the parity of the sum.
Thus β : An − ~1 7→ Bn − ~1 is a bijection. The parity of n tells us
which of An and Bn the vector ~1 is in, hence
f0 (n) + f2 (n) = f1 (n) + f3 (n) + (−1)n . (8)

Combining (1), (2), (3) and (4) give

3n + 2 + (−1)n
f0 (n) = . (9)
4
Marking Scheme:
Part (a): 5 marks. Part (b) 5 marks.

Partial:
• Noticing that
fi (n + 1) = f(i−1) mod 4 (n) + fi (n) + f(i+1) mod 4 (n)
for all 0 ≤ i ≤ 3. [2 marks]

Also, the quadruple of numbers (f0 (n), f1 (n), f2 (n), f3 (n)) could
be thought of as Pascals cylinder of circumference 4 since every
entry is the sum of the 3 entries above it, where f−1 (n) := f3 (n)
and f4 (n) := f0 (n):

n f0 (n) f1 (n) f2 (n) f3 (n)

0 1 0 0 0
1 1 1 0 1
2 3 2 2 2
3 7 7 6 7
4 21 20 20 20
5 61 61 60 61
.. .. .. .. ..
. . . . .
I would suggest a mark of 1 or 2 for noticing this.
 • Realizing equation (6). [2 marks]
• Realizing equation (7). [2 marks]
• Realizing equation (8). [3 marks]
5. Solution to Part (a). One can establish this in several ways. For
instance, using the convexity of t → t2 we have
µ ¶2 µ ¶2
3 + 2(x + y + z) (1 + x + y) + (1 + y + z) + (1 + z + x)
=
3 3
(1 + x + y) + (1 + y + z)2 + (1 + z + x)2
2
≤ ,
3
with equality iff
1 + x + y = 1 + y + z = 1 + z + x,
i.e., x = y = z. Hence
(3 + 2(x + y + z))2
≤ (1 + x + y)2 + (1 + y + z)2 + (1 + z + x)2 ,
3
with equality iff x = y = z.
This also follows from an application of the Cauchy-Schwarz inequality
to the expression
3 + 2(x + y + z) = (1 + x + y) + (1 + y + z) + (1 + z + x).
To continue, by the AM–GM inequality,
√
3 + 2(x + y + z) ≥ 3 + 6 3 xyz ≥ 9,
with equality iff x = y = z = 1. Hence
81 (3 + 2(x + y + z))2
27 = ≤ ≤ (1+x+y)2 +(1+y +z)2 +(1+z +x)2 ,
3 3
with equality iff x = y = z.
Alternatively, applying the AM–GM inequality termwise we see that
p p √
(1 + x + y)2 + (1 + y + z)2 + (1 + z + x)2 ≥ (3 3 1xy)2 + (3 3 1yz)2 + (3 1zx)2
3

p p √
3
= 9( 3 x2 y 2 + 3 y 2 z 2 + z 2 x2 )
p
≥ 27 9 x4 y 4 z 4
≥ 27,
with equality throughout iff x = y = z = 1.
OR, expanding the RHS,

(1 + x + y)2 + (1 + y + z)2 + (1 + z + x)2

= 3 + 2(x2 + y 2 + z 2 ) + 2(xy + yz + zx) + 4(x + y + z)
p √ √
≥ 3 + 6 3 x2 y 2 z 2 + 6 3 xyyzzx + 12 3 xyz
≥ 3 + 6 + 6 + 12
= 27,

with equality throughout iff x = y = z = 1.

Marking Scheme: 5 for complete solution mentioning equality. 2+2
for either (i) Convexity +AM-GM ; (ii) CS+AM-GM; (iii) AM-GM
twice ; (iv) expansion + AM-GM. thrice. Any correct application of a
relevant inequality merits 2 marks.
Solution to Part (b). Now consider the inequality on the right-hand
side:

(1 + x + y)2 + (1 + y + z)2 + (1 + z + x)2 ≤ 3(x + y + z)2 .

Expanding and simplifying this leads to the followings equivalent state-

ment that

3+2(x2 +y 2 +z 2 )+2(xy+yz+zx)+4(x+y+z) ≤ 3(x2 +y 2 +z 2 )+6(xy+yz+zx),

i.e.,

3+4(x+y+z) ≤ x2 +y 2 +z 2 +4(xy+yz+zx) = (x+y+z)2 +2(xy+yz+zx).

Equivalently,
7 ≤ (u − 2)2 + 2v,
where u = x + y + z, v = xy + yz + zx. But u ≥ 3 and v ≥ 3, with
equality in both of these inequalities iff x = y = z = 1. Hence

(u − 2)2 + 2v ≥ 1 + 6 = 7,

with equality in iff x = y = z = 1. This establishes the desired result.

Marking Scheme: 5 for complete solution mentioning equality. 1
mark only for correctly deriving an equivalent inequality, e.g..

3 + 4(x + y + z) ≤ x2 + y 2 + z 2 + 4(xy + yz + zx).

An extra mark only for rewriting this as

3 + 4(x + y + z) ≤ (x + y + z)2 + 2(xy + yz + zx),

OR using one or both of the inequalities

3 ≤ x2 + y 2 + z 2 , 3 ≤ xy + yz + zx.
 TWENTY SECOND IRISH MATHEMATICAL OLYMPIAD

Saturday, 9 May 2009

First Paper.

Time allowed: Three hours.

1. Hamilton Avenue has eight houses. On one side of the street are the houses
numbered 1,3,5,7 and directly opposite are houses 2,4,6,8 respectively. An
eccentric postman starts deliveries at house 1 and delivers letters to each of
the houses, finally returning to house 1 for a cup of tea. Throughout the
entire journey he must observe the following rules. The numbers of the houses
delivered to must follow an odd-even-odd-even pattern throughout, each house
except house 1 is visited exactly once (house 1 is visited twice) and the postman
at no time is allowed to cross the road to the house directly opposite. How
many different delivery sequences are possible?
2. Let ABCD be a square. The line segment AB is divided internally at H
so that |AB|.|BH| = |AH|2 . Let E be the mid point of AD and X be the
midpoint of AH. Let Y be a point on EB such that XY is perpendicular to
BE. Prove that |XY | = |XH|.
3. Find all positive integers n for which n8 + n + 1 is a prime number.
4. Given an n-tuple of numbers (x1 , x2 , . . . , xn ) where each xi = +1 or −1, form
a new n-tuple
(x1 x2 , x2 x3 , x3 x4 , . . . , xn x1 ),
and continue to repeat this operation. Show that if n = 2k for some integer
k ≥ 1, then after a certain number of repetitions of the operation, we obtain
the n-tuple
(1, 1, 1, . . . , 1) .

5. Suppose a, b, c are real numbers such that a + b + c = 0 and a2 + b2 + c2 = 1.

Prove that
1
a2 b2 c2 ≤ ,
54
and determine the cases of equality.
 Solutions

1. Proposed by Gordon Lessells.

Considering the order in which the houses 1,2,3,4 are visited we have six pos-
sibilities
1234, 1243, 1324, 1342, 1423, 1432
For each of these we can find an ordering of 5678 which describes the order
in which the houses 5,6,7,8 are visited. In each case there turn out to be two
orderings which satisfy the criteria. Intertwining these with the ordering of
1234 gives a total of twelve delivery sequences as follows:

17283546, 18253647, 17254638, 18274536, 16382547, 18352746

16354728, 18364527, 16472538, 17452836, 16453827, 17463528

Alternatively,there are six choices for the second and eighth houses to be vis-
ited. Each of these gives rise to two possibilities.

2. Proposed by Jim Leahy

Let the square ABCD have length 2a. Let |AH| = x. Then

x2 = 2a(2a − x)

Hence
x2 + 2ax = 4a2
√
Solving we find x = 5 − 1.
Observe that 4BXY and 4BEA are similar. Hence
|BX|
|XY | = √
5
√ √
5−1
3− 5+ 2
= √
5
√
5−1
=
2
Hence |XH| = |XY |.
3. Proposed by Bernd Kreussler.

Let f (x) = x8 + x + 1. Numerical values get large very quickly:

f (1) = 3
f (2) = 259 = 7 × 37
f (3) = 6565 = 5 × 13 × 101
f (4) = 65541 = 3 × 7 × 3121.

These numbers may suggest that f (n) will be a prime number only if n = 1.
To prove this, we try to factorise the polynomial x8 + x + 1. Progress can be
made if it is suspected that x2 + x + 1 is a factor. This can quickly be tested
by using a cubic root of unity ω 6= 1. It satisfies ω 2 + ω + 1 = 0 and ω 3 = 1,
hence ω 8 = ω 2 from which we directly see f (ω) = 0. Polynomial division gives
now easily the factorisation f (x) = (x2 + x + 1)(x6 − x5 + x3 − x2 + 1).
Another way to obtain this factorisation is the following. We write x8 +x+1 =
x8 − x2 + x2 + x + 1 and observe

x8 − x2 = x2 (x6 − 1) = x2 (x3 + 1)(x3 − 1) = x2 (x3 + 1)(x − 1)(x2 + x + 1).

¡ ¢
This gives f (x) = x8 + x + 1 = (x2 + x + 1) x2 (x3 + 1)(x − 1) + 1 .
If n ≥ 2, we have n2 + n + 1 ≥ 7 and n2 (n3 + 1)(n − 1) + 1 ≥ 37, hence f (n)
is not a prime number if n ≥ 2. As f (1) = 3 is a prime number, we conclude
that n = 1 is the only positive integer for which n8 + n + 1 is a prime number.

4. Proposed by Donal Hurley.

First Solution: Use induction on k. Result clear for k = 1. Assume it is

true for some k > 1 and now consider an arbitrary n-tuple (x1 , x2 , . . . , xn ) of
length n = 2k+1 . Since xi 2 = 1 for all i, the second iteration

(x1 x2 2 x3 , x2 x3 2 x4 , . . . , xn−1 xn 2 x1 , xn x1 2 x2 )

can be written as

(x1 x3 , x2 x4 , x3 x5 , . . . , xn−1 x1 , xn x2 )
which is the result of the interlacing of the two ( n2 )-tuples

(x1 x3 , x3 x5 , . . . , xn−1 x1 ) and (x2 x4 , x4 x6 , . . . , xn x2 ) (∗∗).

The same rule can be used to obtain the fourth iteration of the original n-
tuple by interlacing the second iteration of the two ( n2 )-tuples of (∗∗), the
sixth iteration of the original by interweaving the third iterations etc.. Thus
2 j iterations (j ≥ 2) of the original n-tuple yields the same result as the
interlacing of the j th iterations of the two ( n2 )-tuples of (∗∗). But the induction
hypothesis guarantees that these iterations of the (∗∗) tuples consist only of
ones for sufficiently large j. Thus we conclude that, for j sufficiently large, 2 j
iterations of the original n-tuple gives the n-tuple

(1, 1, . . . , 1)

as required.
Second Solution: Throughout, we assume that all subscripts are read “mod-
ulo n”. For example, if n = 4, then x5 is the same as x1 , x6 means x2 , etc.
More generally, xn+i is to identified with xi .
Let (x1,r , x2,r , . . . , xn,r ) be the rth iterate. So

(x1,0 , x2,0 , . . . , xn,0 ) = (x1 , x2 , . . . , xn ),

(x1,1 , x2,1 , . . . , xn,1 ) = (x1 x2 , x2 x3 , . . . , xn x1 ),

(x1,2 , x2,2 , . . . , xn,2 ) = (x1 x22 x3 , x2 x23 x4 , . . . , xn x21 x2 )
etc...
Lemma: For all r ≥ 0,
r
Y (rj)
xi,r = xi+j .
j=0

Proof: We can prove this by induction on r. The cases r = 0 and r = 1 are

clearly true. Now suppose that r ≥ 2 and that
l
Y (jl )
xi,l = xi+j
j=0
for all l ≤ r − 1. Then

xi,r = xi,r−1 xi+1,r−1

Qr−1 (r−1 j )
Qr−1 (r−1j )
= x
j=0 i+j j=0 xi+1+j
Qr−1 (r−1 j )
Qr (r−1
j−1)
= j=0 x i+j j=1 x i+j
µ ¶
Qr−1 (r−1 j ) (j−1)
r−1

= xi j=1 xi+j xi+j xi+r

µ ¶
Qr−1 (r−1 j ) (j−1)
+ r−1
= xi j=1 xi+j xi+r
µ ¶
Qr−1 (rj)
= xi j=1 xi+j xi+r

Qr (rj)
= j=0 xi+j

as required. QED
Observe that µn−1 ¶
Y n−1
Y
(nj) (nj)
xi,n = xi xi+j xi+n = xi+j
j=1 j=1

since xi xi+n = (xi )2 = 1

Now, we complete the solution by proving...
¡2k ¢
Lemma. For all k ≥ 1 and 1 ≤ j ≤ 2k − 1, the binomial coefficient j
is
even.
Proof: There are various ways to prove this. For example, we can use the
fact that the power of 2 that divides m! is
¹ º ¹ º ¹ º
m m m
+ 2 + 3 + ...
2 2 2
¡2k ¢
Therefore, the power of 2 that divides j
is
µ¹ º ¹ º¶ µ¹ k º ¹ k º¶
k−1 j j 2 −j 2 −j
2 +···+1− + · · · + k−1 − +···+ (1)
2 2 2 2k−1
 However, the expression in (??) is strictly bigger than 0 since for all 1 ≤ j ≤
2k − 1 and 1 ≤ s ≤ k, ¹ º ¹ k º
k−s j 2 −j
2 ≥ s +
2 2s
and the inequality is strict for at least one s between 1 and k (this happens
whenever 2s does not divide j). QED
Thus, if n = 2k then xi,n = 1 for all i.

5. Proposed by Finbarr Holland.

First Solution. Since the result is trivial if one of a, b, c is zero, and not all
of the nonzero ones can have the same sign, we may suppose without loss of
generality that a, b > 0 and c < 0.
Then c = −(a + b), whence

1 = a2 + b2 + c2 = 2(a2 + ab + b2 ),

and so
1
a2 + ab + b2 = .
2
Hence, by AM–GM,
1
3ab = ab + 2ab ≤ ab + a2 + b2 = ,
2
i.e., ab ≤ 1/6, with equality if and only if
1
a=b= √ .
6

It follows that
1 1 1 2
c2 = a2 + 2ab + b2 = + ab ≤ + = ,
2 2 6 3
with equality as before. So
1 2 1
a2 b2 c2 = (ab)2 (c2 ) ≤ = ,
36 3 54
with equality if and only if
1 2
a = b = √ , c = −√ .
6 6
Finally, removing the sign restriction imposed at the outset, we see that the
result
√ holds and that there is equality√ only when two of a, b, c are equal to
±1/ 6 and the third is equal to ∓2/ 6.
Second Solution. First of all,
0 = (a + b + c)2 = a2 + b2 + c2 + 2(ab + bc + ca),
i.e.,
1
ab + bc + ca = − ,
2
and so a, b, c are the roots of the cubic
1
x3 − x − abc = 0.
2
Since this cubic is in normal form x3 − px − q = 0, and the roots of this are
real, 27q 2 ≤ 4p3 , with equality if and only if two of the roots are equal, and
the third is the negative of one of these, (*)
we may deduce that
1 1 1
27(abc)2 ≤ 4( )3 = , a2 b2 c2 ≤ ,
2 2 54
with equality as before.
Remark. Proof of (*). This follows from the following well-known statement
about the discriminant of a cubic in normal form:-
Suppose a, b, c are the roots of the cubic x3 − px − q. Then
(a − b)2 (b − c)2 (c − a)2 = 4p3 − 27q 2 .

Alternatively, an examination of the graph of a real cubic reveals that its

roots are real if and only if the product of its local extrema is nonnegative.
p A
calculus argument reveals that the extrema occur when x = ± p/3. Hence
the requirement is that
√ √ √ √ √ √
p p p p p p p p 2p p 2p p
( √ − √ − q)(− √ + √ − q) ≤ 0 ⇔ (− √ − q)( √ − q) ≤ 0,
3 3 3 3 3 3 3 3 3 3
i.e.,
4p3
− + q 2 ≤ 0,
27
with equality if and only if 0 is either a local max or a local min, in which case
the cubic has a double root, etc..
 TWENTY SECOND IRISH MATHEMATICAL OLYMPIAD

Saturday, 9 May 2009

Second Paper

Time allowed: Three hours.

1. Let p(x) be a polynomial with rational coefficients. Prove that there exists a
positive integer n such that the polynomial q(x) defined by

q(x) = p(x + n) − p(x)

has integer coefficients.

2. For any positive integer n define

E(n) = n(n + 1)(2n + 1)(3n + 1) · · · (10n + 1).

Find the greatest common divisor of E(1), E(2), E(3), . . . , E(2009).

3. Find all pairs (a, b) of positive integers, such that (ab)2 − 4(a + b) is the square
of an integer.

4. At a strange party, each person knew exactly 22 others.

For any pair of people X and Y who knew one another, there was no other
person at the party that they both knew.
For any pair of people X and Y who did not know one another, there were
exactly 6 other people that they both knew.
How many people were at the party?

5. In the triangle ABC we have |AB| < |AC|. The bisectors of the angles at B
and C meet AC and AB at D and E respectively. BD and CE intersect at
the incentre I of 4ABC.
Prove that 6 BAC = 60◦ if and only if |IE| = |ID|.

1
 Solutions
1. Proposed by Stephen Buckley.

Solution

Each term in p(x) is of the form ai xi , where ai is rational. Expanding ai xi −

ai (x + k)i , we see that k is a factor in all terms. Thus it suffices to pick k to
equal the LCM of the denominators of the numbers ai .

2. Proposed by Marius Ghergu.

Solution

Let m be the g.c.d. of E(1), E(2), E(3), . . . , E(2009).

Since m|E(1) = 2 · 3 · . . . · 11, it follows that any prime divisor of m is less than
or equal to 11. Let p be a prime number such that p|m. Since p ≤ 11 < 2009,
it follows that m|E(p) = p(p + 1)(2p + 1)(3p + 1) · · · (10p + 1). Remark that
p + 1, 2p + 1, 3p + 1, . . . , 10p + 1 are relatively prime to p, so E(p) (and thus
m) is divisible by p but not by p2 . We have thus proved that m is not divisible
by the square of any prime number.
Since m|E(1) = 2 · 3 · . . . · 11, it follows that m divides the product of all prime
numbers less than or equal to 11, that is, m|2310.
To show that m = 2310 it is enough to prove that for all n ≥ 1, the number
E(n) is divisible by 2310.
Let n ≥ 1. Then, one of the numbers n or n + 1 is divisible by 2, so 2|E(n).
Similarly, one of the numbers n, n + 1, 2n + 1 is divisible by 3 so 3|E(n). Then,
one of the numbers n, n + 1, 2n + 1, 3n + 1, 4n + 1 is divisible by 5 which yields
5|E(n). In the same manner we obtain 7|E(n) and 11|E(n). Therefore E(n)
is a multiple of 2 · 3 · 5 · 7 · 11 = 2310 and so, the g.c.d. is 2310.

3. Proposed by Bernd Kreussler.

Solution

If (ab)2 − 4(a + b) = x2 with positive integers a, b and an integer x ≥ 0, we

have x < ab. As (ab)2 − (ab − 1)2 = 2ab − 1 is odd, we even have x ≤ ab − 2.
This implies (ab)2 − 4(a + b) ≤ (ab − 2)2 = (ab)2 − 4ab + 4, from which we
obtain
ab ≤ a + b + 1. (1)
After swapping a and b if necessary, we may assume a ≤ b. If a ≥ 3, we get
ab ≥ 3b ≥ a + b + b > a + b + 1 in contradiction to (1). Hence a = 2 or a = 1.
If a = 1, we have b2 −4(b+1) = x2 , which is equivalent to (b−2−x)(b−2+x) =
8. Because (b − 2 − x) + (b − 2 + x) = 2b − 4 is even and b − 2 − x ≤ b − 2 + x,

2
 the only possibility is b − 2 − x = 2 and b − 2 + x = 4. This yields (a, b) = (1, 5)
as the only possible solution with 1 = a ≤ b.
If a = 2, we have 4b2 − 4(b + 2) = x2 , which is equivalent to (2b − 1 − x)(2b −
1 + x) = 9. Here we have two possibilities. Either 2b − 1 − x = 2b − 1 + x = 3
or 2b − 1 − x = 1, 2b − 1 + x = 9. In the first case we obtain b = 2 and in the
second b = 3. So we have shown that (a, b) = (2, 2) and (a, b) = (2, 3) are the
only possible solutions with 2 = a ≤ b.
A simple calculation verifies that the five pairs (1, 5), (5, 1), (2, 2), (2, 3) and
(3, 2) indeed satisfy the requirements of the problem.

4. Proposed by Tom Laffey.

Solution

Suppose there were n people at the party. For each person Pi at the party, let

Si = {j : Pi knows Pj }.

Fix i. We count the number of distinct pairs (j, k) such that j ∈ Si and
k ∈ Sj . There are 222 = 484 such pairs in all. There are 22 such pairs with
k = i. Suppose k 6= i. Then Pk is one of the n − 22 − 1 people different from
Pi that Pi does not know and there are 6 corresponding j for which we must
include (j, k) in our count. Hence 484 = 22 + 6(n − 23) and n = 100.

5. Proposed by Jim Leahy.

Solution

Let 6 BAC = 2α, 6 CBA = 2β and 6 ACB = 2γ. Assume first that 2α =
6 BAC = 60◦ . This implies 2β + 2γ = 120◦ , i.e. β + γ = 60◦ . Hence, 6 DIE =
6 BIC = 120◦ . Therefore, 6 BAC + 6 DIE = 180◦ and the quadrilateral EIDA

is cyclic. As AI bisects 6 BAC, the chords EI and DI subtend angles of 30◦

at the circumference of the circumcircle of EIDA. This implies |IE| = |ID|.

γa
α
β
D
B
E
C
A
DIcb 0

3
Conversely, assume |IE| = |ID|. The bisector BD divides CA in the ratio
|AB| : |BC|. This can easily be seen from the sine rule for the two triangles
4BDA and 4BCD and using that sin(180◦ − x) = sin(x).
Let |BC| = a, |CA| = b and |AB| = c. From |CD|
|DA|
= ac and |CD|+|DA| = b we
bc bc
obtain obtain |DA| = a+c . Similarly we get |AE| = a+b . Because |CA| > |AB|
bc bc
by assumption, we have b > c and so a+c > a+b , hence |DA| > |AE|.
Let D0 be the reflection of D in AI. Since |DA| > |AE|, D0 will lie between
E and B on AB. Then 4AID ≡ 4AID0 , hence |ID0 | = |ID| and 6 ID0 A =
6 ADI = 2γ + β. Since |IE| = |ID| we have |IE| = |ID 0 | from which we get
6 D 0 EI = 6 ID 0 A = 2γ + β. From 6 IEA = 2β + γ we obtain now

180◦ = 6 IEA + 6 D0 EI = 2β + γ + 2γ + β = 3(β + γ) ,

which implies β + γ = 60◦ . Since α + β + γ = 90◦ , we get α = 30◦ and so

6 BAC = 2α = 60◦ .

4
 XV Mexican Mathematical Olympiad

1. Find all 7-digits numbers which are multiples of 3 and 7, and whose digits are 3 or 7.
Solution:
Answer
3373377, 7373373, 7733733, 3733737, 7337337, 3777333.

Considering the sum of the digits mod 3, we see that there must be seven 3s, four 3s
and three 7s, or one 3 and six 7s. 7 does not divide 1111111, so the first case does not
work.
The last case can be written as 7777777 − 4 · 10k for k = 0, 1, ..., 6. So if this is a
multiple of 7, then so is 4 · 10k . But it obviously is not.
In the middle case we must have 10a + 10b + 10c + 10d ≡ 0 mod 7 for some distinct a, b,
c, d in {0, 1, 2, 3, 4, 5, 6}. We have 10a ≡ 1, 3, 2, 6, 4, 5, 1 mod 7 for a = 0, 1, 2, 3, 4, 5, 6.
So we have 14 = 6 + 5 + 2 + 1 = 6 + 4 + 3 + 1 = 5 + 4 + 3 + 2, 7 = 3 + 2 + 1 + 1, and
hence the possible a, b, c, d are {2, 3, 5, 6}, {0, 2, 3, 5}, {0, 1, 3, 4}, {1, 3, 4, 6}, {1, 2, 4, 5},
{0, 1, 2, 6}.

2. Given some colored balls (at least three different colors) and at least three boxes, the
balls are put into the boxes so that no box is empty and we cannot find three balls of
different colors which are in three different boxes. Prove that there is a box such that
all the balls in all the other boxes have the same color.
Solution:
If we take three balls of different colors, these balls are in one or two boxes only. Let
us see that we can find three balls of different colors in two different boxes.
Suppose that any three balls of different colors are always in only one box. Let A,
B, C be three balls of different colors (there are at least three different colors). They
should be in the same box. Let P be any other ball. The color of the ball P is different
of the color of at least two of the balls A, B and C, then it belongs to the same box
that contains A, B and C. Then, all the balls are in one box, and since there are at
least three boxes, there are empty boxes, which is a contradiction. Therefore there are
three balls A, B and C of different colors such that A and B are in the same box, and
C is in other box.
Now, consider a ball P that is not in the boxes that contain A, B, and C (there are
at least three boxes). Then, A, C and P are in different boxes, and therefore, P is of
the same color as A or of the same color as C. Similarly, taking B, C and P , we see
that P is of the same color as B or of the same color as C. Therefore, P has the same
color as C. Thus, any ball in a different box than the box that contains A (and B)
and C is of the same color as C. If we choose now a ball C 0 from one of these boxes
(that is, from a box different from the ones that contain A, B and C) and repeating

1
 the argument, we see that, all the balls in the same box as C are of the same color as
C 0 , and therefore, they are of the same color as C. In this way all the balls that are
not in the box that contains A and B are of the same color.

3. Let ABCD be a cyclic quadrilateral and let M be the midpoint of CD. The diagonals
meet at P . The circle through P which touches CD at M meets AC again at R and
BD again at Q. The point S on BD is such that BS = DQ. The line through S
parallel to AB meets AC at T . Prove that AT = RC.
Solution:
By the power of a point theorem, we have

CR · CP = CM 2 = DM 2 = DQ · DP = BS · DP,
CP BS CP BP
so DP = CR . But AP B and CP C are similar triangles, then DP
= AP
. Also ST is
parallel to AB, so BP
AP
= BS
AT
. Hence AT = RC.

A Q
M
T P
S R

B C

4. For positive integers n, m define f (n, m) as follows. Write a list of 2001 numbers ai ,
where a1 = m, and ak+1 is the residue of a2k mod n (for k = 1, 2, ..., 2000). Then put
f (n, m) = a1 − a2 + a3 − a4 + a5 − ... + a2001 . For which n ≥ 5 can we find m such that
2 ≤ m ≤ n2 and f (m, n) > 0?.
Solution:
The answer is n ≥ 6.
Let a1 , a2 , ..., a2001 be the list (without signs) that we get for m = 2 (where a1 = 2).
If f (2, n) ≤ 0, we take m = a2 , its associated list is a2 , a3 , ..., a2001 , a2002 . Then
f (a2 , n) = −(f (2, n) − 2) + a2002 = −f (2, n) + 2 + a2002 > 0. Therefore, if 2 does not
satisfy the condition, a2 does, whenever 2 ≤ a2 ≤ n2 . If n ≥ 9, a2 = 4 and there is
in the desired interval, therefore, for n ≥ 9 there exist m that satisfies the conditions.
Let us analyze n = 5, 6, 7, 8:
For n = 5, the only m in the interval allowed is 2, and it does not work. Its associated
list is 2, 4, 1, 1, 1, ..., and f (2, 5) = −1.
For n = 6, we take m = 2. Its list is 2, 4, 4, 4, ... and f (2, 6) = 2.

2
 For n = 7, we take m = 3. Its list is 3, 2, 4, 2, 4, 2, 4, ... and f (3, 7) = 3+2·1000 = 2003.
For n = 8, we also choose m = 3. Its list is 3, 1, 1, 1, ... and f (3, 8) = 3.

5. Let ABC be a triangle with AB < AC and ∠A = 2∠C. Let D be a point on AC

such that CD = AB. Let l be the line through B parallel to AC. Let l meet the
external bisector of ∠A at M and the line through C parallel to AB at N . Prove that
M D = N D.
Solution:
Note that ∠BAM = ∠AM B, so BM = AB. Hence M BCD is a parallelogram.
Then ∠BM D = α, where α = ∠C. Thus ∠N BC = α, and ∠ABC = 180◦ − 3α, so
∠ABN = 180◦ − 2α . But ABN C is a parallelogram, then we have that ∠DCN =
180◦ − 2α. Now CN = AB = CD, then ∠CN D = ∠CDN = α. Using again
that ABN C is a parallelogram, we conclude that ∠BN C = ∠BAC = 2α. Hence
∠DN M = α = ∠BM D. Thus DM = DN .

A
M D

B C

6. A collector of rare coins has coins of denominations 1, 2, ..., n (several coins for each
denomination). He wishes to put the coins into 5 boxes such that:
(a) In each box there is at most one coin of each denomination.
(b) Each box has the same number of coins and the same denomination total.
(c) Any two boxes contain all the denominations.
(d) No denomination is in all 5 boxes.
For which n is this possible?
Solution:
Let C1 , ..., C5 be the boxes. We introduce auxiliary boxes D1 , ..., D5 . In the box Di we
put a coin of each denomination that is missing in Ci . Note that (a) and (b) are true
for the boxes Di , but (c) and (d) change to:
(c0 ) Any two different Di0 s can not have common denominations.
(d0 ) For any denomination, there is a coin of this denomination in some Di .

3
Then, (c0 ) and (d0 ) tell us that the denominations are distribute in the boxes Di . We
will solve the problem for these boxes.
By (a), (b) and (c0 ), if each Di has t coins, then n = 5t. Here, t can not be 1 by (b).
We will prove that n can be any 5t with t ≥ 2. For t = 2 we distribute the coins as
follows:

D1 D2 D3 D4 D5
1 2 3 4 5
10 9 8 7 6

For t = 3, we distribute it as follows:

D1 D2 D3 D4 D5
1 2 3 4 5
8 9 10 6 7
15 13 11 14 12

Observe that, given a distribution of the coins in boxes that satisfy (a), (b) and (c0 ), we
can add the same value to each coin in these boxes, and the distribution satisfy the three
conditions. Also, if we have two distributions that satisfy these three conditions and
that do not have common denominations, we can join them in a unique distribution that
also satisfies the conditions. We simply put in the first box of the new distribution, the
coins that are in the first boxes of the other two distributions, in the second box of the
new distribution, the coins that are in the second boxes of the other two distributions,
and so on.
Then, we can make a distribution for t = 5 as follows: we make the distribution for
t = 3 (this uses the denominations 1, 2, ..., 15), then we make the distribution for t = 2
adding to each coin 15 ( this uses the denominations 16, 17, ..., 25); finally, we join these
two distributions.
Now, with the distribution for t = 5 we can make one for t = 7, 9, 11, .... Also, instead
of using the distributions for t = 3 and t = 2, we can use several times the one for
t = 2; in this way we obtain distributions for t = 4, 6, 8, ....

4
 XVI Mexican Mathematical Olympiad

1. The numbers 1 to 1024 are written one per square on a 32 × 32 board, so that the first
row is 1, 2, ..., 32, the second row is 33, 34, ..., 64 and so on. Then, the board is divided
into four 16 × 16 boards and the position of these boards is moved round clockwise, so
that

A B D A
goes to
D C C B

then each of the 16 × 16 boards is divided into four equal 8 × 8 parts and each of these
is moved around in the same way (within the 16 × 16 board). Then, each of the 8 × 8
boards is divided into four 4 × 4 parts and these are moved around, then each 4 × 4
board is divided into 2 × 2 parts which are moved around, and finally the squares of
each 2 × 2 part are moved around. What numbers end up on the main diagonal (from
the top left to bottom right)?
Solution:
The answer is 993, 962, ..., 63, 32, (originally the other main diagonal, from bottom to
top).
We show by induction on n that for a 2n × 2n board we get the other main diagonal
(OM D) from bottom to top. For n = 1 this is obvious. Suppose it is true for n. Now
consider a 2n+1 × 2n+1 board. After the first move, the top left quadrant is the original
bottom left quadrant which contains the bottom half of the OM D as its other main
diagonal. Similarly, the bottom right quadrant is the original top right quadrant which
contains the top half of the OM D as its other main diagonal. Hence, by induction,
the subsequent moves give the OM D from bottom to top.

2. Let ABCD be a parallelogram and let K be the circumcircle of ABD. The lines BC
and CD meet K again at E and F . Prove that the circumcenter of CEF lies on K.
Solution:
Let O be the center of K. Let AO meet K again at G. We show that G is the
circumcenter of CEF . Note that AD is parallel to BE, then AB = DE. Similarly,
AB is parallel to DF , then AD = BF . Hence the arcs AE and AF are equal and the
arcs GE and GF are equal, therefore ∠AOE = 2∠AGE = ∠EGF . Hence triangles
AOE and F GE are similar.
CF CB DA
We have CD · CF = CB · CE, so CE = CD = DE . Also ∠ADE = ∠DAB = ∠F CE.
Then, the triangles ADE and F CE are similar. Thus the figures ADEO and F CEG
are similar. But O is the circumcenter of ADE, so G is the circumcenter of F CE.

1
3. Let n be a positive integer. Does n2 have more divisors of the form 4k + 1 or of the
form 4k − 1?.
Solution:
It has more of the form 4k + 1. It is sufficient to prove the result for all odd n, because
the odd divisors of n are the same as the odd divisors of n0 , where n0 is the largest
odd factor of n. Then, we will assume n odd. We proceed by induction on n. It is
true for n = 1. Suppose p is a prime factor of n and that p2a is the highest power of
2
p dividing n2 . Let h be the number of divisors of pn2a which are congruent to 1 mod
4 and k the number which are congruent to 3 mod 4. By induction h > k. Let H
be the number of divisors of n2 which are congruent to 1 mod 4 and K the number
which are congruent to 3 mod 4. If p ≡ 1 mod 4, then H = (2a + 1)h, K = (2a + 1)k,
so H > K. If p ≡ 3 mod 4, then there are a + 1 powers of p dividing n2 which are
congruent to 1 mod 4 (namely p0 , p2 , ..., p2a ) and there are a powers of p which are
congruent to 3 mod 4 (namely p1 , p3 , ..., p2a−1 ), then H = (a + 1)h + ak = a(h + k) + h,
K = ah + (a + 1)k = a(h + k) + k, therefore H > K.

4. A domino piece has two numbers (which may be equal) between 0 and 6, one at each
end. The domino pieces may be turned around, that is, 4 5 is the same domino piece
as 5 4 . There is one domino of each type, so 28 in all. We want to form a chain in
the usual way, so that adjacent dominoes have the same number at the adjacent ends.
Dominoes can be added to the chain at either end. We want to form the chain so that
after each domino has been added the total of all the numbers is odd. For example, we
could place first the domino 3 4 , total 3 + 4 = 7. Then 1 3 , total 1 + 3 + 3 + 4 = 11,
then 4 4 , total 11 + 4 + 4 = 19.
What is the largest number of dominoes that can be placed in this way?
How many maximum-length chains are there?
Solution:
The answer is 16 and 3456, respectively.
The first domino put down must have an odd total, subsequent dominoes must have
an even total. One of the numbers of the first domino will be odd and the other even.
So any domino put next to the odd number must be odd, odd. Similarly any domino
put next to the even number must be even, even. There are 12 dominoes with one odd
and one even number, 6 with two odd numbers, and 10 with two even numbers. Thus
there are 12 choices for the first domino.
Consider first the dominoes added at the odd end. Suppose the odd number on the
first domino is a. Let the other two odd numbers (in {1, 3, 5}) be b and c. We have
two choices for the first non-double. It must be ab or ac. Ignoring doubles, the order
is then determined. For example, after ab we must put bc, then ca. Now the aa can
be put either in front of the ab or after the ca. There is only one position for each of
the other two doubles. Thus we have 4 possible ways to add the 6 odd-odd dominoes
at the odd end.
Now consider the even end. Again we start by ignoring the doubles. There are 6

2
 non-doubles: 02, 04, 06, 24, 26, 46. Each number not at one of the two ends of the
subchain of even-even non-doubles must occur an even number of times. But we have
four numbers each of which appears an odd number of times in the complete set of
six. So we must exclude at least one even-even non-double. Suppose the even number
on the odd-even domino is A and that the other even numbers are B, C, D. The first
non-double in the even-even subchain can be AB, AC or AD. There are then two
choices for the second. For example, if the first was AB, the second can be BC or
BD. Suppose the second was BC. Then the remaining choices are AB, BC, CA, AD,
DC or AB, BC, CA, AD, DB or AB, BC, CD, DA, AC. In each case we have two
choices for AA (before AB or next to CA) and two choices for the other double which
can go at an end, but only one choice for the other two doubles, so 4 ways of placing
the doubles. Hence 3 · 2 · 3 · 4 = 72 possibilities in all for the even-even subchain.
Thus the maximum length is 1 + 6 + 9 = 16, and there are 12 · 4 · 72 = 3456 maximal
length chains.

5. A trio is a set of three distinct integers such that two of the numbers are divisors or
multiples of the third. Which trio contained in {1, 2, ..., 2002} has the largest possible
sum? Find all trios with the maximum sum.
Solution:
The answer is (a, 2002 − a, 2002), where

a = 1, 2, 7, 11, 13, 14, 22, 26, 77, 91, 143, 154, 182, 286.

Let the numbers be a < b < c. There are 3 possibilities: (1) a and b are divi-
sors/multiples of c; (2) a and c are divisors/multiples of b, so a must divide b and b
must divide c; (3) b and c are both divisors/multiples of a. In case (1) a and b must
both divide c, so b ≤ 2c and a ≤ 3c , so a + b + c ≤ 11 6c < 4004. In case (2), a and b both
divide c, so this is a special case of (1). In case (3), b and c must both be multiples of
a. Hence b ≤ c − a, so a + b + c ≤ 2c ≤ 4004.
We can achieve 4004 by, for example, (a, b, c) = (1, 2001, 2002). So 4004 is the maxi-
mum sum. Evidently it can only be achieved in case (3) and only by taking c = 2002,
a + b = 2002, and a a divisor of 2002. There are 14 divisors of 2002 (apart from 2002
and 1001, which do not work, since a, b, c must be distinct), and each gives a solution.
For example, (1, 2001, 2002), (2, 2000, 2002), (7, 1995, 2002).

6. Let ABCD be a quadrilateral with ∠A = ∠B = 90◦ . Let M be the midpoint of

AB and ∠CM D = 90◦ . Denote by K the foot of the perpendicular from M to CD.
Suppose AK meets BD at P , and BK meets AC at Q. Prove that AKB = 90◦ and

KP KQ
+ = 1.
PA QB
Solution:
The quadrilateral BCKM is cyclic since the opposite angles are 90◦ . Then, ∠ABK =

3
∠M BK = ∠M CK = ∠M CD. Similarly, ∠BAK = ∠M DC, then ∠ABK +∠BAK =
∠M CD + ∠M DC = 180◦ − ∠CM D = 90◦ , for which ∠AKB = 90◦ .
Since AKB is a right triangle, the midpoint M of AB is the circumcenter of AKB,
and therefore, AM = M K. Then, the right triangles AM D and KM D have an equal
corresponding leg and share the hypotenuse M D, for which are congruent. Similarly
M KC and M BC are congruent. In particular, AD = DK and KC = BC.
Let E be the intersection point between AC and BD. By Menelaus’s theorem in the
triangle BKD with the line EQC, we obtain:
KQ BE DC
· · = 1.
QB ED KC
BE BC
Since AD and BC are parallel, ED
= AD
. Using this, KC = BC and that CD =
AD + BC, we get
KQ BC
= .
QB AD + BC
Similarly
KP AD
= ,
PA AD + BC
and the addition of the two fractions is one.

4
 XVII Mexican Mathematical Olympiad

1. Find all positive integers with two or more digits such that if we insert a 0 between
the units and tens digits we get a multiple of the original number.
Solution:
The answer is 15, 18, 45, or any multiple of 10.
Let the number be n. Let n0 be the number obtained by inserting the 0. Then n must
divide 10n − n0 . If the last digit of n is d, then 10n − n0 = 9d. Then n must divide 9d.
In particular, n must be a 2 digit number. For example if d = 9, we need a two digit
number ending in 9 that divides 81. There are none. Similarly, we check d = 8 giving
n = 18, d = 7 giving no solutions, d = 6, giving no solutions, d = 5 giving n = 15
or 45, d = 4 giving no solutions, d = 3 giving no solutions, d = 2 giving no solutions,
d = 1 giving no solutions. Finally if d = 0, then any number works.
2. Let A, B, C be three collinear points, with B between A and C. Let K1 be the circle
with diameter AB, and let K2 be the circle with diameter BC. Another circle touches
AC at B and meets K1 again at P and K2 again at Q. The line P Q meets K1 again
at R and K2 again at S. Show that the lines AR and CS meet on the perpendicular
to AC at B.
Solution:
We show that ARSC is cyclic. We have ∠P RA = ∠P BA (circle diameter AB) = P QB
(circle P BQ) = ∠SQB (same angle) = ∠SCB. Hence ∠ARS + ∠SCA = 180◦ , so
ARSC is cyclic. Let K3 be the circle ARSC. Then AR is the radical axis of K3 and
K1 , CS is the radical axis of K3 and K2 , and the perpendicular to AC at B is the
radical axis of K1 and K2 , and the three radical axes concur.
3. At a party there are n women and n men. Each woman likes r of the men, and each
man likes s of then women. For which r and s must there be a man and a woman who
like each other?
Solution:
Consider the number of pairs (W, M ), where W is a woman and M a man. If no pair
like each other, then the nr pairs (W, M ) where W likes M and the ns pairs (W, M ),
where M likes W must all be distinct. But the total number of available pairs is n2 ,
so we must have nr + ns ≤ n2 and hence r + s ≤ n.
Conversely, suppose r + s ≤ n. Label the women W1 , W2 , ..., Wn and the men
M1 , M2 , ..., Mn . Let woman Wi like men Mi+k for k = 0, 1, 2, ..., r − 1, and let man Mi
like women Wi+k for k = 1, 2, ..., s (we use the cyclic subscript convention, so Wn+1
means W1 etc). Then it is clear that no woman and man like each other.
4. The quadrilateral ABCD has AB parallel to CD. Let P be a point on the side AB
and Q on the side CD such that PAPB = DQ
CQ
. Let M be the intersection point of AQ

1
 and DP , and N be the intersection point of P C and QB. Find M N in terms of AB
and CD.
Solution:
AB · CD
The answer is M N = .
AB + CD
D P B

M N

A Q C

AM AP
Since AM P and QM D are similar triangles, we have M Q
= DQ . Also, P N B and
PN PB AP PB AM
CN Q are similar triangles, then N C = CQ . But DQ = CQ (given), then M Q
= PNNC
and hence M N is parallel to AB. Also, MCD
N
= P MP+M
M
D
= APAP
+DQ
. Similarly, M
AB
N
=
QM DQ MN MN
QM +M A
= AP +DQ . Hence CD + AB = 1.
5. Some cards have a pair of numbers written on each of them. There is just one card for
each pair (a, b) with 1 ≤ a < b ≤ 2003. Two players play the following game. Each
removes a card in turn and writes the product ab of its numbers on the blackboard. The
first player who causes the greatest common divisor of the numbers on the blackboard
to fall to 1 loses. Which player has a winning strategy?
Solution:
Consider the numbers on the board before the losing move. They must have gcd d > 1.
If d is not a prime number, then it has some proper factor k, so the card (1, k)
can still be played. Contradiction. So d must be a prime number, and every pair
(a, b) with d dividing ab must have been played. There are 2002 pairs (a, d), be-
cause a can be any of 1, 2, 3, ..., 2003 except d. Similarly, there are 2002 pairs (a, 2d),
provided that 2d ≤ 2003. However, this double-counts the pair (d, 2d). So if d
is chosen so that 2d < 2003 < 3d, then there will be 4003 possible pairs. The
first player can bring this about by playing (1, 997). Then the possible plays are
(k, 997) for k = 2, 3, ..., 996, 998, 999, ..., 2003 (2001 possibilities), and (k, 1994) for
k = 1, 2, ..., 996, 998, 999, ..., 1993, 1995, 1996, ..., 2003 (2001 possibilities). So there are
an even number of moves available and the first player will win (the only way the
second player can reduce the number of moves available is by losing).
6. Given a positive integer n, an allowed move is to form 2n + 1 or 3n + 2. The set Sn
is the set of all numbers that can be obtained by a sequence of allowed moves starting
with n. For example, we can form 5 → 11 → 35 so 5, 11 and 35 belong to S5 . We call
m and n compatible if Sm ∩ Sn is non-empty. Which members of {1, 2, 3, ..., 2002} are
compatible with 2003?.
Solution:
The answer is 166, 333, 500, 667, 1001, 1335, 1502.

2
Let D be the operation a → 2a + 1, and T the operation a → 3a + 2. Note first that D
and T commute, and they are obviously injective. Now we claim that if a = D m b = T n c,
then we can find d such that b = T n d and c = D m d. We use induction on m.
Consider first m = 1. Note that k is odd iff T k is odd. Now a = Db, so a is odd. But
a = T n c, so c is odd. Hence we can find d such that c = Dd. Then Db = DT n d, so
b = T n d, and the result is true for m = 1.
Suppose it is true for m and that a = D m+1 b = T n c. Then T n c is odd, so c is odd,
and c = De. Hence D m b = T n e. Then, by induction, we can find d such that b = T n d,
e = D m d. Then c = D m+1 d, b = T n d, as required. So the result is true for m + 1 and
hence for all m.
It follows that if a = D r T s b = D m T n c, then b and c can both be obtained from some
d. (wlog r ≥ m, so D r−m T s b = T n c. If s ≥ n, then we are home since c = D r−m T s−n b,
if not then D r−m b = T n−s c and we use the result just proved.)
Thus, if we take m to be the smallest number which can lead to n, then we have
n = D r T s m for some r, s and then the numbers which can lead to n are just D a T b m
for 0 ≤ a ≤ r, and 0 ≤ b ≤ s. (If k leads to n, then we can find d which leads to m
and k, but d cannot be smaller than m, so d = m.)
Thus if k ∈ S2003 ∩ Sn , then 2003 and n can both be obtained from some d. Working
backwards, we find that 2003 = D 2 T 166, where 166 is not odd and not 2 mod 3,
then, it cannot be obtained from anything else. Hence the only numbers that lead
to numbers derived from 2003 are those that derive from 166. We get 4333= D166,
500 = T 166, 667 = D 2 166, 1001 = DT 166, 1335 = D 3 166, 1502 = T 2 166 (the others
are all ≥ 2003).

3
 XVIII Mexican Mathematical Olympiad

1. Find all the prime number p, q and r with p < q < r, such that 25pq + r = 2004 and
pqr + 1 is a perfect square.
Solution:
We have that pqr + 1 is a perfect square, then pqr = m2 − 1 = (m + 1)(m − 1), for
some integer m.
We analyze two cases:
(i) If p | m − 1,
(a) If m + 1 = qr, then m − 1 = p and qr − p = 2, which is impossible since
qr > p2 ≥ 2p ≥ p + 2.
(b) If m + 1 = q, then m − 1 = pr, and q > pr > r, which is not possible.
(c) If m + 1 = r, then m − 1 = pq, and pq = r − 2 = (2004 − 25pq) − 2.
It follows then,
26pq = 2002 =⇒ pq = 7 · 11 =⇒ p = 7 and q = 11. Thus, r = pq + 2 = 7 · 11 + 2 = 79.

(ii) If p | m + 1,
(a) If m − 1 = qr, then m + 1 = p, but then p > qr > r, which is not possible.
(b) If m − 1 = q, then m + 1 = pr and pr − q = 2, but this is impossible since
pr > 2r > 2q > q + 2.
(c) If m − 1 = r, then m + 1 = pq, and pq = r + 2 = (2004 − 25pq) + 2.
2006
Therefore 26pq = 2006, and there is not solution since 26
is not an integer.
Thus, the only solution is p = 7, q = 11 and r = 79.

2. Find the maximum number of positive integers such that any two of them a, b (with
ab
a 6= b) satisfy that |a − b| ≥ 100 .
Solution:
Suppose that a1 < a2 < ... < an is a collection with the maximum number of integers
that satisfy the property of the problem.
It is clear that ai ≥ i, for all i = 1, ..., n.
ab
If a and b are two integers in the collection with a > b, since |a − b| = a − b ≥ 100
, we
b 100b
have that a(1 − 100 ) ≥ b, then, if 100 − b > 0, we have a ≥ 100−b .
Note that there are not two integers a and b in the collection greater than 100, in fact,
ab
if a > b > 100, then a − b = |a − b| ≥ 100 > a, which is false.

1
 Also, we have that for integers a and b less than 100, it is true that
100a 100b
≥ ⇐⇒ 100a − ab ≥ 100b − ab ⇐⇒ a ≥ b.
100 − a 100 − b

It is clear that {1, 2, 3, 4, 5, 6, 7, 8, 9, 10} is one collection with the property.

100a10 100·10 100
Now, a11 ≥ 100−a10
≥ 100−10
= 9
> 11, which implies that a11 ≥ 12.

100a11 100 · 12 1200

a12 ≥ ≥ = > 13 =⇒ a12 ≥ 14.
100 − a11 100 − 12 88
100a12 100 · 14 1400
a13 ≥ ≥ = > 16 =⇒ a13 ≥ 17.
100 − a12 100 − 14 86
100a13 100 · 17 1700
a14 ≥ ≥ = > 20 =⇒ a14 ≥ 21.
100 − a13 100 − 17 83
100a14 100 · 21 2100
a15 ≥ ≥ = > 26 =⇒ a15 ≥ 27.
100 − a14 100 − 21 79
100a15 100 · 27 2700
a16 ≥ ≥ = > 36 =⇒ a16 ≥ 37.
100 − a15 100 − 27 73
100a16 100 · 37 3700
a17 ≥ ≥ = > 58 =⇒ a17 ≥ 59.
100 − a16 100 − 37 63
100a17 100 · 59 5900
a18 ≥ ≥ = > 143 =⇒ a18 ≥ 144.
100 − a17 100 − 59 41
And since there are not more than two integers greater than 100, the maximum col-
lection has 18 numbers.
One collection with 18 integers that satisfy the condition is,

{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 14, 17, 21, 27, 37, 59, 144} .

3. Let Z and Y be the tangency points of the incircle of the triangle ABC with the sides
AB and CA, respectively. The parallel line to Y Z through the midpoint M of BC,
meets CA in N . Let L be the point in CA such that N L = AB (and L on the same
side of N than A). The line M L meets AB in K. Prove that KA = N C.
Solution:
Let N 0 be the intersection point between M N and AB. Since AZ and AY are tangent
to the incircle of ABC, then AZ = AY , and since N 0 N is parallel to ZY , we have that
AN 0 = AN .
By Menelaus’s theorem in the triangle ABC, with N 0 , M and N collinear, we get
AN 0 BM CN
· ·
N 0B M C N A
= −1, and BM = M C ⇒ CN = BN 0 .
Observe that AC = AN + N C = AN 0 + N C = (AB + BN 0 ) + N C = AB + 2N C and
that AC = AL + LN + N C = AL + AB + N C, therefore AL = N C.

2
 K

A
L
Y
Z

B
M C
N0
Applying again Menelaus’s theorem to the triangle ABC, with K, M and L collinear,
AK
we get KB · BM
MC
· CL
LA
= −1 ⇒ CL LA
= KB
KA
⇒ CL
LA
− 1 = KB
KA
− 1 ⇒ CL−LA
LA
= KB−KA
KA
⇒
CL−CN AB NL AB
LA
= KA
⇒ LA
= KA
, and N L = AB ⇒ LA = KA ⇒ KA = CN .

4. At the end of a soccer tournament in which any pair of teams played between them
exactly once, and in which there were not draws, it was observed that for any three
teams A, B and C, if A defeated B and B defeated C, then A defeated C.
Any team calculated the difference (positive) between the number of games that it won
and the number of games it lost. The sum of all these differences was 5000. How many
teams played in the tournament? Find all possible answers.
Solution:
Let n be the number of teams in the tournament, since any team played against all
the other teams exactly once, then every team played n − 1 games.
Suppose that two teams X and Y , both won k games, then if X defeated Y , X defeated
all the teams that Y defeated (by hypothesis), and then X won at least k + 1 games,
which contradicts that X and Y won k games; the same argument is used to show that
Y did not defeat X. Therefore, since there are not draws, there are not two teams with
the same number of wins. Since there are at most n possible values for the number of
games won, we have that the teams won n − 1, n − 2, ..., 1, 0 in some order.
If one team won h games, then it lost n−1−h games, for which its difference (positive)
between the number of games won and the number of games lost is |n − 1 − 2h|.
Let D be the sum of all these differences, then:
a) If n is even, n = 2m with m an integer and
D = (2m − 1) + (2m − 3) + ... + 1 + 1 + ... + (2m − 3) + (2m − 1) = 2[(2m − 1) + (2m −
3) + ... + 1] = 2m2 .
b) If n is odd, n = 2m + 1 with m an integer and
D = 2m + (2m − 2) + ... + 2 + 0 + 2 + ... + (2m − 2) + 2m = 2[2m + (2m − 2) + ... + 2] =
4[m + (m − 1) + ... + 1] = 4 m(m+1)
2
= 2m(m + 1).

3
 Since D = 5000, then:
a) 2m2 = D = 5000 ⇒ m = 50 ⇒ n = 2m = 100.
b) 2m(m + 1) = D = 5000 ⇒ m(m + 1) = 2500, but 49 · 50 < 2500 < 50 · 51, therefore
there is not solution in this case.
Thus the only possible answer is n = 100.

5. Let A and B be two circles such that the center O of B lies in A. Let C and D be the
two intersection points of the circles. Let A be a point on A and let B be a point on
B such that AC is tangent to B in C and BC is tangent to A in C. The line segment
AB meets B again in E and also meets A again in F . The line CE meets A again in
G and the line CF meets the line GD in H. Prove that the intersection point of GO
and EH is the center of the circumcircle of the triangle DEF .
Solution:

A E

F B
H


D
G

Since O is the center of B, we have CO = OD and ∠DCO = ∠CDO, and since CGDO
is cyclic, ∠CGO = ∠CDO = ∠DCO = ∠DGO, therefore GO is an internal angle bi-
sector of ∠CGD.
Since CA is tangent to B, we have ∠ACO = 90◦ , then, ∠AF O = 90◦ (this follows
since AO is the diameter of A or since ACOF is cyclic). Then OF is the perpendicular
bisector of EB, which implies EF = F B.
Now, since CA is tangent to B and CB is tangent to A, we have ∠ACE = ∠CBE = α
and ∠BCF = ∠CAF = β, then ∠CEF = ∠ACE + ∠EAC = α + β = ∠CBF +
∠BCF = ∠CF E, hence EC = CF .
Also, CADF is cyclic, then ∠DAF = ∠DCF = γ, ∠CDF = ∠CAF = β and
∠COD = 180◦ −∠DAC, and since CBDE is also cyclic, we obtain ∠DBE = ∠DCE =
θ.
Since the triangles CAE and BCF are similar by having corresponding equal angles,
AE CE AE
then CF = BF ⇒ CE = CF
BF
. Taking the power of E with respect to A, we get
AE
AE · EF = CE · EG ⇒ CE = EG EF
CF
, and then BF = EG
EF
⇒ EG
CF
EF
= BF = 1 ⇒ EG = CF .

4
 In the triangle ABC we have 180◦ = ∠CAB +∠ABC +∠BCA = β +α+(β +γ +θ+α),
then ∠COD = 180◦ −∠DAC = 180◦ −(∠CAF +∠F AD) = 180◦ −(β +γ) = 2α+β +θ;
and on the other hand ∠COD = 2∠CBD = 2(∠CBE + ∠EBD) = 2(α + θ), then
2α + β + θ = 2α + 2θ ⇒ β = θ. ⇒ ∠CDF = ∠DCE ⇒ F D k CG.
Now, since F D and CG are parallel, DF CG is an isosceles trapezoid (since it is also
cyclic), and then we have GD = CF and HD DG
= HF
FC
⇒ HD = HF ⇒ HG = HC,
therefore the triangles CHG and F HD are isosceles. Moreover, since GE = EC, then
HE is the perpendicular bisector of GC, and since F DGC is an isosceles trapezoid,
HE is also a perpendicular bisector of F D.
Finally, since GD = CF = EG, the triangle EGD is isosceles, and since GO is an
internal angle bisector of ∠EGD, then GO is also a perpendicular bisector of ED,
therefore, the intersection of GO and EH is the circumcenter of the triangle DEF .

6. What is the maximum number of possible change of directions in a path traveling on

the edges of a rectangular array of 2004 × 2004, if the path does not cross the same
place twice?.
Solution:
Let R be a path with the maximum number of possible change of directions. We
can assume without loss of generality that it started in a vertical edge. We color
the horizontal edges of the rectangular array, black and white in alternating fashion,
starting with black. Then each vertex of the array is either white or black.
For the path R, we can make a list L(R) of the colors of the vertices that the path
visit, for instance: BBN N BN N N B (where B is white, and N is black).
Observe that R has a change of direction in a vertex if and only if the adjacent vertices
to it in R have different color.
We separate L(R) in two lists: L1 (R) formed by the odd positions of L(R), and L2 (R)
formed by the even positions of L(R). Then, the number of change of directions of R
is the sum of the change of color of L1 (R) and L2 (R); moreover, by the way we color
the array, some of the lists start with B, suppose that this is L1 (R).
Let i be the number of change of color in L1 (R) and let j be the number of change
of directions in L2 (R), then the number of B in L1 (R) is at least i+1
2
and in L2 is at
j i+j+1
least 2 , therefore, the number of B’s in L(R) is at least 2 , then since 2004·2005
2
is
i+j+1 2004·2005
the number of B in all the array, 2 6 2
⇒ i + j 6 2004 · 2005 − 1.

5
Alberta High School
Mathematics Contest

2006/2007-2013/2014
 Alberta High School Mathematics Competition
First Round November 21, 2006.

1. The value of 24 48 816 is

(a) 216 (b) 252 (c) 268 (d) 284 (e) none of these

2. The number of non-congruent rectangles with integer sides and area 2006 = 2 × 17 × 59 is

(a) 3 (b) 4 (c) 6 (d) 8 (e) none of these

3. The number of pairs (m, n) of positive integers such that m4 + n = 100000001 is

(a) 100 (b) 101 (c) 200 (d) 201 (e) none of these

4. In a city, all streets run north-south or east-west, dividing the city into squares. A, B, C and
D are four students who live at four street intersections which define a rectangle, with A and
C at opposite corners of this rectangle. They all go to the same school, which is at some street
intersection within this rectangle. Each goes to school by a most direct route along streets.
A travels 10 blocks, B 20 blocks and C 50 blocks. The number of blocks D travels is

(a) 10 (b) 20 (c) 30 (d) 40 (e) 50

5. When (1 + x)(1 + x2)(1 + x3)(1 + x4 )(1 + x5 ) is expanded, the coefficient of the term x9 is

(a) 0 (b) 1 (c) 2 (d) 3 (e) 4

6. On a plane are two points A and B at a distance 5 apart. The number of straight lines in
this plane which are at a distance 2 from A and at a distance 3 from B is

(a) 0 (b) 1 (c) 2 (d) 3 (e) 4

7. A country has three provinces, each province has three cities, each city has three wards and
each ward has three electors. In a two-way election, a candidate wins a ward by getting more
votes in the ward, wins a city by winning more wards in the city, wins a province by winning
more cities in the province, and wins the election by winning more provinces. Only electors
may vote, and they must vote. The minimum number of votes needed to guarantee winning
the election is

(a) 41 (b) 54 (c) 66 (d) 81 (e) none of these

8. The teacher asked, “What is the largest possible diameter of a circular coin of negligible
thickness which may be stored in a rectangular box with inner dimensions 7 × 8 × 9?” Ace
said less than 8, Bea said 8, Cec said strictly between 8 and 9, Dee said 9 and Eve said more
than 9. The one who was right was

(a) Ace (b) Bea (c) Cec (d) Dee (e) Eve
 9. If a and b are two real numbers which satisfy a + b − ab = 1 and a is not an integer, then b

(a) is never an integer (b) must be some positive integer

(c) must be some negative integer (d) must be equal to 0

(e) may either be an integer or a non-integer

x y z
10. The non-zero numbers a, b, c, x, y and z are such that = = . The value of
a b c
xyz(a + b)(b + c)(c + a)
is
abc(x + y)(y + z)(z + x)
1
(a) 0 (b) 3
(c) 1 (d) 3
x y z
(e) dependent on the common value of = =
a b c
q √ √ √ √
3
11. If k is a rational number such that 9 3 − 11 2 = 3 + k 2, then the value of k is

(a) −2 (b) −1 (c) 1 (d) 2 (e) none of these

12. All positive integers which can be expressed as a sum of one or more different integer powers
of 5 are written in increasing order. The first three terms of this sequence are 1, 5 and 6. The
fiftieth term is

(a) 3751 (b) 3755 (c) 3756 (d) 3760 (e) 3761
√ √
13. Consider all polynomials whose coefficients are all integers, whose roots include 23 and 32
and whose degree is as small as possible. Among the coefficients of these polynomials, the
smallest positive coefficient is

(a) 1 (b) 6 (c) 35 (d) 36 (e) none of these

14. Colleen used a calculator to compute a+b

c
, where a, b and c are positive integers. She pressed
a, +, b, /, c and = in that order, and got the answer 11. When she pressed b, +, a, /, c
and = in that order, she was surprised to get a different answer 14. Then she realized that
the calculator performed the division before the addition. So she pressed (, a, +, b, ), /, c
and = in that order. She finally got the correct answer, which is

(a) 4 (b) 5 (c) 20 (d) 25 (e) none of these

15. The base of a tetrahedron is an equilateral triangle of side 1. The fourth vertex is at a distance
1 above the centre of the base. The radius of the sphere which passes through all four vertices
of the tetrahedron is
√ √
1 3 1 3 2
(a) 3
(b) 6
(c) 2
(d) 3
(e) 3

16. If p(x) is a polynomial of degree 4 such that p(−1) = p(1) = 5 and p(−2) = p(0) = p(2) = 2,
then the maximum value of p(x) is

(a) 5 (b) 6 (c) 7 (d) 10 (e) none of these

 Alberta High School Mathematics Competition
Solution to the First Round, 2006.

1. The value of 24 48 816 is 24+8×2+16×3 = 268. The answer is (c).

2. There are 4 such rectangles, namely, 1 × 2006, 2 × 1003, 17 × 118 and 34 × 59. The answer
is (b).

3. Since 1004 = 100000000, (100,1) is a desired pair. In fact, m can be any positive integer up to
100, and there is a unique positive integer n such that m4 + n = 100000001 for this particular
m. Hence the total number of pairs is 100. The answer is (a).

4. A and C together travel as many north-south blocks as B and D together. A and C together
also travel as many east-west blocks as B and D together. Hence A and C together travel
as many blocks as B and D together. It follows that the number of blocks D travels is
10 + 50 − 20 = 40. The answer is (d).

5. We are counting the number of ways of making 9 from 1, 2, 3, 4 and 5, using each number at
most once. There are 3 ways, namely 1+3+5, 2+3+4 and 4+5. The answer is (d).

6. Draw a circle with centre A and radius 2. Draw another circle with centre B and radius 3.
The lines we seek are the common tangents of these two circles, of which there are 3. The
answer is (d).

7. We can steal the election by winning as little as 16 votes. We can win 2 provinces, 2 cities
within each of these 2 provinces, 2 wards within each of these 4 cities, and 2 electors within
each of these 8 wards. Hence (81 − 16) + 1 = 66 votes are required to guarantee winning the
election. The answer is (c).

8. Let A and B be the respective midpoints of two opposite edges of length 9. Let C and D be
the respective centres of the two 7 × 8 faces. (See the daigram below.) Then AB and CD
intersect at the centre O of the box. Place a coin of diameter 9 on the plane determined by
AB and CD with centre O.√Its circumference will pass through C and D, but A and B are
not covered up since AB = 72 + 82 > 9. Rotate the coin about AB so that it just comes off
C and D. We can then expand the coin slightly and still have it fit inside the box. The answer
is (e). We do not know the actual maximum value. That is why the question is phrased in its
current form.
.
.......................
............. .....
..............
............. .....
............. .....
. .
........ . . ....... ....
........
....
C
..... ......................................................................................................................................................
. ...
.....
.
..
. ..... ........
.... .... .
.
. ..
.... ...
.... ...... .
.
. ...... ....
... ...
. ..
. .
... ..
... .... ..
. .... ...
.... ..
. ...
. ... ....
... .. .
. . ..
.
... ... .... ...
.. ...
.... ... ... .. ...
... .. ... . ..
....................................................................................................................................................
A ....
..
....
...
O
...
.
.
....
.
...
...
.... .. .... .. ...
...
...
..
..
..
.
.
..
.
. .
..
.
.
..
. ...
...
B
... .. .
. .
. ...
...
... ... .... ..
.
. . ...
... .... .
. .. ...
... .... .
.
. ..... ...
..... ...
... ....... .... .. .
... ..
...............................................................................................................................................................................

D
 9. Solving for b, we have b(1 − a) = 1 − a. Since a is not an integer, 1 − a 6= 0 and can be
cancelled. Hence b = 1, and is a positive integer. The answer is (b).

10. Let xa = yb = zc = r. Then xyz

abc
= r3 , x = ra, y = rb and z = rc. It follows that x+y = r(a+b),
y + z = r(b + c) and z + x = r(c + a), so that x+y
a+b
= y+z
b+c
= z+x
c+a
= r. The given expression is
3 1 3
equal to r ( r ) = 1. The answer is (c).
√ √ √ √ √ √ √ √
11. We have 9 3 − 11 2 = ( 3 + k 2)3 = 3 3 + 9k 2 + 6k 2 3 + 2k 3 2. From 9 = 3 + 6k 2 ,
we have k = ±1. From −11 = 9k + 2k 3 , k < 0. Hence k = −1. The answer is (b).

12. We change the base-ten number 50 to base-two and obtain the number 110010, representing
25 + 24 + 22 . We now interpret this as a base-five number, representing 55 + 54 + 52 . Changing
this number to base-ten, we obtain 3755. The answer is (b).
√ √ √ √
13. Since 32 and 23 are roots, we must also have − 32 and − 23 as roots in order to have rational
coefficients. Thus the degree of the polynomial cannot be less than 4, but 4 is sufficient. Of
these polynomials, the one with 1 as the leading coefficient is
√ ! √ ! √ ! √ !   
2 2 3 3 2 2 2 3 35 1
x− x+ x− x+ = x − x − = x4 − x2 + .
3 3 2 2 9 4 36 6

Since we want integral coefficients, we clear the denominators to obtain 36x4 − 35x2 + 6. Since
the coefficients are relatively prime (though not pairwise relatively prime) and the smallest
absolute value among the coefficients is 6, the smallest positive coefficient is also 6. The
answer is (b).

14. We have a + bc = 11 and b + ac = 14. Adding the two equations yields (a + b) c+1 c
= 25 or
(a + b)(c + 1) = 25c. Since c + 1 and c are relatively prime, c + 1 must divide 25. Hence c = 4
or c = 24. If c = 24, then a + b = 25 but at least one of ac and bc is not an integer. Hence
c = 4, a + b = 20 and a+bc
= 5. The answer is (b).

15. Let the base of

√ the pyramid be √ the equilateral triangle ABC. Let D be the midpoint of BC.
Then AD = AB − BD = 23 . Let G be the centre of ABC. Then G lies on AD and
2 2
√
AG = 2AD3
= 3
3
. Let V be the fourth vertex of the pyramid and O be the centre of the sphere
passing through all four vertices. Then O lies on V G and let r = V Oq= OA be the radius of
√
the sphere. Since V G − V O = OG = OA2 − AG2 , we have 1 − r = r2 − 13 . Squaring both
sides, we have 1 − 2r + r2 = r2 − 13 . It follows that r = 23 . The answer is (e).

16. Note that p(−2) < p(−1) > p(0) < p(1) > p(2). Hence the graph of this fourth degree
polynomial opens down. Let p(x) = −x4 + ax3 + bx2 + cx + d. Then d = p(0) = 2. From
p(±2) = 2, we have −16±8a+4b±2c+2 = 2. Hence 4b−16 = ±(8a+2c). This is only possible
if both sides are equal to 0, so that b = 4. Similarly, from p(±1) = 5, we have 25a + 5c = 0.
Combining with 8a + 2c = 0, we have a = c = 0, so that p(x) = −x4 + 4x2 + 2 = 6 − (x√
2
− 2)2 .
It follows that the maximum value of p(x) is 6, occurring when x2 − 2 = 0 or x = ± 2. The
answer is (b).
 Alberta High School Mathematics Competition
Second Round February 7, 2007.

Problem 1.
Determine all positive integers n such that n is divisible by any positive integer m which satisfies
m2 + 4 ≤ n.
Problem 2.
The numbers 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 and 15 are arranged to form a 5 × 3 table
in each of the 15! possible ways. For each table, we compute the sum of the three numbers in
each row, and record in a list the largest and the smallest of these sums. Determine the sum of the
2 × 15! numbers on our list.
Problem 3.
One angle of a triangle is 36◦ while each of the other two angles is also an integral number of degrees.
The triangle can be divided into two isosceles triangles by a straight cut. Determine all possible
values of the largest angle of this triangle.
Problem 4.
1 − a3 1 − b3 1 − c3
Let a, b and c be distinct non-zero real numbers such that = = . Determine
a b c
all possible values of a3 + b3 + c3 .
Problem 5.
A survey in Alberta was sent to some teachers and students, a total of 2006 = 2 × 17 × 59 people.
Exactly a% of the teachers and exactly b% of the students responded, yielding an overall response
rate of exactly c%, where a, b and c are integers satisfying 0 < a < c < b < 100. For each possible
combination of values of a, b and c, determine the total number of teachers and the total number
of students who responded to the survey.
 Alberta High School Mathematics Competition
Solutions and Comments to the Second Round, 2007.

Problem 1.
This problem really consists of two parts, finding values for n and proving that there are no more.
Most contestants got somewhere with the first part, but many faltered in the second.
For n = 1, 2, 3 and 4, there are no positive integers m such that m2 + 4 ≤ n. Hence these four
values have the desired property vacuously. While not an essential part of the problem, these values
should be included for completeness.
If the maximum value of m is 1, then 12 + 4 ≤ n < 22 + 4 and n = 5, 6 or 7. Since 1 divides
all of them, these three values have the desired property. If the maximum value of m is 2, then
22 + 4 ≤ n < 32 + 4 and n = 8, 9, 10, 11 or 12. Of these, only 8, 10 and 12 are divisible by both 1
and 2. If the maximum value of m is 3, then 32 + 4 ≤ n < 42 + 4 and n = 13, 14, 15, 16, 17, 18
or 19. Of these, only 18 is divisible by all of 1, 2 and 3. If the maximum value of m is 4, then
42 + 4 ≤ n < 52 + 4 and n = 20, 21, 22, 23, 24, 25, 26, 27 or 28. Of these, only 24 is divisible by
all of 1, 2, 3 and 4. It will turn out that no positive integers other than 1, 2, 3, 4, 5, 6, 7, 8, 10, 12,
18 and 24 have the desired property.
For the second part, Jeffrey Mo of William Aberhart High School argued as follows. Suppose the
maximum value of m is k for some integer k ≥ 5. Then k 2 + 4 ≤ n < (k + 1)2 + 4. In order for n to
be divisible by just k − 1 and k, it has to be a multiple of k(k − 1) since k − 1 and k are relatively
prime. Now k(k − 1) < k 2 + 4 while 2k(k − 1) − ((k + 1)2 + 4) = k 2 − 4k − 5 = (k + 1)(k − 5) ≥ 0
for k ≥ 5. Hence n cannot be a multiple of k(k − 1), so that there are no solutions for m ≥ 5.
Jerry Lo of Ross Sheppard High School argued as follows. Suppose we have solutions n for some
m ≥ 5. Then n must be a multiple of m. Now m2 +4 < m2 +m < m2 +2m < (m+1)2 +4 ≤ m2 +3m,
with equality holding in the last case only for m = 5. If n = m2 + 3m, then we must have m = 5
so that n = 40, but 40 is not divisible by 3. Hence n = m(m + 1) or m(m + 2). Note that n must
also be a multiple of m − 1. Note that m − 1 and m are relatively prime. If n = m(m + 1), then
m − 1 must divide m + 1 = (m − 1) + 2. Hence it must divide 2, so that m ≤ 3. If n = m(m + 2),
then m − 1 must divide m + 2 = (m − 1) + 3. Hence it must divide 3, so that m ≤ 4. Either case
contradicts m ≥ 5. Hence there are no solutions n for m ≥ 5.
Problem 2.
Far too many contestants did not know that the total number of tables is 15!. For those who did,
the majority merely observed that the maximum sum of a row is 13+14+15=42 and the minimum
sum is 1+2+3=6. From these, they concluded that the total of the two sums must be 48, in that
if the maximum sum drops, the minimum sum would rise and compensate. While this may be a
loose description of what is the case, it did not explain why this is the case. The argument really
rests on one simple fact. The following solution by Linda Zhang of Western Canada High School
is typical of those of the top contestants.
For each table A, there is a table B which may be obtained from A by subtracting each number in
A from 16. Note that A and B are distinct tables. Now the row in A with the largest sum turns
into the row in B with the smallest sum, and the row in A with the smallest sum turns into the row
in B with the largest sum. The largest row sum of A plus the smallest row sum of B is 48, as is the
largest row sum of B plus the smallest row sum of A. Since the 15! tables may be divided into 15!2
such pairs, the sum of the 2 × 15! numbers on our record is 48 × 15!.
Problem 3.
This turned out to be the problem in which most contestants could make some progress. However,
many approached it haphazardly, and managed to find only some of the answers. Others found
all the answers but did not prove that there are no more. We give the solution by Jarno Sun of
Western Canada High School.
Let ABC be the triangle. Let 6 ABC = 36◦ . We may assume that 6 CAB ≥ 6 BCA. Then
◦ ◦
6 CAB ≥ 180 −36 = 72◦ > 6 ABC. In order for ABC to be divided into two triangles with a
2
straight cut, the cut must pass through a vertex. We consider three cases:
Case 1. The cut passes through B.
Let the cut meet CA at E. Since 6 CAB > 6 ABC > 6 ABE, 6 BEA must be one of the equal
angles in triangle BEA. It follows that 6 BEA is acute so that 6 BEC is obtuse. (See the diagram
below.) Let 6 EBC = 6 BCA = x◦ . Then 6 BEA = 2x◦. We consider two subcases:
Subcase 1a. 6 BEA = 6 CAB.
Then 6 ABE = 180◦ − 4x◦ and 36◦ = 6 ABC = (180◦ − 4x◦ ) + x◦. This yields x = 48 but then
6 ABE = −12◦ . This is impossible.

Subcase 1b. 6 BEA = 6 ABE.

Then 6 ABE = 2x◦ and 36◦ = 6 ABC = 2x◦ + x◦ . This yields x = 12. It follows that ABC is a
(132◦ , 36◦ , 12◦ ) triangle. A ...
... ..............................
................ E
.... ................
....
.... .
...
. ......................... ... ............. ................
.... ............... ................
.... ................ ................
.... ................ ................
....... ............................. ................
................
....................... ......
.. ............................................................................................................................................................................................................................................................
..
.
..

B C
Case 2. The cut passes through A.
Let the cut meet BC at D. We consider three subcases:
Subcase 2a. 6 BDA = 6 ABC = 36◦
◦
Then 6 ADC is obtuse. (See the first diagram below.) We must have 6 BCA = 6 CAD = 362 = 18◦ ,
so that ABC is a (126◦ , 36◦ , 18◦ ) triangle.
Subcase 2b. 6 BAD = 6 ABC = 36◦ .
◦
Then 6 BDA = 108◦ . If AD = CD, then 6 DAC = 6 BCA = 108 2
= 54◦ and 6 CAB = 90◦ . (See
◦ ◦ ◦
the second diagram below.) It follows that ABC is a (90 , 54 , 36 ) triangle. If AD = AC, then
6 BCA = 6 ADC = 72◦ and 6 CAB = 180◦ − 36◦ − 72◦ = 72◦ . (See the third diagram below.) It

follows that ABC is a (72◦ , 72◦ , 36◦ ) triangle. Finally, if AC = CD, then 6 CAD = 6 ADC = 72◦ .
Hence 6 BCA = 36◦ and 6 ABC = 108◦ . (See the fourth diagram below.) It follows that ABC is a
(108◦ , 36◦ , 36◦ ) triangle.
Subcase 2c. 6 BAD = 6 BDA = 72◦ .
◦
Then 6 ADC is obtuse. We must have 6 BCA = 6 CAD = 722 = 36◦ and ABC is again a
(108◦ , 36◦ , 36◦ ) triangle.
A.......
.... ...............
A ...
..........
A ..
.........
A ....
....... ....
... ..... ........... ... . .. . ..... .. ........
.... .... ........... .... ... ..... .... .. ..
.... .... ........... ..... .... .... ... ... .... ..
.... ....
....
.... ..... ........... .....
... .... .... ... .... .....
..... .... ...........
....
. ... ..... .. .... . ....
... .... .......... ..
. ... ... .
..... .
.. .. ... .. ....
..... .... ...........
........... ..... .. ... .... .. .. ..... ..
. ....
... . ..... .. ... . .
.. ... ..... ... . ... . .
.. ....
.....................................................................................................................................................................
... . ..............................................................................
.. . ................................................................. .....................................................................................................
... .

B D C B D C B D C B D C
Case 3. The cut passes through C.
Let the cut meet AB at F . Since 6 CAB ≥ 6 BCA > 6 ACF , 6 AF C must be one of the equal
angles in triangle AF C. It follows that 6 CF B is obtuse. It follows that 6 BCF = 6 ABC = 36◦
and 6 AF C = 72◦ . Since 6 CAB ≥ 6 BCA, ABC is again a (72◦ , 72◦ , 36◦ ) triangle.
In summary, the largest angle of ABC, namely 6 CAB, may be 72◦ , 90◦ , 108◦ , 126◦ or 132◦ .
Problem 4.
With greater reliance on graphing calculators and computer software, the majority of students
nowadays are very uncomfortable with algebraic manipulations. A problem such as this has become
inaccessible to most contestants.
Brett Baek of Western Canada High School used the following approach.
3 3
From 1−aa
= 1−bb
, we have b − a3b = a − ab3. Hence a − b = ab(b2 − a2) so that ab(a + b) = −1.
Similarly, bc(b + c) = −1. From these two equations, we have a2 − c2 = bc − ab or (a + c)(a − c) =
−b(a − c). Since a 6= c, we have a + b + c = 0. Hence abc(a + b) = −c = a + b. Since a + b + c = 0
but c 6= 0, a + b 6= 0 and we have abc = 1. Now

0 = (a + b + c)3
= a3 + 3a2 (b + c) + 3a(b + c)2 + (b + c)3
= (a3 + b3 + c3) + 3(a + b + c)(bc + ca + ab) − 3abc
= (a3 + b3 + c3) + 0 − 3.

It follows that the only possible value of a3 + b3 + c3 is 3.

For those with more knowledge of algebra, this problem was practically trivial. Jerry Lo had the
most succinct write-up.
The given conditions show that a, b and c are roots of the equation x3 + kx − 1 = 0 where k is the
common value of the three fractions. Hence

x3 + kx − 1 = (x − a)(x − b)(x − c)
= x3 − (a + b + c)x2 + (bc + ca + ab)x − abc.

It follows that a+b+c = 0, abc = 1 and a3 +b3 +c3 = (a+b+c)(a2 +b2 +c2 −bc−ca−ab)+3abc = 3.
Problem 5.
This problem, which looks deceptively easy, is very annoying. Only one contestant gave a complete
argument, and a handful of others came close. There is a relatively easy part of the problem that
many contestants got, that is, showing that c = 50. Suppose z people in all responded to the survey.
z c
Then 2006 = 100 or 50z = 1003c. Since 50 and 1003 are relatively prime, c must be a multiple of 50.
Since we are given that 0 < a < c < b < 100, the only possible value is c = 50. After this, things
get messy. What follows is the approach by Jeffrey Mo.
Let the total number of teachers be d and the number of those teachers who responded be x.
Then xd = 100a
and 1003−x
2006−d
b
= 100 . From the first, we have ad = 100x. From the second, we have
2006b − bd = 100300 − 100x = 100300 − ad. This may be rewritten as (b − a)d = 2006(b − 50). It
follows that 1003 = 17 × 59 divides (b − a)d. Now b − a < 100 < 1003 and we also have d < 1003
since a < c = 50. Hence there are two cases.
Case 1. 59 divides b − a.
This means of course that b − a = 59 and d = 2 × 17(b − 50). Hence 50x = ad2
= 17a(a + 9). Since
25 is relatively prime to 17 and to at least one of a and a + 9, it must divide either a < 50 or
a + 9 < 59. We consider three subcases.
Subcase 1a. a = 25.
We have b = 25 + 59 = 84, x = 17×25(25+9)50
= 289 and d = 2 × 17(84 − 50) = 1156. Hence
1003 − 289 = 714 and 2006 − 1156 = 850. It follows that 289 of 1156 teachers and 714 of 850
students responded to the survey.
Subcase 1b. a + 9 = 25.
We have a = 25 − 9 = 16, b = 16 + 59 = 75, x = 17×16×2550
= 136 and d = 2 × 17(75 − 50) = 850.
Hence 1003 − 136 = 867 and 2006 − 850 = 1156. It follows that 136 of 850 teachers and 867 of 1156
students responded to the survey.
Subcase 1c. a + 9 = 50.
We have a = 50 − 9 = 41 and b = 41 + 59 = 100. This contradicts b < 100 and there are no
solutions in this subcase.
Case 2. 17 divides b − a.
Then b−a = 17n where n ≤ 5. We have nd = 2×59(b−50). Hence 50nx = and 2
= 59a(17n−50+a).
We consider five subcases, none of which yields additional solutions.
Subcase 2a. n = 1.
We have 50x = 59a(a − 33) and 25 must divide either a or a − 33. The former means a = 25,
but then a − 33 < 0. The latter means a − 33 ≥ 25, but then a ≥ 58 > 50 = c. Both lead to
contradictions.
Subcase 2b. n = 2.
We have 100x = 59a(a − 16) and 25 must divide either a or a − 16. The former means that a = 25,
but then 59a(a − 16) is odd. The latter means a − 16 = 25, but then 59a(a − 16) is again odd.
Subcase 2c. n = 3.
We have 150x = 59a(a + 1) and 25 must divide either a or a + 1. The former means that a = 25,
but then 59a(a + 1) is not divisible by 3. The latter means a + 1 = 25 or 50. If a = 49, 59a(a + 1)
is again not divisible by 3. If a = 24, then x = 236, but then d = 100×236
24
is not an integer.
Subcase 2d. n = 4.
We have 200x = 59a(a + 18) and 25 must divide either a or a + 18. The former means that a = 25,
but then 59a(a + 18) is odd. The latter means a + 18 = 25 or 50. If a = 7, 59a(a + 18) is again
odd. If a = 32, then b = 32 + 68 = 100, and this contradicts b < 100.
Subcase 2e. n = 5.
We have 250x = 59a(35+ a) and 25 must divide either a or 35+ a. However, this means that a ≥ 25
and b = a + 85 > 100, a contradiction.
A shorter approach goes as follows. Let s and t be the respective numbers of students and teachers in
ta sb ta sb
the survey. Then s+t = 2006, both 100 and 100 are integers, and 100 + 100 = 1003. Note that 5 cannot
divide both s and t. If 5 does not divide s, then b must be a multiple of 25. Since 50 < b < 100, we
must have b = 75. If 5 does not divide t, then a must be a multiple of 25. Since 0 < a < 50, we must
have a = 25. If 5 does not divide either s or t, then a = 25 and b = 75 and we have t + 3s = 4012.
Subtract from this s + t = 2006 and we have 2s = 2006, so that s = t = 1003. However, neither
ta sb
100
nor 100 is an integer. Henceforth, we assume that 5 divides exactly one of s and t. We consider
two cases. Suppose 5 divides t. Then b = 75 and we have ta + 75s = 2006 × 50. Subtract this from
75t +75s = 2006× 75, we have (75 − a)t = 2006× 25. Since 5 divides t, 75 − a divides 2 × 17× 59× 5.
Since 0 < a < 50, 25 < 75 − a < 75. Hence we must have 75 − a = 34 or 59. If a = 41, both t
ta
and a are odd, and 100 will not be an integer. If a = 16, we have t = 850. This leads to s = 1156.
Thus 867 students and 136 teachers responded to the survey, yielding a total of 1003, as required by
c = 50. Finally, since (100 − a)t + (100 − b)s = 100(s + t) − (at + bs) = 200600 − 100300 = 100300,
the only other solution is b = 100 − 16 = 84 and a = 100 − 75 = 25, as indicated above since now 5
divides s instead of t. Solving s + t = 2006 and 84s + 25t = 100300, we have s = 850 and t = 1156.
Thus 714 students and 289 teachers responded to the survey, again yielding the desired total of
1003.
 Alberta High School Mathematics Competition

1. A positive integer has 1001 digits all of which are 1s. When this number is divided by 1001,
the remainder is

(a) 1 (b) 10 (c) 11 (d) 100 (e) none of these

2. Some cats have got into the pigeon loft because the total head count is 34 but the total leg
count is 80. The number of cats among the pigeons is

(a) 6 (b) 12 (c) 17 (d) 22 (e) 28

3. In triangle ABC, AB ≤ 1 ≤ BC ≤ 2 ≤ CA ≤ 3. The maximum area of triangle ABC is

3 5
(a) 1 (b) 2
(c) 2 (d) 2
(e) none of these

4. The number of ways in which five As and six Bs can be arranged in a row which reads the
same backwards and forwards is

(a) 1 (b) 5 (c) 10 (d) 15 (e) none of these

5. Among twenty consecutive integers each at least 9, the maximum number of them that can
be prime is

(a) 4 (b) 5 (c) 6 (d) 7 (e) 8

6. The non-negative numbers x and y are such that 2x + y = 5. The sum of the maximum value
of x + y and the minimum value of x + y is

5 15
(a) 0 (b) 2
(c) 5 (d) 2
(e) none of these

7. We wish to choose some of the positive integers from 1 to 1000 inclusive, such that no two
differ by 3 or 5. The maximum number of positive integers we can choose is

(a) 200 (b) 300 (c) 333 (d) 500 (e) none of these

8. The number of polynomials p with integral coefficients such that p(9) = 13 and p(13) = 20 is

(a) 0 (b) 1 (c) 2 (d) 3 (e) infinitely many

9. In the quadrilateral ABCD, AB = CD, AD = 2 and BC = 6. AD and BC are parallel lines

at a distance 8 apart. The radius of the smallest circle which can cover ABCD is
√ √ √
85
(a) 18 (b) 20 (c) 2
(d) 5 (e) none of these
10. The number of pairs (a, b) of positive integers such that all three roots of the cubic equation
x3 − 10x2 + ax − b = 0 are positive integers is

(a) 3 (b) 8 (c) 10 (d) 66 (e) none of these

2 8 2
11. The real numbers x and y are such that x + y
= 3
and y + x
= 3. The value of xy is

3 4
(a) 2
(b) 3
(c) 2 (d) 4

(e) not uniquely determined

12. Let θ be an acute angle such that sec2θ+tan2 θ = 2. The value of csc2θ+cot2 θ is

(a) 2 (b) 3 (c) 4 (d) 5 (e) none of these

13. The diameter AC divides a circle into two semicircular arcs. B is the midpoint of one these
arcs, and D is any point on the other arc. If the area of ABCD is 16 square centimetres, the
distance, in centimetres, from B to AD is
√ √
(a) 2 (b) 2 2 (c) 4 (d) 4 2

(e) dependent on the radius of the circle

14. Five students took part in a contest consisting of six true-or-false questions. Student #i gave
the answer T to question #j if and only if i < j. The total number of incorrect answers is 8
or 9, and there are more incorrect answers of T than incorrect answers of F. The student who
has both an incorrect answer of T and an incorrect answer of F is

(a) #1 (b) #2 (c) #3 (d) #4 (e) #5

15. An integer n is randomly chosen from 1099 to 10100 − 1 inclusive. The real number m is
defined by m = 9n5
. Of the following five numbers, the one closest to the probability that
99 100
10 ≤ m ≤ 10 − 1 is

1 4 1 5 2
(a) 3
(b) 9
(c) 2
(d) 9
(e) 3

16. The smallest value of the real number k such that (x2 + y 2 + z 2 )2 ≤ k(x4 + y 4 + z 4 ) holds for
all real numbers x, y and z is

(a) 1 (b) 2 (c) 3 (d) 6 (e) 9

 Alberta High School Mathematics Competition
Solution to Part I – 2007

1. The number 111111 is divisible by 1001. Now 1001 = 6×166+5. Hence the desired remainder
is the same when we divide 11111 by 1001, which is 100. The answer is (d).
2. Tell the cats to put their front legs up. Now there are still 34 heads, but only 34 × 2 = 68 legs
on the ground. Hence 80 − 68 = 12 legs are up in the air, and each cat puts up 2 of them. It
follows that the number of cats is 12 ÷ 2 = 6. The answer is (a).
3. Since AB ≤ 1 and BC ≤ 2, the area of triangle ABC is at most 2. This maximum value √
can
be attained when AB = 1 and BC = 2 and are perpendicular to each other. Now CA = 5,
and we indeed have 2 ≤ CA ≤ 3. The answer is (a).
4. The middle symbol must be an A. The first five symbols consist of two As and three Bs, and
they can be arranged in all possible ways. The last five symbols, consisting also of two As
and three Bs, must be in reverse order with respect to the first five symbols. In the first five
symbols, we only have to find
 the number of ways of choosing two of the five positions for the
5
two As. This is given by 2 = 10. The answer is (c). An exhaustive analysis also works.
5. The integers from 11 to 30 include six primes, namely 11, 13, 17, 19, 23 and 29. In the ten odd
numbers among any twenty consecutive integers each at least 7, at least three are multiples
of 3 and exactly two are multiples of 5, but at most one can be a multiple of 15. Hence the
maximum is indeed six. The answer is (c).
6. We have 2x+2y = 5+y ≥ 5, so that the minimum value of x+y is 52 , attained at (x, y) = ( 52 , 0).
Also, x + y = 5 − x ≤ 5, so that the maximum value of x + y is 5, attained at (x, y) = (0, 5).
The answer is (d).
7. If we take all the even numbers, clearly no two will differ by 3 or 5. Hence we can take at least
500 numbers. Now partition the integers from 1 to 1000 into blocks of 10. From each of the
following five pairs, we can take at most one number: (10n + 1, 10n + 4), (10n + 2, 10n + 5),
(10n + 3, 10n + 8), (10 + 6, 10n + 9) and (10n + 7, 10n + 10). Hence we can take no more than
500 numbers. The answer is (d).
8. Suppose there exists such a polynomial p. Since an − bn is divisible by a − b for all positive
integers a, b and n with a 6= b, 13 − 9 must divide p(13) − p(9). However, 4 does not divide
7, and we have a contradiction. The answer is (a). A parity argument also works.
9. The centre O of the circle lies on the axis of symmetry of ABCD. Let y be its height above
BC. Then OB 2 = y 2 + 32 while OA2 = (8 − y)2 + 12 . Equating these two values yields y = 72 .
q √
Hence the radius is ( 72 )2 + 32 = 85
2
. The answer is (c).
A .................... .. D
.............. ............. .. ....................
........ . .. .. .
.. . .. .....
..... .. ... ... ... ....
.... .. ... ... ... ....
... .. .. .. ...
..... ... ... .... .. ...
...
.
.... ... .. ...
.
.. ...
..
..
..
..
.
. . .. ..
... ... .. .. ..
.
.. .... .
..
... .. ... ... ..
..
... ... .... . .
... ... .... .. ...
....
... .
..
.
. .
.
.O
.
....
... ... . ..
.. ..
..
.
.
.. ... ... ...
. ..
. ..
.. . .... .
.
.
..
. .
.
.
.. . ... .. ..
... ... ... .... .. ..
... .. ... ... .. ...
... ... ..... ... .. .....
.... ...... .
. . ...
....................................................................................
..... .
....... .....
B ...........
....................................
...... C
10. Let the positive integral roots be r ≤ s ≤ t. Then x3 − 10x2 + ax + b = (x − r)(x − s)(x − t).
Expansion yields x3 − (r + s + t)x2 + (st + tr + rs)x − rst. Hence r + s + t = 10. The possible
partitions are (1,1,8), (1,2,7), (1,3,6), (1,4,5), (2,2,6), (2,3,5), (2,4,4) and (3,3,4). The answer
is (b).
4
11. Multiplying one equation by the other, we have xy + 4 + xy = 8. This may be rewriten as
2 2
0 = (xy) − 4xy + 4 = (xy − 2) . Hence xy = 2. The answer is (c).

12. Let s = sin2 θ and c = cos2 θ. We have sec2θ + tan2 θ = 1+s c

= 2. Since s + c = 1, 2 − c = 2c
so that c = 23 . It follows that s = 13 . Now csc2θ+cot2 θ = 1+c
s
= 5. The answer is (d).

13. Let E be the point on AD such that BE is perpendicular to AD. Complete the rectangle
BEDF . Now AB = BC, 6 AEB = 90◦ = 6 CF B and 6 ABE = 90◦ − 6 CBE = 6 CBF . Hence
ABE and CBF are congruent triangles and they have equal area. It follows that BEDF is
a square, and its area is also 16. Hence BE = 4. The answer is (c).

B .............................................
F

QQ
.......... .......
....... ....
....... ....
. ....
....

Q
....
.
. ... ...
Q
..
. ...

... ...
Q
. ..
... ..

...
QC
..
... ...
.. ..

..
.... ...
... ..

..
...
..
.. ....
...
..

A ..
..
... E ..
..
.. D
... ..
... ...
...
.... ....
.... ....
..... ....
....... .....
. .....
......... .
...............................................

14. The number of incorrect answers for each of questions 1 and 6 is 0 or 5. The number of
incorrect answers for each of questions 2 and 5 is 1 or 4. The number of incorrect answers
for each of questions 3 and 4 is 2 or 3. A total of 8 incorrect answers can only be made up
from 0+1+3+3+1+0. However, we would have an equal number of incorrect answers of T
and incorrect answers of F. Hence the total must be 9, and it can be made up from either
0+1+2+2+4+0 or 0+4+2+2+1+0. However, the latter yields more incorrect answers of F
than incorrect answers of T. It follows that the correct answers for the six questions are F, F,
F, T, F and T respectively. Only student #4 has both an incorrect answer of T (for question
5) and an incorrect answer of F (for question 4). The answer is (d).

15. We have 1099 × 59 ≤ n ≤ (10100 − 1) 59 . Since n is an integer, 5 × 1098 < n < 5 × 1099 . However,
we must eliminate those values of n where 5×1098 < n < 1099 . Thus the number of acceptable
values of n is about 4.5 × 1099 . Since 1099 ≤ n ≤ 10100 − 1, the desired probability is very
close to 12 . The answer is (c).

16. We can rewrite the inequality as (k − 3)(x4 + y 4 + z 4 ) + (y 2 − z 2)2 + (z 2 − x2)2 + (x2 − y 2)2 ≥ 0,
from which it is clear that k ≥ 3. The answer is (c).
 Alberta High School Mathematics Competition
Part II February 6, 2008.

Problem 1.
The function f (n) = an + b, where a and b are integers, is such that for every integer n, f (3n + 1),
f (3n) + 1 and 3f (n) + 1 are three consecutive integers in some order. Determine all such f (n).
Problem 2.
In a contest, no student solved all problems. Each problem was solved by exactly three students
and each pair of problems was solved by exactly one student. What is the maximum number of
problems in this contest?
Problem 3.
From a jar of candies, Autumn takes a% of the candies plus a more candies. Brooke takes b% of
the remaining candies plus b more candies. Here, a and b are positive integers less than 100. If
Autumn and Brooke have taken the same number of candies, determine all possible values of a and
b.
Problem 4.
Let a, b, c and d be real numbers. The equation x2 + ax + b = 0 has two real roots. On the
other hand, the equation (x2 − 2cx + d)2 + a(x2 − 2cx + d) + b = 0 has no real roots. Prove that
d2 + ad + b > c4.
Problem 5.
In triangle ABC, AB = AC and 6 A = 100◦ . D is the point on BC such that AC = DC, and F is
the point on AB such that DF is parallel to AC. Determine 6 DCF .
 Alberta High School Mathematics Competition
Solutions and Comments to Part II, 2008.

Problem 1.
We have f (3n + 1) = 3an + b + a, f (3n) + 1 = 3an + b + 1 and 3f (n) + 1 = 3an + b + (2b + 1).
Hence a, 1 and 2b + 1 must be three consecutive integers in some order. Since one of them is 1,
they can only be (−1, 0, 1), (0,1,2) or (1,2,3). Since 2b + 1 is odd, the second case is impossible. In
the first case, we must have 2b + 1 = −1 and a = 0, yielding f (n) = −1. In the third case, we must
have 2b + 1 = 3 and a = 2, yielding f (n) = 2n + 1.
Problem 2.
Let problem 1 be solved by students A, B and C. Since each pair of problems was solved by exactly
one student, every other problem was solved by exactly one of A, B and C. Suppose the number of
problems is at least eight. By the Pigeonhole Principle, at least three of the other seven problems
were solved by one of A, B and C, say A. Let these be problems 2, 3 and 4. Then apart from
A, no student solved more than one of problems 1, 2, 3 and 4. Since A did not solve all prolems,
there is one, say problem 5, which A did not solve. In order to have a common solver with each of
problems 1, 2, 3 and 4, problem 5 must be solved by at least four students, which is a contradiction.
Hence the number of problems is at most seven. The following scheme, which has all the desired
properties, shows that the number of problems can be exactly seven.

Students Problems Solved

A 1,2,3
B 1,4,5
C 1,6,7
D 2,4,6
E 2,5,7
F 3,4,7
G 3,5,6

Problem 3.
na
Let the total number of canides be n initially. After Autumn has taken 100 + a candies, Brooke
na b
will take (n − 100 − a) 100 + b candies. Equating these two expressions and simplifying, we obtain
na + 100a = nb − nab100
− ab + 100b. This may be rewritten as (n + 100)(b − a − 100 ab
) = 0. Since
100a
n 6= −100, the second factor must be zero so that b = 100−a . Note that a < 50 so that 100 − a > 50.
Let p be any prime divisor of 100 − a. Then p must divide 100a so that it divides either 100 or a.
It follows that p must divide 100, so that p can only be 2 or 5. This means that 100 − a = 64 or
80. However, 36 does not divide 6400. Hence (a, b)=(20,25) is the only possibility. This does work
if we take n = 5k where 3k > 40. Autumn will take k + 20 candies, leaving behind 4k − 20. Then
Brooke will take (k − 5) + 25 = k + 20 candies.
Problem 4.
Let r and s be the two real roots of x2 + ax + b = 0. Then r + s = −a and rs = b. That
(x2 − 2cx + d)2 + a(x2 − 2cx + d) + b = 0 has no real roots means that neither x2 − 2cx + d = r
nor x2 − 2cx + d = s has any real roots. It follows that (−2c)2 < 4(d − r) or c2 < d − r. Similarly,
c2 < d − s. Multiplication yields c4 < d2 − d(r + s) + rs = d2 + ad + b.
Problem 5.
Since AB = AC and 6 CAB = 100◦ , 6 ABC = 6 BCA = 40◦ . Now 6 CAD = 6 CDA = 70◦ since
AC = DC, so that 6 BAD = 30◦ . Since ED is parallel to AC, 6 F DA = 70◦ and 6 BDF = 40◦ . Let
P be the point on CD such that 6 P AD = 6 F AD = 30◦ . Then 6 P AC = 40◦ so that P A = P C.
Now triangles P AD and F AD are congruent, so that AP = AF . Since 6 P AF = 60◦ , P AF is an
equilatral triangle. Hence 6 AP F = 60◦ so that 6 F P D = 20◦ . Moreover, P F = P A = P C. Hence
6 P CF = 6 P F C = 10◦ .
A ......
............ .. . . .
.... .. ... .....
..... .. .. .......
.... ... .... ....
.
...... .. ... ....
....
.....
. .. .. ....
...... ...
.
..
. ....
... . .. ....
.. . .
... ....
...... ..
. . ....
...... .
.. ...
...
....
....
..... .. ....
.... .. ... ....
. . . . ....
... . ..
F ...
...
....... .
... ................................................. .
.
.
.
.
.. .
...
...
..
.....
....
....
....
..... .... ........... ..................... . ....
. . .... .
............ . ..... .
................... .
...... ...... ...
. .
.
.......... ..
......
...
....................
....
....
...... .. . .
..... . ..
. ..
..... ....
..... ....
.. .. ...............
. ... ...................
...
. ..
.. .
.. ....
.... .... . ......... .. ................... ....
. .
. ... . . ....
...... . . .
.................. ...
...........................................................................................................................................................................................................................................................................................

B D P C
 Alberta High School Mathematics Competition
Part I November 18, 2008.

1. The numbers m and n are related by m = 9n 5

+ 32. If n is an integer between 1 and 99, the
number of corresponding values of m which are also integers is

(a) 1 (b) 19 (c) 20 (d) 50 (e) 99

2. When the positive integers a, b and c are divided by 13, the respective remainders are 9, 7
and 10. When a + 2b + 3c is divided by 13, the remainder is

(a) 1 (b) 3 (c) 4 (d) 5 (e) 7

3. A polygon has area greater than 0. All side lengths are integers, the largest of which is 10.
The smallest possible perimeter of this polygon is

(a) 10 (b) 11 (c) 20 (d) 21 (e) none of these

4. Let f and g be functions such that f (x) = g(2x) and g(x) = 2f (x) for all real numbers x. If
g(2) = 3, the value of f ( 12 ) is

1 1
(a) 3
(b) 2
(c) 2 (d) 3 (e) 6

5. If a certain positive integer is divided by 7, the remainder is 6. If it is divided by 4, the

remainder is 1. The remainder when it is divided by 28 is

(a) 5 (b) 6 (c) 7 (d) 13 (e) 17

6. A building is in the shape of a right circular cylinder, with mirrors for its interior wall. A
horizontal ray starts from a point on the wall, reflects off the wall exactly 11 times and returns
to its starting point for the first time. The number of possible directions along which it may
have started is

(a) 4 (b) 8 (c) 10 (d) 11 (e) infinite

7. For a real number t ≥ 0, |t| = t, and for a real number t < 0, |t| = −t. The number of real
numbers x satisfying |||||x − 1| − 2| − 3| − 4| − 5| = 0 is

(a) 0 (b) 1 (c) 2 (d) 5 (e) 32

8. A rectangular box has integral dimensions and volume 2008 cubic centimetres. In square
centimetres, the minimum value of its surface area is

(a) 3028 (b) 4024 (c) 4534 (d) 5028 (e) none of these
 9. The positive integers a and b are such that there are exactly 10 integers greater than a and
less than b, and exactly 1000 integers greater than a2 and less than b2. The value of b is
(a) 50 (b) 51 (c) 52 (d) 55 (e) 102
10. Linda is walking at constant speed halfway between two parallel train tracks. On each track
is a train of the same length. They are approaching Linda from different directions, both at
V kilometres per hour. The one going the same way as Linda takes t1 seconds to pass her,
while the other takes t2 seconds to pass her. In kilometres per hour, Linda’s walking speed is
t1 −t2 t2 −t1 t1 +t2 t2 +t1
(a) t1 +t2
V (b) t2 +t1
V (c) t1 −t2
V (d) t2 −t1
V

(e) dependent on the length of the trains

11. In base 10, the positive integer a has 2 digits, the positive integer b has a digits and the
positive integer c has b digits. The smallest possible value for c is
9 9 −1 9 10 −1 10
(a) 109 (b) 1010 (c) 1010 (d) 1010 (e) 1010 .
12. Of the eight members of a Sports Club, two are senior boys, two are senior girls, two are
junior girls and two are junior boys. An official team is of size four, and must consist of two
seniors and two juniors, as well as two girls and two boys. The number of different official
teams is
(a) 2 (b) 8 (c) 10 (d) 16 (e) 18
13. On a vertical wall are drawn four geometric figures with equal area. The first is an equilateral
triangle with a side along the floor. The second is an isosceles right triangle with one of the
short sides along the floor. The third is a square with a side along the floor. The fourth is a
circle tangent to the floor. The figure which reaches the highest point above the floor is
(a) only the equilateral triangle (b) only the isosceles right triangle
(c) either triangle (d) only the square (e) only the circle

14. The sum of the fifth powers of the roots of x3 − 3x + 1 = 0 is

(a) −15 (b) −9 (c) 0 (d) 6 (e) none of these
15. A country has 100 cities numbered from 1 to 100. For all m < n, there is a road linking the
n
m-th and the n-th city if and only if m is a prime number. One can travel in either direction
on a road. The minimum number of roads one needs to use to go from the 99th city to the
100th city is
(a) 4 (b) 5 (c) 6 (d) 7 (e) 8
16. In each lottery ticket, you choose two of the numbers 1, 2, 3, 4 and 5. Eventually, two of these
five numbers will be drawn. Your ticket wins if neither of the numbers drawn are chosen on
your ticket. The smallest number of tickets you must have in order to guarantee that at least
one of them wins is
(a) 3 (b) 4 (c) 5 (d) 6 (e) 7
 Alberta High School Mathematics Competition
Solution to Part I – 2008

1. In order for m to be an integer, n must be a multiple of 5. Since there are 19 integer values
of n within range, the number of corresponding values of m that are integers is also 19. The
answer is (b).

2. When a+2b+3c is divided by 13, the remainder will be the same as when 9+2×7+3×10 = 53
is divided by 13. The latter is clearly 1. The answer is (a).

3. The total length of all but one side in a polygon with positive area must be greater than the
length of that side. To minimize the perimeter, choose that side to be the longest. Hence the
perimeter is greater than 20. Since all side lengths are integers, the minimum is at least 21.
This minimum value may be achieved with a triangle of side lengths 5, 6 and 10. The answer
is (d).

4. Note that f (x) = x3 and g(x) = x6 satisfy the hypothesis. In any case, we have f ( 12 ) = g(1) =
2f (1) = 2g(2) = 6. The answer is (e).

5. From the first condition, the desired remainder is one of 6, 13, 20 and 27. From the second
condition, the desired remainder is one of 1, 5, 9, 13, 17, 21 and 25. The only common value
is 13. The answer is (d).

6. Consider a horizontal cross-section of the room as a clock, and let the ray start at the 12
o’clock position. It must travel directly at another point on the clock in order to return to
its starting position after 11 bounces. However, some of these 11 paths take it back to the
starting point before 11 bounces. The initial destination must be a point which is relatively
prime to 12, namely, 1, 5, 7 or 11. The answer is (a).

7. We have ||||x − 1| − 2| − 3| −4| = 5. Now |||x − 1| − 2| − 3| = 4 ± 5. Since the left side can never
be equal to −1, we must have |||x − 1| − 2| − 3| = 9. Similarly, ||x − 1| − 2 = 12, |x − 1| = 14
and x = 1 ± 14 = −13 or 15. The answer is (c).

8. Since 2008 = 2 × 2 × 2 × 251, the box must be one of 2 × 4 × 251, 2 × 2 × 502, 1 × 8 × 251,
1 × 4 × 502, 1 × 2 × 1004 and 1 × 1 × 2008, with respective surface area 3028, 4024, 4534,
5028, 6028 and 8034. The answer is (a).

9. We have b − a = 11 and b2 − a2 = 1001, so that b + a = 91. Hence 2b = 102 and b = 51. The
answer is (b).

10. Let Linda’s walking speed be v kilometres per hour. Then the length of the trains is given by
t1(V − v) = t2 (V + v). It follows that v = tt11 −t
+t2
2
V . The answer is (a).

11. The smallest value of a leads to the smallest value of b, and of c. The smallest value of a is
10, a 1 followed by 2 − 1 0s. The smallest value of b is 109 , a 1 followed by 10 − 1 0s. The
9
smallest value of c is 1010 −1 , a 1 followed by 109 − 1 0s. The answer is (b).
12. If we take both senior boys, we must complement them with both junior girls. If we take
neither senior boy, we must take both senior girls and both junior boys. If we take only one
senior boy, then we must take one senior girl, one junior girl and one junior boy. In each
category, there are two choices. Hence the number of different teams is 1 + 1 + 24 = 18. The
answer is (e).

13. The height of the circle is 2r where r is its radius. From πr2 = 1, we have 2r = √2π . Since
√
2 < π < 4, this is greater than √24 = 1, the height of the square, and less than √22 = 2, the
√
height of the right isosceles
√ triangle. The height
√ of √
the equilateral triangle
√ is√ 3s √
where 2s
2 4 4 4
is its side length. From 3s = 1, we have 3s = 3. Since 3 < 4, 3 < 4 = 2. The
answer is (b).

14. Let a, b and c be the roots. Then a + b + c = 0 and bc + ca + ab = −3. It follows that
a2 + b2 + c2 = (a + b + c)2 − 2(bc + ca + ab) = 6. Since x3 = 3x − 1, we have

x5 = x2(3x − 1) = 3(3x − 1) − x2 = −3 + 9x − x2 .

Hence a5 + b5 + c5 = −9 + 9(a + b + c) − (a2 + b2 + c2 ) = −15. The answer is (a).

15. In moving from city to city, we either pick up an extra prime or discard one. The numbers of
prime divisors of 99 = 3 × 3 × 11 and 100 = 2 × 2 × 5 × 5 are 3 and 4 respectively, and there
are no common divisors. To change from one lot to the other, we need to make 7 moves. The
answer is (d).

16. We can guarantee that at least one ticket wins by buying the tickets (1,2), (1,3), (2,3) and
(4,5). If neither 4 nor 5 is drawn, the last ticket wins. If at least one of them is drawn, then at
most one of 1, 2 and 3 is drawn. Then one of the first three tickets wins. If we only buy three
tickets, we have to choose 6 numbers. By the Pigeonhole Principle, one of the numbers 1, 2,
3, 4 and 5 will appear on at least two of our tickets. If it appears on all three, none of them
wins if this number is drawn. Suppose it appears only on the first two tickets. If this number
is drawn along with one of the numbers on our third ticket, we will not have a winning ticket.
It follows that the minimum number of tickets to guarantee that at least one of them wins is
4. The answer is (b).
 Alberta High School Mathematics Competition
Part II, 2009.

Problem 1.
Let w, x, y and z be non-negative numbers whose sum is 100. Determine the maximum possible
value of wx + xy + yz.
Problem 2.
1
Determine all positive integers a and b, a < b, so that exactly 100
of the consecutive integers
a2, a2 + 1, a2 + 2, . . . , b2 are the squares of integers.
Problem 3.
A game is played on a 7 × 7 board, initially blank. Betty Brown and Greta Green make alternate
moves, with Betty going first. In each of her moves, Betty chooses any four blank squares which
form a 2 × 2 block, and paints these squares brown. In each of her moves, Greta chooses any blank
square and paints it green. They take alternate turns until no more moves can be made by Betty.
Then Greta paints the remaining blank squares green. Which player, if either, can guarantee to be
able to paint 25 or more squares in her colour, regardless of how her opponent plays?

Problem 4.
A, B and C are points on a circle Ω with radius 1. Three circles are drawn outside triangle ABC
and tangent to Ω internally. These three circles are also tangent to BC, CA and AB at their
respective midpoints D, E and F . If the radii of two of these three circles are 23 and 11 2
, what is
the radius of the third circle?
A...............................................................................................
........................ ... .. .....
.............
. ..........
......... .... .. ....
.... ... ....
...........E ......
.......... .. ..
............... ..............
...
...
.... ...... ....
.......... ...
.... F
.. ... .
... ..
B....................................................................................................................................................................................C
.
.
.
.
. ............
....
.. .....
.. ........ ..
.....
....
...
..
...
...
.... D ................ ..
...
...
... ... ... ..
... ... .. ...
... .. .. ..
...
...
.
..
.
. .
...
.. . .. .
... ..... ... ..
.. ... . ...
.. .. .. ..
.. ... ... ...
...
...
...
... ....
..
.. ..
.
...
...
.
... Ω
.. ....
... ..
.... ... .
... ......
.... ... . .
..... .... ... .....
............ ..... ......
................ .... ......
................. ................
....................................

Problem 5.
Prove that there are infinitely many positive integers k such that k k can be expressed as the sum
of the cubes of two positive integers.
 Alberta High School Mathematics Competition
Solutions and Comments to Part II, 2009.

Problem 1.
Since (w + y) + (x + z) = 100, we have w + y = 50 + t and x + z = 50 − t for some real number t.
Hence wx + xy + yz ≤ (w + y)(x + z) = (50 + t)(50 − t) = 2500 − t2 ≤ 2500. This maximum value
may be attained for instance when w = x = 50 and y = z = 0.
Problem 2.
The number of integers between a2 and b2 inclusive is b2 − a2 + 1. The number of squares between
a2 and b2 inclusive is b − a + 1. From b2 − a2 + 1 = 100(b − a + 1), we have (b + a − 100)(b − a) = 99.
Since 99 has 6 positive divisors, there are 6 solutions, as shown in the chart below.

b + a − 100 b − a 2b − 100 b a
1 99 100 100 1
99 1 100 100 99
3 33 36 68 35
33 3 36 68 65
9 11 20 60 49
11 9 20 60 51

Here is a slightly different approach. Let d = b − a. Then there are (a + d)2 − a2 + 1 = 2ad + d2 + 1
integers under consideration, d + 1 of which are the squares of integers. It follows that we need
100(d + 1) = 2ad + d2 + 1, so that

100(d + 1) − d2 − 1 100 − d 99
a= = + .
2d 2 2d
If d is even, the first term is an integer and the second is not. Hence d must be odd. Then the
first term is a fraction with denominator 2, so that the second term must also be a fraction with
denominator 2. This means that d must be a divisor of 99, that is, d is 1, 3, 9, 11, 33 or 99.
If d = 1, then a = 992
+ 99
2
= 99 and b = 99 + 1 = 100.
If d = 3, then a = 2 + 99
97
6
= 65 and b = 65 + 3 = 68.
91 99
If d = 9, then a = 2 + 18 = 51 and b = 51 + 9 = 60.
If d = 11, then a = 89 2
+ 99
22
= 49 and b = 49 + 11 = 60.
If d = 33, then a = 67 2
+ 99
66
= 35 and b = 35 + 33 = 68.
1 99
If d = 99, then a = 2 + 198 = 1 and b = 1 + 99 = 100.
Problem 3.
There are 9 squares at the intersections of even-numbered rows and even-numbered columns. Any
2 × 2 block chosen by Betty must include one of these 9 squares. Hence Greta should play only on
these squares in her first four moves. This will ensure that Betty has at most five moves, and can
paint at most 20 squares brown. Hence Greta wins.
 . ... .... . . ... .... . . ... ... ..
. ... .... . . ... .... . . ... ... ..
. ... .... . . ... .... . . ... ... ..
. ... .... . . ... .... . . ... ... ..
. . .. . . . .. . . . .. .

. ... .... . . ... .... . . ... ... ..

. ... .... . . ... .... . . ... ... ..
. ... .... . . ... .... . . ... ... ..
. ... .... . . ... .... . . ... ... ..
. . .. . . . .. . . . .. .

. ... .... . . ... .... . . ... ... ..

. ... .... . . ... .... . . ... ... ..
. ... .... . . ... .... . . ... ... ..
. ... .... . . ... .... . . ... ... ..
. . .. . . . .. . . . .. .

Problem 4.
Denote the circumcentre of Ω by O and note that it lies within the circle with radius 23 . We have
 
2 1 2 2 2 32
BC = 4BD = 4(OB − OD ) = 4 1 − =
9 9
and  
2 49 288 2 2 2
CA = 4CE = 4(OC − OE ) = 4 1 − = .
121 121
2 +OC 2 −BC 2 2 +OA2 −CA2
By the Cosine
√
Law, cos BOC = OB
√ 2OB·OC
= − 79 and cos COA = OC 2OC·OA 23
= − 121 . Hence
4 2 84 2
sin BOC = 9 and sin COA = 121 . It follows that
833
cos AOB = cos(6 BOC − 6 COA)) = (cos BOC)(cos COA) + sin BOC sin COA = .
1089
512
By the Cosine Law again, AB 2 = OA2 + OB 2 − 2OA · OB cos AOB = 1089 . Hence we have
OF = OA −AF = 1089 and OF = 33 . It follows that the radius of the third circle is 12 (1− 31
2 2 2 961 31
33
1
) = 33 .

.....................................................
.................. ..........
............ ..
... .......... ............. .....................
........ ....
A .......
............
........ ............. ..
.......
...
. .
.. ..................
.. ......
. ........ ... ............. . ... .. .....
....
. .. .......... . . ..
..... ... ... .......... ..... ..
.. .... ....
.. . .... ..... ... .. ...
. ............ . .
. .
....
....
...... ..... .. ..
... ..........
..
... . . ..
. ....
... ..... ... .. . . . . . ...
.. ... ...
......... ... . . ...
. ........... .... . . . .. . ..
............ .. .
.
. .... ...
. ...... .......... ... .... . ....... ...
...
.......................... .
.... .................. ..
.... ... . .......... ...
.. .... .... . . ...
.
.
.. .....
.. ....
.
F .....
. ....
......
..
...
. .
E ..
.
.
.
. ..
...........
. .
.............
.. .
.. .........
...
...
..
.. ..... .......
...
... .. . . . . . .......... ..
.. ....
. ..... ... ... ..
.............. ..
..
B ......... ....
.....
.
...............
. .
.
D ..
...............................................................................................................................................................................................................................................................................................................
.
.... ................... ...............
...
.
..
.....
.
... ... C
.......... .... ......... .. .. . .
. . . . .......... .. ...... .
. ..
... .......... ... .... . .
. . . ........... ........... ..
.. .......... .......... ...
.... .
.. ..... .......... ..
... .......... ..... ........ ..
... .... ..... ......... ..
... .......... ....... .... ...
. . . ..... ................... ...
... .. .. . . . . .
.... .................... ....
..
... .. ..
. . .. .
.... ............. ..
... ..... ..........
..
.... .. .. .
. . . . ...
. .... .....
. .... ..
.. .......... .... .. .. . .......... ... ...
... .. .. ... ...
.......... ... .. . . . . .. ...
. . ....
... ...
. .................... ................... .... ...
... ... .................... . ....
... . .. ..
.
..
. ..
....
...
.. .
...
..
. ...
.
.
.
..
O ..
..
... .
.
.
..
.
..

... . . ... .
.. ..... ...
... .. ..
...
... ... .. ..
..
.. .. ... ..
..
... .... .
. ... ..
.. ... .... .... ..
.. ... ..
.. ... .. ..
.. ..... ... .... ..
.. . . . ...
.. ... . . .
.. ... .... ..
.. ..
..
... .. ... ... ...
... ... .. ..
...
... ... .
. ... ....
... ..
.. .... ..
.. ...
... .. ... .. ...
... .. ... .. ...
....
... ... .
. ... .
.....
... ....
....
.... ... .... ...
.
....
.... ... ... ... ....
.... .... .... ... .....
..... ... ....
..... ....... .. ... .. .
..... . ... .... ........
...... ........ ....
....
.... ......
.
....... ..... ..... .......
........ ..... .. ......................
......... ....... .
. ...
....
....................
...................... ..
. ..... .
..
..
..
....................
.
.........................................................
Problem 5.
We have (a + 1)a+1 = (a + 1)a (a + 1) = a(a + 1)a + (a + 1)a . Choose a = 33t for an arbitrary
3t−1
positive integer t. Then a = (3t )3 and (a + 1)a = ((33t + 1)3 )3 are both cubes. If we take
3t−1 3t−1
k = 33t + 1, m = 3t (33t + 1)3 and n = (33t + 1)3 , then k k = m3 + n3 . Since t is an arbitrary
positive integer, the number of possible choices for k is infinite.
Essentially the same solution is given by Jarno Sun of Western Canada High School, who took
k = 27n3 + 1 for any positive integer n. Then
3 +1 3 +1 3 3
(27n3 + 1)27n = (27n3 + 1)27n − (27n3 + 1)27n + (27n3 + 1)27n
3 3
= (27n3 + 1)27n (27n3 + 1 − 1) + (27n3 + 1)27n
3 3
= ((27n3 + 1)9n 3n)3 + ((27n3 + 1)9n )3 .

Yuri Delanghe of Harry Ainlay High School argued as follows. Let t be any integer congruent to
t2t −1
4 modulo 6. Then t ≡ 1 ≡ 2t (mod 3). Hence t2t − 1 is divisible by 3. Let m = n = 2 3 and
t t
k = 2t . Then m3 + n3 = 2t2 = (2t )2 = k k .
When t = 4, k = 16, and this was also the initial example of Danny Shi of Sir Winston Churchill
High School. He showed inductively that for any solution k, 64k is also a solution, in that

(64k)64k = (464k k 21k )3 k k = (464k k 21k a)3 + (464k k 21k b)3 ,

where k k = a3 + b3 .
 The Alberta High School Mathematics Competition
Part I, November 17, 2009

1. If 2x = 3y , then 4x is equal to

(a) 5y (b) 6y (c) 8y (d) 9y (e) none of these

2. Caroline bought some bones for her 7 dogs. Had she owned 8 dogs, she could have given each
the same number of bones. As it was, she needed two more bones to give each dog the same
number of bones. The number of bones she could have bought was

(a) 16 (b) 24 (c) 32 ( d) 40 (e) 48

3. Ace calculates the average of all the integers from 1 to 100. Bea calculates the average of
all the integers from 1001 to 1100 and subtracts 1000. Cec calculates the average of all the
integers from 1000001 to 1000100 and subtracts 1000000. The largest answer is given by

(a) Ace only (b) Bea only (c) Cec only (d) exactly two of them
(e) all three of them

4. A large rectangular gymnasium floor is covered with unit square tiles, most of them blank,
in the pattern shown in the diagram below. Of the following fractions, the one nearest to the
fraction of tiles which are not blank is
1 1 1 1 1
(a) 12
(b) 8
(c) 6
(d) 5
(e) 4

@@ @@
@@ @@
@@ @@
@@ @@
@@ @@
@@ @@
@@ @@
@@ @@
@@ @@
@@ @@
@@ @@
@@ @@

5. The number of integers between 1 and 2009 inclusive which can be expressed as the difference
of the squares of two integers is

(a) 1 (b) 502 (c) 1005 (d) 1507 (e) 2009

6. Among the positive integers with six digits in their base-10 representation, the number of
those whose digits are strictly increasing from left to right is

(a) between 1 and 50 (b) between 51 and 100 (c) between 101 and 500

(d) between 501 and 1000 (e) greater than 1000

 7. The number of arrangements of the letters AABBCC in a row such that no two identical
letters are adjacent is

(a) 30 (b) 36 (c) 42 (d) 48 (e) none of these

8. If 22009 has m digits and 52009 has n digits in their base-10 representations, then the value of
m + n is

(a) 2007 (b) 2008 (c) 2009 (d) 2010 (e) 2011
√
9. An equilateral triangle has area 2 3. From the midpoint of each side, perpendiculars are
dropped to the other two sides. The area of the hexagon formed by these six lines is
√ √
(a) 23 (b) 1 (c) 3 (d) 2 (e) none of these

10. Two sides of an obtuse triangle of positive area are of length 5 and 11. The number of possible
integer lengths of the third side is

(a) 3 (b) 4 (c) 6 (d) 8 (e) 9

11. Q(x) is a polynomial with integer coefficients such that Q(9) = 2009. If p is a prime number
such that Q(p) = 392, then p can

(a) only be 2 (b) only be 3 (c) only be 5

(d) only be 7 (e) be any of 2, 3, 5 and 7

12. A parallelogram has two opposite sides 5 centimetres apart and the other two opposite sides
8 centimetres apart. Then the area, in square centimetres, of the parallelogram

(a) must be at most 40 and can be any positive value at most 40

(b) must be at least 40 and can be any value at least 40
(c) must be 40 (d) can be any positive value (e) none of these
r q √
13. The number of positive integers n such that n+ n + ··· + n < 10 for any finite number
of square root signs is

(a) 10 (b) 90 (c) 91 (d) 99 (e) 100

14. A chord of a circle divides the circle into two parts such that the squares inscribed in the two
parts have areas 16 and 144 square centimetres. In centimetres, the radius of the circle is
√ √ √
(a) 2 10 (b) 6 2 (c) 9 (d) 85 (e) 10

15. The number of prime numbers p such that 2p + p2 is also a prime number is

(a) 0 (b) 1 (c) 2 (d) 3 (e) more than 3

√
16. Suppose that 2 − 99 is a root of x2 + ax + b where b is a negative real number and a is an
integer. The largest possible value of a is

(a) −4 (b) 4 (c) 7 (d) 8 (e) none of these

 Alberta High School Mathematics Competition
Solution to Part I – 2009

1. If 2x = 3y , then 4x = (2x )2 = (3y )2 = 9y . The answer is (d).

2. The number of bones Caroline bought is a multiple of 8 but 2 less than a multiple of 7. The
answer is (d).

3. The calculation is (n+1)+(n+2)+···+(n+100)

100
− n = 100n+(1+2+···+100)
100
− n = 1+2+···+100
100
, with n = 0
for Ace, n = 1000 for Bea and n = 1000000 for Cec. The answer is (e).
4. Almost the entire gymnasium floor may be divided into 2 × 3 non overlapping rectangles each
with exactly one non-blank square at the lower left corner. The answer is (c).

5. Observe that x2 − y 2 = (x − y)(x + y) is the product of two integers of same parity. Hence
x2 − y 2 is either odd or divisible by 4. Thus a number which is neither odd nor divisible by
4 cannot be expressed as a difference of two squares. On the other hand, if n is odd, then
n = 2k + 1 = (k + 1)2 − k 2 . If n is divisible by 4, then n = 4k = (k + 1)2 − (k − 1)2 . Between
1 and 2009 inclusive, there are 1005 numbers that are odd and 502 that are divisible by 4.
The answer is (d).
 
9
6. We can choose any six of the nine non-zero digits. The number of choices is 6
= 84. Each
choice gives rise to a unique number. The answer is (b).
7. Assume that the first A appears before the first B, and the first B before the first C. Then we
must start with AB and continue with A or C. If we continue with A, the last three letters
must be CBC. If we start with ABC, we must continue with A or B. In either case, either of
the last two letters can appear before the other. So the total is 1 + 2 × 2 = 5. Relaxing the
order of appearance, the total becomes 5 × 3! = 30. The answer is (a).
8. Since 10n−1 < 22009 < 10n and 10m−1 < 52009 < 10m , we have 10m+n−2 < 2200952009 < 10m+n .
It follows that 10m+n−1 = 2200952009 = 102009. Hence m + n − 1 = 2009. The answer is (d).
9. Let ABC be the triangle and DREP F Q be the hexagon, as shown in the diagram below.
Triangles AP E, AP F, ERD and F QD are all congruent to one another.√ Hence DREP F Q
has the same area as the parallelogram AF DE, which is one half of 2 3. The answer is (c).
A
..........
... ... ...
.. ... ...
... ... ...
.... .... ....
. ... ...
... ... ...
..
.. ..... ...
.
... ..
. ..
...
..
.. ...... .........
.
.. ...........
... ...... ....

F .
.....
.
........... P ........ ...
...... ..
E
........
.....
........ .........
.. .. ... .. .. ..
.... .... .... ... .... ....
. .
... .... .... ... ... ....
. ... ... ... ..... ..
... .... .. .. ...
.. ... .. .. ..
... ..
. .... .. .
. ...
... ...
... ... .
.
. ...
... ........ ... ... . .. . ...
... ...... . .
.. ........
. ..
.. ...... ... . ......... ...
.
.
.
.
Q ...
...... .. .. ......
..... . .
.
.
....
. R ...
..
.....................................................................................................................................................................

B D C
10. By the Triangle Inequality, the third side must be from 7 to 15. Now 92 < 112 − 52 < 102 , so
that (5,7,11), (5,8,11) and (5,9,11) are obtuse triangles. Also, 122 < 112 + 52 < 132 , so that
(5,11,13), (5,11,14) and (5,11,15) are obtuse triangles. The other three are not. The answer
is (c).
11. Since Q(9) = 2009, x − 9 is a factor of Q(x) − 2009. Since Q(x) has integral coefficients, p − 9
divides Q(p) − 2009 = 392 − 2009. Since this number is odd, p − 9 must be odd. Since p is
prime, we must have p = 2. Now 392 − 2009 = −1617 = −231 × 7, so that Q(x) does exist.
For instance, we may have Q(x) = 231x − 70. The answer is (a).

12. Let 5 centimetres be the height of the parallelogram. Its base is a segment intercepted by the
other pair of parallel lines 8 centimetres apart. Hence its length is at least 8 centimetres and
can be arbitrarily large. The answer is (b).
√
13. q
With one square root sign, n < 10 is equivalent to n < 100. With two square root signs,
√ √
n + n < 10 is equivalent to n + n < 100, which is in turn equivalent to n < 91. With
r q √ √
three square root signs and n < 91, we have n + n + n < n + 10 ≤ 10. With more
square root signs, the same inequality will hold. There are 90 positive integers which satisfy
n < 91. The answer is (b).

14. Let x be the distance from the centre of the circle to the bottom edge of the larger square.
2
The square of the radius of the circle is√given by 6√ + x2 = 22 + (4 + 12 − x)2 . This yields
x = 7 so that the radius of the circle is 62 + 72 = 85. The answer is (d).
......................................................................................................
................. ... .............
......... ....
.. ..
......... ....
...... ...
... ........
......
.. ...
..... ... .. ... .....
.
....... .
. ... .
.
.
.....
.....
....
..... .. ..
. ..
. . ....
.... .. ... .. ..
. ....
.. . .
. .. .
. .... ....
.. .
. .
. ...
.... . ..
.
. .
.
.
.
.
. ...
... .. .... .. ... ..
...............................................................................................................................................................................................................
.
. . .
.. ... .. .
. .
. ..
. . . . ...
..
.
.
.
.
.. ..
.. . . .
.
. ..
.
.. ..... .. ... .... ..
..
..
. .
.. . . .
. .
. ..
.. . .. ... .
.
. ..
... ..
... ...
.. .. . .
.
. ...
.. .... .. .. ..
. ..
...
.. .. ... ... ...
. ..
... .... .. ... ... ..
.... .. ....
......
.
. ..
... .... ..
. ...
.... ... .
. ..
. ...
... . ... .
.
... .
... ...
. . . ..
... ... . . ..
. ..
.
... ... . ... ... .
.
. .
.
.. .. .. . . .
.
... . ..
. .
. .
. ..
.
... .... .... .... ..
.. ... ...
.. . ... ... ... ...
... ... .... .... ..
.. .
. ...... . . .
.
..
.. ... ... ..
.. . ... .... .... ..
..
.. ..... ... .... .... ..
.. ... ... .. .. ..
... . ..... .
. .
. ....
... . . ...
... .... .... .... .... ...
... ... ... ... ...
.... .... ...... ... ... ......
... ... .... .
. .
. ..
.... .. ..
........................................................................................................................................................................
.... .
.... ....
..... .....
...... ......
........
......... ..
............
............ ........
...............................................................

15. If p = 2, 2p + p2 = 8 is not prime. If p = 3, 2p + p2 = 17 is prime. If p > 3, then p = 6k ± 1

for some integer k, so that p2 = 36k 2 ± 12k + 1 is 1 more than a multiple of 3. On the other
hand, when divided by 3, successive powers of 2 leave remainders of 2 and 1 alternately. Since
p > 3 is odd, 2p is 2 more than a multiple of 3. Hence 2p + p2 is divisible by 3, and cannot be
prime. The answer is (b).
√ √
16. Let √the other root be t. Then −a = t + 2 − 99. Now b = t(2 − 99) < 0, so that
t = 99 − 2 − a > 0. Since a is an integer, a ≤ 9 − 2 = 7. The answer is (c).
 The Alberta High School Mathematics Competition
Part II, February 3, 2010

1. Of Melissa’s ducks, x% have 11 ducklings each, y% have 5 ducklings each and the rest have
3 ducklings each. The average number of ducklings per duck is 10. Determine all possible
integer values of x and y.

2. (a) Find all real numbers t 6= 0 such that tx2 − (2t − 1)x + (5t − 1) ≥ 0 for all real numbers
x.
(b) Find all real numbers t 6= 0 such that tx2 − (2t − 1)x + (5t − 1) ≥ 0 for all x ≥ 0.

3. Points A, B, C and D lie on a circle in that order, so that AB = BC and AD = BC + CD.

Determine 6 BAD.

4. Let n be a positive integer. A 2n × 2n board, missing a 1 × 1 square anywhere, is to be

partitioned into rectangles whose side lengths are integral powers of 2. Determine in terms of
n the smallest number of rectangles among all such partitions, wherever the missing square
may be.

5. Let f be a non-constant polynomial with non-negative integer coefficients.

(a) Prove that if M and m are positive integers such that M is divisible by f (m), then
f (M + m) is also divisible by f (m).
(b) Prove that there exists a positive integer n such that each of f (n) and f (n + 1) is a
composite number.
 The Alberta High School Mathematics Competition
Solution to Part II, 2010.

1. We have 11x + 5y + 3(100 − x − y) = 1000 or 4x + y = 350. Since y ≥ 0, we get x ≤ 87.

Since x + y ≤ 100, we also have that 3x ≥ 250, so x ≥ 84. Thus the only solutions are
(x, y) = (84, 14), (85,10), (86,6) and (87,2).
2. For either (a) or (b), clearly the leading coefficient t of the quadratic must be positive.
(a) For the inequality to hold for all real x, the discriminant must be non-positive, that is,
0 ≥ (2t − 1)2 − 4t(5t − 1) = 1 − 16t2 = (1 − 4t)(1 + 4t).
Since t > 0, 1 + 4t > 0, so we need 1 − 4t ≤ 0. Thus t ≥ 14 .
(b) We now have the additional possibility that the two roots of the quadratic are real and
non-positive. This holds if and only if 0 < t ≤ 14 , 2t − 1 ≤ 0 and 5t − 1 ≥ 0. This is
equivalent to 15 ≤ t ≤ 14 . Combining with the answer to (a), we have t ≥ 15 .
3. First Solution:
Putting AB = BC = b and CD = c, we get AD = b + c. Let 6 BAD = α. Since ABCD
is cyclic, 6 BCD = 180◦ − α. Applying the cosine law to triangles BAD and BCD, we have
BD2 = b2 + (b+ c)2 − 2b(b+ c) cos α and BD2 = b2 + c2 − 2bc cos(180◦ − α) = b2 + c2 +2bc cos α.
Hence b2 + (b + c)2 − 2b(b + c) cos α = b2 + c2 + 2bc cos α, so that b2 + 2bc = (2b2 + 4bc) cos α.
This yields cos α = 12 , so that α = 60◦ is the only possibility.
..... ....
..................... ...................................
........ . . .......
...... ... .............. ....
..... ...
..
.
.....
...
... ........
B .. ... .....
.. .. ....
. .
.. ... .....
....
....
...
...
. .. ... .... .....
.... .....
.... ...
.
. . .... .. ..
. . . ....
...
...
.
..... .. ... . ..
. .. .. .... .
... ...... .. ... .... ...
... ...... ..
..
... ... ..
.... ..
.. .... .. ..... ......
...... . ... ...
A ...................
...
...
.............
.............
............
..
..
..
..
...
...
.
.
....
.......
C
.. ............. . .... .... . .
.. ............. .... . .. . .
.. .............
.. ............. .... ......
... .................
...
... E ...
.
...
....
..... .
......
...
...
D
..... .
....... ......
........... .......
......................................
Second Solution:
Let E be the point on AD such that DE = DC, so that AE = AD − DE = BC = AB. Now
6 BDE = 6 BDC since they are subtended by the equal arcs BA and BC. It follows that

triangles BED and BCD are congruent, so that BE = BC = BA = AE, triangle BAE is
equilateral and 6 BAD = 60◦ .
4. First Solution:
The area of the punctured board is 22n − 1. The base-2 representation of this number consists
of 2n 1s. Since the area of each rectangle in the partition is a power of 2, we must have at
least 2n rectangles. There exist such partitions with exactly 2n rectangles. Divide the board
in halves by a horizontal grid line. Set aside the one with the missing square and cover the
other with a rectangle of height 2n−1 . Repeating the process with the strips set aside, we
obtain rectangles with decreasing heights 2n−2 , 2n−3 , . . . , 21 and 20 , a total of n rectangles.
We now divide the resulting 2n × 1 board in halves by a vertical line. Set aside the one with
the missing square and cover the other with a rectangle of width 2n−1 . Repeating the process
with the strips set aside, we obtain another n rectangles with decreasing widths, for a total
of 2n rectangles in the overall partition.
 Second Solution:
Divide the board into four congruent quadrants. Set aside the one with the missing square.
Merge two of the other quadrants into one rectangle and keep the third quadrant as the second
rectangle. In reducing a 2n × 2n board down to a 2n−1 × 2n−1 board, we use two rectangles. It
follows that we will use exactly 2n rectangles in the overall partition. We now prove that we
cannot get by with a smaller number. The area of a rectangle of the prescribed type is a power
of 2. The smallest has area 1, and the largest has area 22n−1 . Thus there are 2n different sizes.
If we use one of each size, the total area of these 2n rectangles is 1 + 2 + · · · + 22n−1 = 22n − 1,
exactly the size of the punctured chessboard. Consider any other collection of rectangles
whose areas are powers of 2 and whose total area is 22n−1 − 1. Replace any pair of rectangles
of equal area by one with twice the area. Repeat until no further replacement is possible. The
resulting collection consists of rectangles of distinct areas which are powers of 2, and with
total area 22n−1 − 1. It can only be our collection, and since mergers only reduce the number
of rectangles, 2n is indeed minimum.

5. (a) Note that f (M + m) − f (m) is a sum of terms of the form ak ((M + m)k − mk ) where ak is
the coefficient of the term xk in f (x). Since each term is divisible by M = (M + m) − m,
so is f (M + m) − f (m). Since M is divisible by f (m), f (M + m) − f (m) is divisible by
f (m). It follows that f (M + m) is divisible by f (m).
(b) Since all the coefficients of f are non-negative and f is non-constant, it is strictly increas-
ing. Let M = f (2)f (3) and n = M + 2. By (a), f (n) is divisible by f (2) and f (n + 1)
is divisible by f (3). Since f (n + 1) > f (n) > f (3) > f (2) > f (1) ≥ 1, both f (n) and
f (n + 1) are composite.
 The Alberta High School Mathematics Competition
Part I, November 16, 2010

1. The number of positive integers n such that the number 4n has exactly two digits is

(a) 21 (b) 22 (c) 23 (d) 24 (e) 25

2. A 4 × 6 plot of land is divided into 1 × 1 lots by fences parallel to the edges of the plot, with
fences along the edges as well. The total length of fences is:

(a) 58 (b) 62 (c) 68 (d) 72 (e) 96

3. The greatest common divisor and least common multiple of two positive integers are 1 and
10 respectively. If neither of them is equal to 10, their sum is equal to

(a) 3 (b) 6 (c) 7 (d) 11 (e) none of these

4. The number of pairs (x, y) of non-negative integers such that 3x + 2y = 27 is

(a) 4 (b) 5 (c) 8 (d) 9 (e) 10

5. In the sequence 1, 2, 3, 4, 6, 7, 8, 9, . . . , obtained by deleting the multiples of 5 from the

sequence of the positive integers, the 2010th term is

(a) 2511 (b) 2512 (c) 2513 (d) 2514 (e) none of these

6. Alice, Brian, Colin, Debra and Ethel are in a hotel. Their rooms are on floors 1, 2, 3, 21
and 40 respectively. In order to minimize the total number of floors they have to cover to get
together, the floor on which the get-together should be is

(a) 18 (b) 19 (c) 20 (d) 21 (e) none of these

7. A square pigeon coop is divided by interior walls into 9 square pigeonholes in a 3 × 3 config-
uration. Each of two pigeons chooses a pigeonhole at random, possibly the same one. The
probability that they choose two holes on the opposite sides of an interior wall is

1 1 4 8 1
(a) 18
(b) 9
(c) 27
(d) 27
(e) 3

1
8. The set of all values of the real number x such that ≤ −3 ≤ x is
x

(a) {x ≤ −1/3} (b) {−3 ≤ x ≤ −1/3} (c) {−3 ≤ x < 0}

(d) {−1/3 ≤ x < 0} (e) none of these

 9. In the quadrilateral ABCD, AB is parallel to DC, DC = 2AB, 6 ADC = 30◦ and 6 BCD = 50◦.
Let M be the midpoint of CD. The measure of 6 AMB is

(a) 80◦ (b) 90◦ (c) 100◦ (d) 110◦ (e) 120◦

10. We are constructing isosceles but non-equilateral triangles with positive areas and integral side
lengths between 1 and 9 inclusive. The number of such triangles which are non-congruent is

(a) 16 (b) 36 (c) 52 (d) 61 (e) none of these

11. In each of the following numbers, the exponents are to be evaluated from top down. For
c c
instance, ab = a(b ) . The largest one of these five numbers is
3 2 2 2 2
22 23 32 22 22
( a) 22 (b) 22 (c) 22 (d) 23 (e) 32

12. A gold number is a positive integer which can be expressed in the form ab + a + b, where a
and b are positive integers. The number of gold numbers between 1 and 20 inclusive is

(a) 8 (b) 9 (c) 10 (d) 11 (e) 12

13. The edges DA, DB and DC of a tetrahedron ABCD are perpendicular to one another. If
the length of DA is 1 cm and the length of each of DB and DC is 2 cm, the radius, in cm,
of the sphere passing through A, B, C and D is

3
√
5+1
√ √ 1
(a) 2
(b) 2
(c) 3 (d) 2+ 2
(e) none of these

14. Let f (x) = x2 and g(x) = x4. We apply f and g alternatively to form

f (x) = x2, g(f (x)) = g(x2) = (x2)4 = x8, f (g(f (x)) = f (x8 ) = (x8)2 = x16,

and so on. After we have applied f 50 times and g 49 times, the answer is xn where n is

(a) 148 (b) 296 (c) 2148 (d) 2296 (e) none of these

15. Triangle ABC has area 1. X, Y are points on the side AB and Z a point on the side AC such
that XY = 2AX, XZ is parallel to Y C and Y Z is parallel to BC. The area of XY Z is

1 2 1 2 1
(a) 27
(b) 27
(c) 9
(d) 9
(e) 3

16. The number of integers n for which n3 − 3n + 2 is divisible by 2n + 1 is

(a) 3 (b) 4 (c) 5 (d) 6 (e) 8

 The Alberta High School Mathematics Competition
Solution to Part I, 2010

1. Since 10 ≤ 4n ≤ 99, 3 ≤ n ≤ 24. Hence there are 24 − 3 + 1 = 22 such values. The answer is
(b).

2. There are 7 fences of length 4 and 5 fences of length 6. The total length is 7 × 4 + 5 × 6 = 58.
The answer is (a).

3. Since the least common multiple is even, at least one number is even. Since the greatest
common divisor is odd, exactly one number is even. We can show in a similar manner that
exactly one of the two numbers is divisible by 5. Since neither is 10, one of them is 2 and the
other is 5, yielding a sum of 7. The answer is (c).

4. Note that x must be odd, and x = 9 − 2y 3

. Since y ≥ 0, x ≤ 9. Thus there are 5 triples
(x, y) = (9, 0), (7,3), (5,6), (3,9) and (1,12). The answer is (b).

5. Note that 2010 × 54 = 2512.5. There are 502 multiples of 5 from 5 to 2510 inclusive. Hence
2512 is the (2512 − 502)-th or 2010-th number in the punctured sequence. The answer is (b).

6. Alice and Ethel are 39 floors apart, and as long as the get-together floor is in between, the
total number of floors they cover is 39. Similarly, the total number of floors Brian and Debra
cover is 19. The minimum number of floors Colin covers is 0, when they get together on floor
3. The answer is (e).

7. The pigeons can choose the pigeonholes in 9 × 9 = 81 ways. There are 12 pairs of rooms
separated by an interior wall. Since the pigeons can choose these two rooms in 2 ways, the
desired probability is 2×12
81
8
= 27 . The answer is (d).

8. Since x1 ≤ −3, x ≤ 0. Multiplying by − x3 , we have − 13 ≤ x. Hence the set of all values of x is

{− 13 ≤ x < 0}. The answer is (d).

9. Since AB = DM and AB is parallel to DM, ABMD is a parallelogram. Similarly, ABCM is

a parallelogram. Therefore, 6 AMD = 6 BCD = 50◦ and 6 BMC = 6 ADC = 30◦ . Therefore,
6 AMB = 180◦ − 6 AMD − 6 BMC = 180◦ − 50◦ − 30◦ = 100◦ . The answer is (c).

A B
..................................................................................................................................................
....... ...... .... ...
...... ... ...... ....
......... .... ...... ...
....
........ ...
. ............ ....
..
.. ... ...
. ...
.. ... ...
...
....
. . ... ...... ....
.... .. ....
......... ...
.. ......... ...
...
......... ...
.... ...... ........ ....
....... . .. ..
................................................................................................................................................................................................................................................................
. .

D M C

10. We first count the triangles in which the equal sides of length k are longer than the third side,
which can be of length from 1 to k−1. Summing from k = 1 to 9, we have 0+1+2+· · ·+8 = 36
such triangles. We now count the triangles in which the equal sides of length k are shorter
than the third side, which can be of length from k + 1 to 2k − 1. Summing from k = 1 to 9,
we have 0 + 1 + 2 + 3 + 4 + 3 + 2 + 1 + 0 = 16. The total is 36 + 16 = 52. The answer is (c).
 256 512 81
11. The first three choices are equal respectively to 22 , 22 and 22 . Clearly, the second one is
16
the largest among them. The fourth number is equal to 23 . Since 2512 = 4256 > 3256 > 316,
16
the second number is larger than the fourth one. The fifth number is equal to 32 . Clearly,
256 255 255 16
22 = 42 > 32 > 32 . Hence the second number is the largest among the five choices.
The answer is (b).

12. Note that ab + a + b + 1 = (a + 1)(b + 1). Every composite number can be written in this
form and no prime number can be written in this form. Therefore, the positive integers that
are not gold numbers are those that are one less than a prime. By simple counting, we see
there are 8 primes from 2 to 21. Therefore, the number of gold numbers between 1 and 20
inclusive is 20 − 8 = 12. The answer is (e).

13. The sphere which passes through A, B, C and D also passes through the other four vertices
of a 1 × 2 × 2 block having
√ A, B, C and D as four of its vertices. Since the space diagonal
2 2 2 3
of this block is of length 2 + 2 + 1 = 3, the radius of the sphere is 2 . The answer is (a).

14. Each application of f doubles the exponent while each application of g quadruples the ex-
ponent. After 50 applications of f and 49 applications of g, the exponent has been doubled
50 + 2 × 49 = 148 times so that n = 2148 . The answer is (c).

15. Denote the area of triangle T by [T ]. Since triangles AXZ and AY C are similar, ZC = 2AZ.
Since triangles AY Z and ABC are similar, Y B = 2AY . It follows that
2 2 2 2
[XY Z] = [AY Z] = [AY C] = [ABC] = .
3 9 27 27
The answer is (b).
A......
.... ..
.... ....
....
X ...
..
.... ....
..
..
...
.... ...... ..
...... ...
.... ....
.
.... .
... . ... ...
.. ..... ...
.
..... .... ...
.
Y .
.
.... ...
....................................................................................
.... .......
Z
..
... .... ..
. ... ..
... .... ..
...... ......
..
..
...... ......
..
..
.... .... ..
.... ..... ..
..
..... ..... ...
.
. ... . ..
... .... ..
.. . .... ..
..... . ... ..
..... .. ...
... .... ..
.. ..... ..
..... ..... ..
..... .
...
..
..
..... .. ...
.... .... ..
. ..... .
..... ..... ....
....
...
. ..... ...
...... . .... ...
.. .. ..
......................................................................................................................................................................................................................................

B C
2
16. When n3 −3n+2 is divided by 2n+1, the quotient is n2 − n4 − 11 8
and the remainder is 27
8
. Hence
3 2 3
8(n − 3n + 2) = (2n + 1)(4n − 2n − 11) + 27, so that 2n + 1 divides n − 3n + 2 if and only if it
divides 27. The set of all values of n for which 2n+1 divides 27 is {−14, −5, −2, −1, 0, 1, 4, 13},
and there are 8 such values. The answer is (e).
 The Alberta High School Mathematics Competition
Part II, February 2, 2011.

Problem 1.
A cross-shaped figure is made up of five unit squares. Determine which has the larger area, the
circle touching all eight outside corners of this figure, as shown in the diagram below on the left, or
the square touching the same eight corners, as shown in the diagram below on the right.
.......................................
............. .........
......... .......
......... ... ........
..................................................................................................................................................................
............ ...... .... ...............
... ... ..... .... ...... ....
...
........ ... ... .... .... ...
...
....... .
. ... ... ... ....
.... ... ..... .... ... ....
....
.... ...
..
.. .
. . .... .....
... .... ... .... .... .... .... ...
... .... .... .... .... ... ... .
. ... .... ..... .... ...
.......... .... ..... ......
...... .... .......
. ....
.... .... .
.. .. ...... ....
........ .... .... ...... ........ .... .... ......
....
. ........ . ...... . . .
. .....
.. ..... ............ ...... ... .... ......
.... ...... .... ..
.
.
. ..... .... . ... ...... ... .... ...... .... ... .... ....
. .... . .... .. .. ... ...
... ... ..... ..... .... .. .... .... .. .... .. ...
.. .... .... ... .... .. .... ... .... ... ...
.. ... ....
... .... ... .... ...
.....
....
...
....
... ... .. ... .... .... .... .... ....
....
.... ... ... . .... ... . .
. ... ... .... ....... .... ....... ..
.... ........ ....... ..
. ... ........ ........ ...
. ...... .......
...
.
..
.. .....
..
.
..
.. .....
.. ... .... . .
... ......
. .
... .....
...
... . . ... .. .... ... ... .. .. ...
.. ..... .
. .... ..... ..... .. ... . .... .... ..... .... ...
...
..
. .. . . .
... . .
.. ... .... ....
.... ....
...
... . .... . ... ... .... .. ....
.... ... .... .... .. ... . . ... . .... ....
..
.. ...... .... ... ... .. .... ......
.. ... .. . .
.. .... .... ....... .... .... ....
.... .... .. .... ...
...... ...... ... .. ........ ....... .......
...... .
. .
.. ...... . ..
... .... .... ... .... .... .... ......
....... ... ...... ..... .... ...... ... ...... .... ....
....... ... .... ...... ... .... ...
... .... ... .... .... .. ... .... ... ... ....
... ... .... ........ .... .... ..... ..
.
.... ....
... ... ...... .
. ... . .... .... .... ... .... ....
.... ...
... .... ... .... ... .... .... .... .... ..
.........
........ ..... .... ............. ... ... ...... .... ...... ..
.... ........... .. .
..........
........ .. .. ..........................................................................................................................................................................
........... ......
.................................................

Problem 2.
There is exactly one triple (x, y, z) of real numbers such that x2 + y 2 = 2z and x + y + z = t.
Determine the value of t.
Problem 3.
On the side BC of triangle ABC are points P and Q such that P is closer to B than Q and
6 P AQ = 1 6 BAC. X and Y are points on lines AB and AC, respectively, such that 6 XP A = 6 AP Q
2
and 6 Y QA = 6 AQP . Prove that P Q = P X + QY .
Problem 4.
Determine all the functions f from the set of integers to the set of positive integers such that
f (n − 1) + f (n + 1) ≤ 2f (n) for all integers n.
Problem 5.
Seven teams gather and each pair of teams play one of three sports, such that no set of three teams
all play the same sport among themselves. A triplet of teams is said to be diverse if all three sports
are played among themselves. What is the maximum possible number of diverse triplets among the
seven teams?
 The Alberta High School Mathematics Competition
Solution to Part II, 2011.

Problem 1. √
The diameter
√ of the circle, being the diagonal of a 1 × 3 rectangle, is 10, so the area of the circle
2
is π( 10/2)2 = 5π 2
. The diagonal of the square is 4, so the side of the square is 42 = 8. Since
π < 3.2 = 16
5
, we have 5π 2
< 8. Thus the square has greater area than the circle.
Problem 2.
Eliminating z, we have x2 + y 2 = 2(t − x − y) so that (x + 1)2 + (y + 1)2 = 2(t + 1). In order to
have a unique solution for x and y, we must have 2(t + 1) = 0 or t = −1.
Problem 3.
Let M be the point on P Q such that 6 MAP = 6 BAP . Then

6 MAQ = P AQ − 6 MAP 6
1
= (6 ABC − 6 MAB)
2
1
= 6 MAC
2
= 6 CAQ.

Since 6 XP A = 6 MP A, triangles XAP and MAP are congruent by the ASA Postulate, so that
P X = P M. Similarly, we can prove that QY = QM, so that P X + QY = P M + QM = P Q.
A..
........... ....... ...
........... .. ..
.......... .. .
... ................ .... ...
... .. .. ... ...
..... ... ... .. ..
.... ..... ... ... ..
..... ... ...
... ... ..... ..... ..
..
..
..
...... ..... .... .. ..
....... .... .... ...
. ..
...
.
..... ..... .... ..
. ...
....... ..... .... .. ..
.
. .... .. .. ...
. ..
. ..
..... ..... ... ... ..
..
X .
...
.... ...
.........
.......
... . ...

....
.
.......
.
.
.
.
.
.
.
..
..
..
..
... . . ... .. ..
.
... . .. .. . ..
... ... ..... .. .
... .. ..
.......................................................................................................................................................................
........
..
C
..
B P M Q .........
........
......... ...
........ ..
Y
.....

Problem 4.
Since f (n) takes on only positive integral values, it has a minimum value m. Let n be such that
f (n) = m. Then 2m ≤ f (n−1)+f (n+1) ≤ 2f (n) = 2m, which implies that f (n−1) = f (n+1) = m
also. It follows easily that f (n) = m for all integers n.
Problem 5.
We first show that the conditions of the problem can be satisfied. Construct a graph where the
teams are represented by vertices Ti , 0 ≤ i ≤ 6. In the diagram below, we partition the graph into
three subgraphs. Two teams play each other in the first sport if and only if the vertices representing
them are joined by an edge in the first subgraph, the second sport in the second subgraph and the
third sport in the third subgraph. None of the subgraphs contains a triangle.

T ... 0
....... ..............
T .. 0
... ...
T......0
........ ........ ... ...... .....
....... ....... ... .. ..
....... ...
T 6 ......... ...
.... ........
..... T
... 1
..
T .... T
6 ............................................................................................... 1 T6 ........
.......... .
.
.. ..
.. ....
.. ....
T
.......
...........
1
... ... .. .. ...
... ....
... ........
... ........ .... . ......... ...
.. ... .. .... ... . ... ......... ................. .....
.... ... .... ... .. ... . ... .
.. . .
.
..
...
... ...... .... ... ........................... ....
. .. ............ ..............
... ... ..... ...... .. . ..... ...........
... .. .... .. ... ........ .......
...
...
...
.
.
... ... .. ...... ............. .
...... .....
.... ........
.......
T ...
5 ..... . .
T2
...
. T .
..........
5 ........... ... .. .
.
.
.. T2...
.......
...............
T 5
...
.
.... ...... ... T2
.......................................................................................................................
... . .. ......... .. ............ ....
...
.... . ............
.....
........... .. ... ... ...
... .. ......... ..
... ... .. ....... .............. ... ... .... ... ..
.... ... .................. .. ... ....
... ..
... ... ..
... .............. ........ ...
... ..
................................................... ......... ......
..
T4 T3 T4 .... ....
T3 T4 T3
The edges in the same subgraph have the same length, and those in different subgraphs have
different lengths. In geometric terms, a diverse triple is a scalene triangle. There is basically one
such triangle, namely T0 T1T3. Six others can be obtained from it by rotation, and seven more by
reflection. Thus we may have as many as 14 scalene triangles.
We now prove that there are at most 14 diverse triples. Construct a complete graph on 7 vertices
which represent the 7 teams. Paint an edge in the i-th colour if the teams represented by its
endpoints play each other in the i-th sport, 1 ≤ i ≤ 3. A triangle is diverse if all three sides are
of different colours, and non-diverse otherwise. Since there are no monochromatic triangles, a non-
diverse triangle has two sides of the same colour. Call the vertex at the junction of the two sides
of the same colour its pivot. The number of pivots is equal to the number of non-diverse triangles.
There are six edges incident with each vertex. If at least 3 of them are of the same colour, then
this vertex is the pivot of at least 3 non-diverse triangles. If not, then exactly 2 edges are of each
colour, so that the vertex is the pivot of exactly 3 isosceles triangles. Hence each vertex is the pivot
of at least 3 non-diverse triangles. Since there are 7 vertices, this brings the total to at least 21, so
that the maximum number of diverse triangles or diverse triples is 14.
 The Alberta High School Mathematics Competition
Part I, November 15, 2011

1. If 22012 + 41006 = 2n , then n is

(a) 2013 (b) 2014 (c) 3018 (d) 4024 (e) not an integer

2. Mini-marshmallows are cubes 1 centimetre on each side, while giant marshmallows are cubes
3 centimetres on each side. The number of mini-marshmallows whose combined surface area
is the same as the surface area of one giant marshmallow is

(a) 3 (b) 6 (c) 9 (d) 27 (e) 54

3. The number of customers in a restaurant on Tuesday is 20% more than the number on Monday,
the number of customers on Wednesday is 50% more than the number on Monday, and the
number of customers on Wednesday is n% more than the number on Tuesday. The value of
n is

(a) 20 (b) 25 (c) 30 (d) 50 ( e) none of these

4. Sawa starts from point S and walks 1 kilometre north, 2 kilometres east, 3 kilometres south
and 4 kilometres west. At this point, her distance, in kilometres, from S is
√ √
(a) 5 (b) 2 2 (c) 4 (d) 8 (e) 10

5. A millenium number is a positive integer such that the product of its digits is 1000. The
number of six-digit millenium numbers is

(a) 60 (b) 120 (c) 140 (d) 180 (e) 200

6. Let x1, x2 , . . . be a sequence of positive rational numbers such that x1 = 16, x2 = 32 and
x +x
xn = n−l 2 n−2 for all positive integers n ≥ 3. Then the value of x6 is

(a) 24 (b) 25 (c) 26 (d) 27 (e) 28

√
7. The number of real solutions of the equation 2x2 − 2x = 2x x2 − 2x + 1 is

(a) 0 (b) 1 (c) 2 (d) 3 (e) 4

8. Let f (x) be a quadratic polynomial. If f (1) = 2, f (2) = 4 and f (3) = 8, then the value of
f (4) is

(a) 12 (b) 14 (c) 15 (d) 16 (e) 18

 9. A lucky number is a positive integer n such that 7 is the largest divisor of n which is less than
n. The number of lucky numbers is

(a) 1 (b) 2 (c) 3 (d) 4 (e) more than 4

10. The perimeter of a square lawn consists of four straight paths. Annabel and Bethany started
at the same corner at the same time, running clockwise at constant speeds of 12 and 10
kilometres per hour respectively. Annabel finished one lap around the lawn in 1 minute.
During this minute, the number of seconds that Annabel and Bethany were on the same path
is

(a) 36 (b) 42 (c) 48 (d) 50 ( e) none of these

11. ABCD is a quadrilateral with AD = BC and AB parallel to DC. It is only given that the
lengths of AB and DC are 20 and 15 centimetres respectively. Adrian puts n copies of this
tile together so that the edge BC of each copy coincides with the edge AD of the next, and
the edges DC of all copies together form a regular n-sided polygon. Then the value of n is

(a) 6 (b) 8 (c) 12 (d) 20 (e) not uniquely determined

12. The sum of twenty positive integers, not necessarily different, is 462. The largest possible
value of greatest common divisor of these numbers is

(a) 21 (b) 22 (c) 23 (d) 33 (e) 42

13. The largest real number m such that (x2 + y 2)3 > m(x3 + y 3)2 for any positive real numbers
x and y is
1 1
(a) 3
(b) 2
(c) 1 (d) 2 ( e) none of these

14. Of the 49 squares of a 7 × 7 square sheet of paper, two are to be coloured black while the
others remain white. Two colourings are called distinct if one cannot be obtained from the
other by rotating the sheet of paper about its centre. The number of distinct colourings is

(a) 288 (b) 294 (c) 296 (d) 300 (e) 588

15. The lengths of the sides of triartgle ABC are consecutive positive integers. D is the midpoint
of BC and AD is perpendicular to the bisector of 6 C. The product of the lengths of the three
sides is

(a) 24 (b) 60 (c) 120 (d) 210 (e) 336

16. For any real number r, brc is the largest integer less than or equal to r. For example, bπc = 3.
Let n be a positive integer. Let a1 = n, a2 = b a31 c, a3 = b a32 c and a4 = b a33 c. The number of
positive integers n from 1 to 1000 inclusive such that none of a1, a2, a3 and a4 is divisible by
3 is

(a) 144 (b) 192 (c) 210 (d) 280 ( e) none of these
 The Alberta High School Mathematics Competition
Solution to Part I, 2011

1. We have 22012 + 41006 = 22012 + 22012 = 22013. The answer is (a).

2. The surface area of a mini-marshmallow is 6 square centimetres while that of a giant marsh-
mallow is 54 square centimetres. Thus the desired number of mini-marshmallows is 54÷6 = 9.
The answer is (c).

3. Suppose there are m customers on Monday. Then there are 1.2m on Tuesday and 1.5m on
Wednesday. The increase of 0.3m from Tuesday to Wednesday is 25% of 1.2m. The answer
is (b).

4. Sawa is 2 kilometres
√ south√and 2 √
kilometres west of S. Her distance from S, by the Pythagorean
2 2
Theorem, is 2 + 2 = 8 = 2 2 kilometres from S. The answer is (b).

5. Since 1000 = 2 × 2 × 2 × 5 × 5 × 5, the digits can onlybe

 1, 2, 4, 5 and 8. Three of them must
be 5s and they can be placed among the six digits in 63 = 20 ways. The product of the other
three digits is 8, and they are (1,1,8), (1,2,4) or (2,2,2). They can be placed in 3, 6 and 1
ways respectively. Hence the total number of six-digit millenium numbers is 20(3+6+1)=200.
The answer is (e).

6. We have x3 = 24, x4 = 28, x5 = 26 and x6 = 27. The answer is (d).

√
7. Squaring both sides of 2x2 − 2x − 1 = 2x x2 − 2x, we have 4x4 − 8x3 + 4x + 1 = 4x4 − 8x3,
which simplifies to 4x+1 = 0. Hence the only solution is x = − 14 . Indeed, 2(− 14 )2 −2(− 14 ) = 58
q
and 2(− 14 ) (− 14 )2 − 2(− 14 ) + 1 = 58 . The answer is (b).

8. Note that g(x) = f (x + 1) − f (x) is a linear polynomial. Since g(1) = f (2) − f (1) = 2 and
g(2) = f (3) − f (2) = 4, we have g(3) = 6. Hence f (4) = f (3) + g(3) = 8 + 6 = 14. The
answer is (b).

9. Since a lucky number n is divisible by 7, it has the form n = 7k for some positive integer k.
If k is not a prime number, then it has a divisor h where 1 < h < k, and 7h is a divisor of n
larger than 7 but not equal to n. Hence k must be a prime number. Moreover, it cannot be
greater than 7. Hence there are only 4 lucky numbers, namely, 14, 21, 35 and 49. The answer
is (d).

10. Annabel spent 15 seconds on each path, and Bethany 18 seconds. On the first path, Bethany
was with Annabel all 15 seconds. On the second path, Bethany joined Annabel 3 seconds late,
and was with her for 12 seconds. On the third path, Bethany was with Annabel for 9 seconds.
On the fourth path, Bethany was with Annabel for 6 seconds. The total is 15+12+9+6=42
seconds. The answer is (b).

11. We can draw a regular polygon of any number of sides such that the side length is 20 centime-
tres. We can then draw a regular polygon of the same number of sides but with side length
15 centimetres, placed centrally inside the larger polygon. Then a tile can be chosen which
can pave the ring-shaped region inside the larger polygon but outside the smaller one. Hence
the answer is (e).
12. Let x1 , . . . , x20 be the given numbers. If d is the greatest common divisor of these numbers
then  
x1 x20
x1 + · · · + x20 = d + ··· + = 462 = 21 · 22.
d d
The value d = 22 is obtained if x1 = · · · = x19 = d and x20 = 2d. For each i, xdi ≥ 1. Hence
d ≤ 462
20
= 23.1. Since d divides 462, the largest value for d is indeed 22. The answer is (b).

13. Dividing throughout by y, we have (z 2 + 1)3 > m(z 3 + 1)2 where z = xy . This is equivalent to

(1 − m)z 6 + 3z 4 + (3 − 2m)z 3 + 1 − m > 0

for any positive real z. Hence it is necessary to have 1 − m ≥ 0, i.e. m ≤ 1. If we take m = 1,

the inequality (x2 + y 2)3 ≥ (x3 + y 3)2 is equivalent to

x2 y 2((x − y)2 + 2x2 + 2y 2) > 0,

which is clearly true. The answer is (c).

 
14. There are 49
2
= 1176 colourings. The number of symmetrical colourings with respect to the
middle square is 49−1
2
= 24. These colourings are counted twice. All the other colourings are
counted four times. The desired number is 24
2
+ 1176−24
4
= 300. The answer is (d).

15. Let AD intersect the bisector of 6 C at G. Then 6 CGA = 90◦ = 6 CGD, 6 GCA = 6 GCD and
GD = GD. Hence triangles GCA and GCD are congruent, so that AC = DC. It follows that
we have BC = 2DC = 2AC. Now among three consecutive positive integers, one is double
another. This is only possible if the integers are 1, 2 and 3, or 2, 3 and 4. The former does
not yield a triangle. Hence AC = 2, AB = 3 and BC = 4, so that AB · BC · CA = 24. The
answer is (a).
A ....
.......... ....
....... .. ......
............ ... ....
.... . ...
...... ... ....
...... .. ....
....
........................ .. ...
....
.... ......... .

..
..................... .
.
.
.
..........
.
.G..
. ............
....
....
....
...... .. ......... ...
.... . ....
......... ....
......... ... ....
......... ....
..... .
..
.... ..
...
.... ........... ......
.
... . ......... ....
........ .. ....... ..
.................................................................................................................................................................................................................
. .

B D C

16. Note that if we write a positive integer m in base 3, then the base 3 representation of b m 3
c
is simply the base 3 representation of m with the rightmost digit removed. Also, a positive
integer m is divisible by 3 if and only if the rightmost digit of m is 0. Hence, in order that none
of a1, a2, a3 and a4 is divisible by 3, the rightmost four digits of the base 3 representation of
n are all non-zero. Note that 1000 > 2(35 + 34 + 33 + 32 + 3 + 1). If n has at most 6 digits
in its base 3 representation, the first two can be any of 0, 1 and 2, while the last four cannot
be 0. There are 32 × 24 = 144 such numbers. Clearly, n cannot have more than 7 digits as
otherwise n ≥ 37 > 1000. Suppose n has exactly 7 digits. As before, the last four cannot be
0. Since 1000 < 36 + 35 + 33 + 32 + 3 + 1, the first one must be 1, the second must be 0 and
the third can be any of 0, 1 and 2. Hence, there are 3 × 24 = 48 such numbers. The total is
192. The answer is (b).
 The Alberta High School Mathematics Competition
Part II, February 1, 2012.

Problem 1.
A rectangular lawn is uniformly covered by grass of constant height. Andy’s mower cuts a strip of
grass 1 metre wide. He mows the lawn using the following pattern. First he mows the grass in the
rectangular “ring” A1 of width 1 metre running around the edge of the lawn, then he mows the
1-metre-wide ring A2 inside the first ring, then the 1-metre-wide ring A3 inside A2, and so on until
the entire lawn is mowed. Andy starts with an empty grass bag. After he mows the first three rings,
the grass bag on his mower is exactly full, so he empties it. After he mows the next four rings, the
grass bag is exactly full again. Find, in metres, all possible values of the perimeter of the lawn.



A4
 A3

A2
A1
Problem 2.  2
PA PD
In the quadrilateral ABCD, AB is parallel to DC. Prove that = , where P is a point
PB PC
on the side AB such that 6 DAB = 6 DP C = 6 CBA. .
Problem 3.
A positive integer is said to be special if it can be written as the sum of the square of an integer
and a prime number. For example, 101 is special because 101=64+37. Here 64 is the square of 8
and 37 is a prime number.

(a) Show that there are infinitely many positive integers which are special.

(b) Show that there are infinitely many positive integers which are not special.

Problem 4. √
In triangle ABC, AB = 2, BC = 4 and CA = 2 2. P is a point on the bisector of 6 B such
that AP is perpendicular to this bisector, and Q is a point on the bisector of 6 C such that AQ is
perpendicular to this bisector. Determine the length of P Q.
Problem 5
Determine the smallest positive integer n for which there exist real numbers x1 , . . . , xn , 1 ≤ xi ≤ 4
for i = 1, 2, . . . , n, which satisfy the following inequalities simultaneously:
7n
x 1 + x2 + · · · + xn ≥
3
1 1 1 2n
and + + ··· + ≥ .
x1 x2 xn 3
 The Alberta High School Mathematics Competition
Solution to Part II, 2012.

Problem 1.
Let the dimensions of the lawn be a metres by b metres. The area of the first three rings is given
by ab − (a − 6)(b − 6) = 6(a + b) − 36. Similarly, the area of the next four rings is given by
(a − 6)(b − 6) − (a − 14)(b − 14) = 8(a + b) − 160. These two regions contain the same amount of
grass and so must be the same area. Thus 6(a + b) − 36 = 8(a + b) − 160. It follows that the only
possible value of the perimeter of the lawn is 2(a + b) = 124 metres.
Problem 2.
Since 6 DAB = 6 CBA and AB is parallel to DC, we have AD = BC. Since AB is parallel to DC,
6 BP C = 6 P CD. It follows that triangles BP C and P CD are similar. A similar argument shows

that triangles ADP and P CD are also similar. Hence ( PP D

C
)2 = BC
BP
AP
· AD = PP B
A
, as desired.
A......................................................................................P B
....................................................
.. .. . ...
.. ...... ......
.. ....... .... ...
.. ....... ... ..
.... .......... .
..
..
..
.. ....
...... ....
.... ..
.. .
..... ..
.. .... ..
... ....
.
........ ..... ....
... ......... .... ...
.
... ...
... .... ...
. ...... .... ..
.. ........ ..
..............................................................................................................................................................
D C
Problem 3.
(a) For any positive integer n, n2 + 2 is special.
(b) We claim that for infinitely many positive integers n, n2 is not special. Suppose n2 = m2 + p
for some integer m and some prime number p. Then p = n2 − m2 = (n − m)(n + m). We
must have n − m = 1 and p = n + m = 2n − 1. If we let n = 3k + 2 for any positive integer
k, then 2n − 1 = 6k + 3 is not a prime number. This justifies the claim.

Problem 4.
Extend AP and AQ to cut BC√ at M and N respectively. √ Then ABM and ACN are isosceles,
so that BM = 2 and NC = 2 2. Hence MN √ = 2 2 − 2. Now P Q is the segment joining the
MN
midpoints of AM and AN . Hence P Q = 2 = 2 − 1.
A ...
............... . ..
.... ... .. .............
.... .. .. .......
.... ... .... .......
....... .. .. .......
.......
.... . ..

............. Q .
.
..
. P ..
..
...................................
.. . .
.......
.......
.......
......
.... .............. ................... .......
.... .. .
......... . .
.. ................... .......
... ...... . ................ ......
...... .............. ..
. ..
. ..................... .............
..
. .. .............. .
. ..
. .............. .......
. . .................. ..
. ..
. ............. ......
.... ..
............................................................................................................................................................................................................................
B N M C
Problem 5.
Suppose the real numbers x1, . . . , xn , 1 ≤ xi ≤ 4 for i = 1, 2, . . . , n, satisfy the two given inequalities.
Then (xi − 1)(xi − 4) ≤ 0 so that xi + x4i ≤ 5. Equality holds for xi = 1 or xi = 4. From these
inequalities and the given ones, we obtain
 
7n 8n 1 1 1
5n = + ≤ x1 + x2 + · · · + xn + 4 + + ··· + ≤ 5n.
3 3 x1 x2 xn
Hence xi = 1 or 4, i = 1, 2, . . . , n. Suppose x1 = x2 = · · · = xk = 1 and xk+1 = xk+2 = · · · = xn = 4
for some index k. Then x1 + x2 + · · · + xn = k + 4(n − k) = 7n 3
. Hence 5n = 9k so that 9 divides
n. It follows that the smallest value of n is 9, with the numbers 1, 1, 1, 1, 1, 4, 4, 4 and 4.
 The Alberta High School Mathematics Competition
Part I, November 21, 2012

1. Each day, Mr. Sod visited pubs A, B, C and D in that order, always spending $35, $12, $40
and $27 at the respective places. His total expenditure at the pubs, from the beginning of
the month up to a certain moment that month, was $1061. Which pub would he be visiting
next?

(a) A (b) B (c) C (d) D (e) impossible total

2. Meeny, Miny and Moe were playing tennis. From the second game on, the one who sat out
the preceding game would replace the loser of that game. At the end, Meeny played 17 games
and Miny played 35 games. How many games did Moe play?

(a) 18 (b) 26 (c) 36 (d) 52

(e) not uniquely determined

3. A circle of diameter r is drawn inside a circle of diameter R. For which of the following pairs
(r, R) is the area of the smaller circle closest to half the area of the larger circle?

(a) (1, 3) (b) (2, 4) (c) (3, 5) (d) (4, 6) (e) (5, 7)

4. A quadratic polynomial f (x) satisfies f (0) = 1, f (1) = 0 and f (2) = 3. What is the value of
f (3)?

(a) −3 (b) 1 (c) 2 (d) 10 (e) none of these

5. ABCD is a square. E and F are points on the segment BC such that BE = EF = F C = 4

cm. The segments AF and DE intersect at G. What, in cm2, is the area of triangle EF G?
√
(a) 6 (b) 4 3 (c) 8 (d) 12 (e) none of these

6. For how many integers n ≥ 2 is the sum of the first n positive integers a prime number?

( a) 0 (b) 1 (c) 2 (d) 3 (e) more than 3

7. In a test, Karla solved four-fifth of the problems and Klaus solved 35 problems. Half of the
problems were solved by both of them. The number of problems solved by neither was a
positive one-digit number. What was this number?

(a) 1 or 2 (b) 3 or 4 (c) 5 or 6 (d) 7 or 8 (e) 9

8. What is the largest possible integer a such that exactly three of the following statements are
true: a < 1, a > 2, a < 3, a > 4 and a < 5?

(a) 0 (b) 1 (c) 2 (d) 3 (e) 4

 9. A rectangle with integer length and width in cm has area 70 cm2. Which of the following, in
cm, cannot be the length of the perimeter of the rectangle?
(a) 34 (b) 38 (c) 74 (d) 98 (e) 142
10. The positive integer n is such that between n2 + 1 and 2n2 there are exactly five different
perfect squares. How many such n can we find?
(a) 0 (b) 1 (c) 2 (d) 3 (e) more than 3
11. ABCD is a rectangle such that AD − AB = 15 cm. P QRS is a square inside ABCD whose
sides are parallel to those of the rectangle, with P closest to A and Q closest to B. The total
area of AP SD and BQRC is 363 cm2 while the total area of AP QB and CRSD is 1113 cm2.
What, in cm2, is the area of P QRS?
(a) 900 (b) 1600 (c) 2500 (d) 3600

(e) not uniquely determined

12. Weifeng writes down 28 consecutive numbers. If both the smallest and the largest number
are perfect squares, what is the smallest number she writes down ?

(a) 9 (b) 36 (c) 100 (d) 169

(e) not uniquely determined

1 1 1
13. If the positive numbers a and b satisfy 2 + 2 = , what is the maximum
a + 4b + 4 b + 4a + 4 8
value of a + b?
3 5
(a) 2
(b) 2 (c) 2
(d) 4 (e) none of these
14. The incircle of triangle ABC is tangent to AB and AC at F and E respectively. If BC = 1,
6 A = 90◦ and 6 B 6= 6 C, what is the distance from the midpoint of BC to EF ?
√ √ √ √
(a) 42 (b) 22 (c) 3 4 2 (d) 2

(e) not uniquely determined

15. At the beginning of the year, there were more robots than androids. On the first day of each
month, each robot made 7 androids and each android made 7 robots. The next day, each old
android would pick a fight with a new android, and they would destroy each other. At the
end of the year, there were 46875 million robots and 15625 million androids. What was the
difference between the numbers of robots and androids at the beginning of the year?

(a) less than 10 (b) at least 10 but less than 100 (c) at least 100 but less than 1000

(d) at least 1000 but less than 10000 (e) at least 10000

16. Let m and n be positive integers such that 11 divides m + 13n and 13 divides m + 11n. What
is the minimum value of m + n?

(a) 24 (b) 26 (c) 28 (d) 30 (e) 34

 The Alberta High School Mathematics Competition
Solution to Part I, 2012

1. Note that 35 + 12 + 40 + 27 = 114 and 1061 = 9 × 114 + 35. Thus Mr. Sod had spent $35 on
the tenth day of that month at pub A. The answer is (b).
2. Since Meeny played 17 games, Miny and Moe played each other at most 17 + 1 = 18 times,
and each could play at most 18 + 17 = 35 games. As Miny played 35 games, Moe did not
play Meeny but played Miny 18 times. The answer is (a).
3. We want ( Rr )2 to be close to 12 . We have ( 13 )2 < ( 24 )2 < ( 35 )2 < ( 46 )2 = 4
9
and ( 57 )2 = 25
49
. Since
25
49
− 12 = 98
1
< 181
= 12 − 49 , the answer is (e).
4. Let f (x) = ax2 + bx + c. Then 1 = f (0) = c, 0 = f (1) = a + b + c and 3 = f (2) = 4a + 2b + c.
We have c = 1, a + b = −1 and 2a + b = 1. Hence a = 2 and b = −3, so that f (3) = 10. The
answer is (d).
5. Triangles GAD and GF E are similar, with AD = 3EF . Hence the vertical height of triangle
EF G is 13 of the vertical height of triangle ADG. Hence it is equal to 14 AB = 3 cm so that
the area of triangle EF G is 12 × 3 × 4 = 6 cm2. The answer is (a).
A D
J

JG

J

J

J
B E F C
n(n+1)
6. The sum of the first n positive integers is 2 . Suppose n is even. Then we must have
either n2 = 1 or n + 1 = 1. Both lead to n = 2. Suppose n is odd. Then we must have either
n+1
2
= 1 or n = 1. However, n = 1 is not allowed by the hypothesis. The answer is (b).
7. The fraction of problems solved only by Karla was 45 − 12 = 10 3
so that the total number of
3 7
problems is a multiple of 10. The fraction of problems solved by Klaus was at most 1− 10 = 10 .
Thus the total number of problems was at least 50. If it were 50, then 10 problems were solved
by Klaus alone, and as Karla solved 45 × 50 = 40 problems, the number of problems solved by
neither is 0. The total number of problems could only be as large as 70 since 35 problems would
be solved by both. In this case, the number of problems solved by neither was 15 × 70 = 14.
It follows that the total number of problems must be 60, of which 30 were solved by both, 5
3
by Klaus alone, 10 × 60 = 18 problems by Karla alone, and 60 − 30 − 5 − 18 = 7 by neither
of them. The answer is (d).
8. Note that a > 2 and a < 3 cannot both be true as there are no integers between 2 and 3.
Similarly, a > 4 and a < 5 cannot both be true. Since exactly three of the statements are
true, a < 1 must be true. Hence the largest possible value is a = 0, and for this value, the
three statements a < 1, a < 3 and a < 5 are true and the two statements a > 2 and a > 4
are false. The answer is (a).
 9. We have 70 = 1 × 70 = 2 × 35 = 5 × 14 = 7 × 10. Thus there are four possible shapes of the
rectangle, with respective perimeters 142 cm, 74 cm, 38 cm and 34 cm. The answer is (d).

10. Solving (n + 5)2 < 2n2 < (n + 6)2 yields 50 < (n − 5)2 and (n − 6)2 < 72. Thus 8 ≤ n − 5 and
n − 6 < 9 or 13 ≤ n ≤ 14. The answer is (c).

11. We shade the regions AP QB and CRSD while leaving the regions AP SD and BQRC un-
shaded. Extend the sides of P QRS to the perimeter of ABCD, creating four rectangles at
the corners each of which consists of two congruent triangles, one shaded and one unshaded.
The difference between the total area of the unshaded regions (not counting P QRS) and the
total area of the shaded regions is 1113 − 363 = 750 cm2. The difference in the lengths of
AD and AB is 15 cm. Hence the side length of P QRS is 750 ÷ 15 = 50 cm, and the area of
P QRS is 2500 cm2, The answer is (c).
A D
...............................................................................................................................................................................................................
........................................................................................................................................................................................................................
................................ ............................................................................
........................................
P
..................................
...............................
S
............................................................
...........................................................................
.............................................
........................ .........................................................................
. ... ...
.......................................... ..............................................................
......................... .............................................
................................ .........................................................................
....................................... ............................................................
.......................... ............................................................
................................ ............................................................
....................................... ............................................................
.......................... ............................................................
....................................... ..............................................................
......................... ..........................................................
................................ ............................................................
....................................... ............................................................
.......................... ............................................................
...................................... ............................................................
............................... ............................................................
.................. ............................................................
. ... ...
........................................ ............................................................
.......................... .............................................
Q R
.........................................................................................................................................................................................................................................................
.................... .. ... .......................................
.......... ... . ..... .....
.............................................................................................................................................................................................
B C

12. Let the smallest and the largest numbers Weifeng writes down be n2 and m2 respectively.
Since they are the ends of a block of 28 consecutive numbers, (m + n)(m − n) = m2 − n2 = 27.
We may have m + n = 27 and m − n = 1, whereby m = 14 and n = 13. We may have
m + n = 9 and m − n = 3, whereby m = 6 and n = 3. Thus the smallest number Weifeng
writes down may be 32 = 9 or 132 = 169. The answer is (e).

13. We have
1 1 1
= 2 + 2
8 a + 4b + 4 b + 4a + 4
1 1
= +
(a − 2)2 + 4a + 4b (b − 2) + 4b + 4a
1 1
≤ +
4(a + b) 4(a + b)
1
= .
2(a + b)

Hence a + b ≤ 4. This maximum value is attained if and only if a = b = 2. The answer is (d).
14. Let M be the midpoint of BC and D the point where the circle is tangent to BC. Let
B 0, C 0 and M 0 be the respective projections of B, C and M on EF . Now AEF is a right
isosceles triangle. Hence so are BB 0F and CC 0E. Hence BB 0 = BF
√ and CC 0 = CE
2
√ , so that
2 √
M M 0 = 12 (BB 0 + CC 0) = 2
1
√
2
(BF + CE) = 2
1
√
2
(BD + CD) = BC
√
2 2
= 4
2
. The answer is (a).
A ......
..... .....
...
B0
.........
.. .......................
....
.......
...
...
...
........................ . ................................ ...
... ....................... .. .
.............. ........... .....
...
...
..
.......................
F
........................ ....... .
. .. .
.........................
... .........................
....... ...
..... ...
......
......
...
.... .
.....
......... ..
.. 0
.......................
E
.............................
........................
...
...
.
.... ..
. . .
..
.. ... M .
.
.
.
..
...
....
.....
........................
....................... C
....................
0
... ...... .... ..
. ..... .
... ..... .
.. .
.
.
. .. .... ...
.
... .. ... ..
.
.. .... ...
. ... .. ... ..
.. .... ..... ... .. .... ..
... .... .. .... ... ..
.... ....... .. ..
.
. . ... .
.
... ... ... .
.. .... ... ... ... ... ..
... .... .. ... ... ..
... ....
... ... .. ... ...
.... ....... ..
. .
... .... ...
.. ..
.
.. .. ...
.. .... ... ... .. ... ..
... ..... ... ... ... ... ...
... .... ... .. ... ... ..
.... ... .. ... ... ..
.... ..... .
.... .. .
. . ... .... .
.
... .... ..... .. .... ... ...
.. ...... ....... .... ... .
. ...... ... ..
... .... ... ......... ......... ..
.............................................................................................................................................................................................................................................................................................................................................
B M D C

15. Let the numbers of robots and androids be r and a respectively. After one month, these
numbers became r+7a and 7r−a. After another month, they became (r+7a)+7(7r−a) = 50r
and 7(r + 7a) − (7r − a) = 50a. Hence after a two-month period, the number of robots became
50 times the original number, and the same goes for the number of androids. There being 6
two-month periods in a year, the initial number of robots was 46875000000 ÷ 506 = 3 and the
initial number of androids was 15625000000 ÷ 506 = 1. The answer is (a).

16. Since 13 divides 6(m + 11n) = (6m + n) + 13(5n), 13 divides 6m + n. Since 11 divides
6(m + 13n) = (6m + n) + 11(7n), 11 also divides 6m + n. Hence 11 × 13 = 143 divides 6m + n,
so that 6m+n = 143k for some integer k. Since 6(m+n) = 143k +5n = 6(24k +n)−(k +n), 6
divides k +n so that k +n ≥ 6. Now 6(m+n) = 143k +5n = 138k +5(k +n) ≥ 138+30 = 168.
Consequently m + n ≥ 28, and this is attained if m = 23 and n = 5. The answer is (c).
 The Alberta High School Mathematics Competition
Part II, February 6, 2013.

Problem 1.
Determine all pairs of positive integers (a, b) with a ≤ b such that
  
6 6
a+ b+ = 25.
b a

Problem 2.
A set S of positive integers is called perfect if any two integers in S have no common divisors greater
than 1. Candy wants to build a perfect set of numbers between 1 and 20 inclusive, in such a way
that her set contains as many numbers as possible.

(a) How many elements will her set have?

(b) How many different such sets can she build?

Problem 3.
Randy plots a point A. Then he starts drawing some rays starting at A, so that all the angles he
gets are integral multiples of 10◦ . What is the largest number of rays he can draw so that all the
angles at A between the rays are unequal, including all angles between non-adjacent rays?
Problem 4.
In a convex pentagon of perimeter 10, each diagonal is parallel to one of the sides. Find the sum of
the lengths of its diagonals.
Problem 5.
Find all integers r > s > t and all quadratic polynomials of the form f (x) = x2 + bx + c such that
b and c are integers, r + t = 2s, f (r) = 1, f (s) = b and f (t) = c.
 The Alberta High School Mathematics Competition
Solution to Part II, 2013.

Problem 1.
36
The given equation may be rewritten as ab + ab
+ 12 = 25. Therefore

(ab)2 − 13ab + 36 = (ab − 4)(ab − 9) = 0.

Hence ab = 4 or ab = 9. Note that a and b are positive integers with a ≤ b. If ab = 4, we have

(a, b) = (1, 4) or (2,2). If ab = 9, we have (a, b) = (1, 9) or (3,3). It is easy to verify that all four
are indeed solutions.
Problem 2.

(a) Candy’s perfect set may be {1, 2, 3, 5, 7, 11, 13, 17, 19}. We claim that this number is the
highest possible. Now a maximal perfect set must contain the element 1, as otherwise we can
add 1 and obtain a larger perfect set. Also, a maximal perfect set cannot contain an element
which is divisible by two distinct primes, as otherwise we can replace that element by the two
primes and obtain a larger perfect set. Hence each element pther than 1 is a positive power
of a prime. Moreover, distinct elements are powers of distinct primes. Since there are only 8
primes less than 20, namely, 2, 3, 5, 7, 11, 13, 17 and 19, we claim is justified.

(b) Every maximal perfect set Candy can build must have the form

S = {1, 2i2 , 3i3 , 5i5 , 7i7 , 11i11 , 13i13 , 17i17 , 19i19 },

where each exponent is a positive integer. Since 52 > 20, the exponent for all primes greater
than or equal to 5 must be 1. Since 24 ≤ 20 ≤ 25 and 32 ≤ 20 ≤ 33 , the exponent for 2 must
be 1, 2, 3 or 4, and the exponent for 3 must be 1 or 2. This yields eight different maximal
perfect sets.

Problem 3.  
Let n ≥ 2 be the number of rays drawn by Randy. Then there are n2 = n(n−1) 2
pairs of rays. Each
◦
pair determines two angles adding up to 360 . Hence the total number of angles between 2 of the n
rays is exactly n(n − 1). The measure of such an angle is clearly less than 360◦ . Since it is supposed
to be an integral multiple of 10◦ , there are at most 35 values for the measures of these angles. Since
they are distinct, n(n − 1) ≤ 35. Now 6 × 5 = 30 < 35 < 42 = 7 × 6. Hence n ≤ 6. It is possible
for Randy to draw 6 rays, determining 30 distinct angles. In the diagram below, 6 U AV = 60◦ ,
6 V AW = 40◦ , 6 W AX = 10◦ , 6 XAY = 20◦ , 6 Y AA = 140◦ and 6 ZAU = 90◦ .

X..........................W ......... V
..............................

Y......................... ........ ......... ..

...
......
.. ........
....
..
.. .... .. .. ... ...
.. ... .. ... .. ...
... ...
... .... ....
.
...
...
.. ... .. ... .. ...
... .... .. .. .
. ..
.
. .... ...... .... ..
..
... ... .... ..
... ..... .. ..
.. ......... .. . .
... .......................................................................
...
...
... A
...
...
.. ..
..
... U
... ... ...
.. .
... .
.
.. ... ..
... ..
... ... ..
... ...
. .....
.... ... ...
.... ... ....
..... ... ....
.... ....
.......
.........
...
. .
...
........
.
...........................................

Z
We now verify that the 30 angles between two rays are distinct. We have 6 W AY = 30◦ , 6 V AX = 50◦ ,
6 V AY = 70◦ , 6 U AW = 100◦ , 6 U AX = 110◦ , 6 U AY = 130◦ , 6 ZAV = 150◦ , 6 XAZ = 160◦ as well

as 6 W AZ = 170◦ . These are 9 different angles distinct from the 6 between adjacent rays. All have
measures less than 180◦ . Corresponding to these 15 angles, we have 15 other angles greater than
180◦ , yielding a total of 30 distinct angles.
Problem 4.
Let L be the point of intersection of EC and DB. Let M be the point on the extension of AB
such that M C is parallel to AE. Then ABLE and AMCE are parallelograms. Note that triangles
DLC and EAB are similar, as are triangles AMC and ELD. It follows that
EC EL + LC LC DL DL AB
= =1+ =1+ =1+ =1+ .
AB AB AB EA CM EC
√
Let x = EC
AB
. Then x = 1 + x1 so that x2 − x − 1 = 0. Hence x = 1+2 5 . Similarly, we have
DB AC AD EB
√
1+ 5
√
AE
= ED
= BC
= DC
= 2
, so that EC + DB + AC + AD + EB = 5(1 + 5).
A .......
.......... ....
.... .. .. ....
.... .... .... .......
........ .
. ..
..
....
....
.. .
..... .. .. ....
.... .. .. ....
. .. .... ..
.
.
..
..
....
....
....... .
.
. ..
. ....
....
.. .. .
. .. ....
B ...
. . ... .
.
...
. .
.. ... E
.............................................................................................................................................................
.
.... .......... .. .. ... ..
.... .
. . .. ..... .
.
.. ..
. . ....... ...
..... .. .... .. .. .
M ........
....
...
..
..
..
....
....
.... ..
.... ..
.
.
..
..
..
..
.
.. ..... .
.. ....
....
..
..
..
.... .. ..... .. ... ..
.. ..
... ..
... .. .. .. ... .. ..
.... .. .. ........ .... .... ..
... .. .. .... .... .. ..
.... .. .. .... ........ .. ...
.... ..
. .
... ........ ... ..
.... . ..
... ..
.. ..
. ..... ....... ..
.. ..
... .... ..
.... ... ....
... .. ..
.... .. .. .......
....
.... L ....
....
..... .... ....
...... .. .
..
.
. .
..
... .. ... ..... ... .. ..
............
.................................................................................................

C D
Remark:
The regular pentagon is used in the illustrative diagram. Many students√may get the correct answer
by treating only this special case, essentially proving that cos 36◦ = 1+4 5 .
Problem 5.
The conditions are:
r2 + br + c = 1, (1)
s2 + bs + c = b, (2)
t2 + bt + c = c, (3)
r+t = 2s. (4)
From (3), t(t + b) = 0 so that either t = 0 or t = −b. We consider these two cases separately.
Case 1: t = 0.
From (4), we have r = 2s. Substituting into (1), we have 4s2 + 2bs + c = 1. Subtracting (2) from
this, we have 3s2 + bs = 1 − b which may be rewritten as (s + 1)(3s − 3 + b) = −2. Hence 2 is
divisible by s + 1, so that s = −3, − 2, 0 or 1. However, since s > t = 0, we may only have s = 1.
It follows that b = −1. Hence f (x) = x2 − x − 1, with r = 2, s = 1 and t = 0.
Case 2: t = −b.
From (4), we have r = 2s + b. Substitoting into (1), we have 4s2 + 6sb + 2b2 + c = 1. Subtracting (2)
from this, we have 3s2 + 5sb + 2b2 = 1 − b which may be rewritten as (3s + 2b + 3)(s + b − 1) = −2.
Hence −2 is divisible by s + b − 1. From r > s > t = −b, we have s + b > 0. Hence s + b − 1 > −1 so
that s + b − 1 = 1 or 2. If s + b − 1 = 1, we have 3s + 2b + 3 = −2 so that s = −9 and b = 11. Hence
f (x) = x2 + 11x + 30 with r = −7, s = −9 and t = −11. If s + b − 1 = 2, we have 3s + 2b + 3 = −1
so that s = −10 and b = 13. Hence f (x) = x2 + 13x + 43, with r = −7, s = −10 and t = −13.
 The Alberta High School Mathematics Competition
Solution to Part I, 2013.

1. Of the first 100 positive integers 1, 2, . . . , 100, the number of those not divisible by 7 is

(a) 14 (b) 15 (c) 85 (d) 86 (e) none of these

Solution:
When 100 is divided by 7, the quotient is 14, with a remainder of 2. Thus 14 of the first 100
positive integers are divisible by 7. It follows that the number of these integers not divisible
by 7 is 100 − 14 = 86. The answer is (d).

2. The total score of four students in a test is 2013. Ace scores 1 point more than Bea, Bea
scores 3 points more than Cec, and Cec scores 2 points more than Dee. None of their scores
is divisible by

(a) 3 (b) 4 (c) 5 (d) 11 (e) 23

Solution:
Since the average score is just over 500, we try 500 as a score for Dee. Then Cec’s score is
502, Bea’s is 505 and Ace’s 506. The total is indeed 2013, so that no adjustment is necessary.
Now 500 is divisible by 4, 500 and 505 are divisible by 5, and 506 is divisible by 11 and 23.
The answer is (a).

3. Of the following five fractions, the largest one is

1 2 3 4 6
(a) 75
(b) 149
(c) 224
(d) 299
(e) 449

Solution:
12
The lowest common numerator is 12. The fractions then become , 12 , 12 , 12
900 894 896 897
and 12
898
respectively. The answer is (b).

4. Two teams A and B played a soccer game on each of seven days. On each day, the first
team to score seven goals wins. There were no ties. Over the seven days, A won more games
than B, but B scored more goals than A overall. The difference in the total numbers of goals
scored by B and A is at most

(a) 17 (b) 18 (c) 19 (d) 20 (e) none of these

Solution:
B won at most three games, and for each of these games, B wins by at most 7 goals with a 7
to 0 score. In the other four games, B loses by at least 1 goal with a 7 to 6 score. The goal
difference is at most 7 × 3 − 1 × 4 = 17. The answer is (a).
5. ABCD is a quadrilateral with AB = 12 and CD = 18. Moreover, AB is parallel to CD
and both 6 ADC and 6 BCD are less than 90◦ . P and Q are points on side CD such that
AD = AP and BC = BQ. The length of P Q is

(a) 6 (b) 7 (c) 8 (d) 9 (e) 10

Solution:
From A and B, drop perpendiculars onto CD at the points E and F respectively. Since both
6 ADC and 6 BCD are less than 90◦ , E and F do lie on the segment CD. Note that

DE + F C = CD − EF = CD − AB = 6.

Since ED = EP and F C = F Q, EP + F Q = 6 < 12 = EF , P is closer to E than Q and Q

is closer to F than P . Therefore, P Q = EF − (EP + F Q) = 12 − 6 = 6. The answer is (a).
A B
C
A
 C
A
 C
A
 C
A
 C
A
 C
A
 C
A
 C
A
D E P Q F C

6. Each of four cows is either normal or mutant. A normal cow has 4 legs and always lies. A
mutant cow has either 3 or 5 legs and always tells the truth. When asked how many legs they
have among them, their respective responses are 13, 14, 15 and 16. The total number of legs
among these four cows is

(a) 13 (b) 14 (c) 15 (d) 16 (e) none of these

Solution:
Since all four responses are different, at least three of them are wrong. If all four are wrong,
then the cows are all normal, and they will have 16 legs among them. However, this makes
one of the responses right. Hence one of the responses is indeed right. The three normal cows
have 12 legs among them. Hence the mutant cow must have 3 legs in order to make one of
the responses right. The answer is (c).
a
7. Let a and b be positive integers such that ab < 100 and b
> 2. Denote the minimum value of
a
b
by m. Then we have

(a) m ≤ 2.15 (b) 2.15 < m < 2.2 (c) m = 2.2

(d) 2.2 < m < 2.25 (e) m > 2.25

Solution:
We have 100 > ab > b(2b+1) so that b ≤ 6. Thus the minimum value of m is 2+ 16 = 2.1666 . . ..
The answer is (b).
8. Let ABCD be a quadrilateral with 6 DAB = 6 CBA = 90◦ . Suppose there is a point P on
side AB such that 6 ADP = 6 CDP and 6 BCP = 6 DCP . If AD = 8 and BC = 18, the
perimeter of the quadrilateral ABCD is

(a) 70 (b) 72 (c) 74 (d) 76 (e) 78

Solution:
Since 6 DAB + 6 CBA = 180◦ , AD is parallel to BC. Therefore, 6 ADC + 6 BCD = 180◦ .
Hence 6 P DC + 6 P CD = 90◦ . Consequently, 6 DP C = 90◦ . Let Q be the foot of the
perpendicular on CD from P . Note that triangle P DA is congruent to triangle P DQ, and
triangle P CB is congruent to triangle P CQ. Hence DQ = DA = 8, CQ = CB = 18 and
P A = P Q = P B. Since 6 DP C = 90◦ , triangle DP Q is similar to triangle P CQ. Hence
DQ
QP
= PQC
Q
. Therefore, P Q2 = 8 × 18 = 144 and P Q = 12. The perimeter of ABCD is
therefore AP + P B + BC + CQ + QD + DA = 12 + 12 + 18 + 18 + 8 + 8 = 76. The answer
is (d).
..........
........ .. ..
C
......... .... ....
.
............... ... ..
.
........ ...
. ...
......... ... ....
........
......... .. ....
..
............. .... ..
............. ..
..
...
. .
........... ... ....
.
........... ... ..
............. ... ...
....
......... .
.. ....
...........
. ... ..
............. . .. ...
.......... .. ...
Q ..
.................
...... ...
... ....
...
...
............. ... ...
. ...
...
......... ..
. .
.. ...
..
........... ..
. ... ...
............
. ..
....
...
.
............ ..
. ... ...
...... .. ..
D
..............
....
................ ...
... ........ ...
...
... .......
... .
.. ....
..... ... .
. ..
... ..... . ... ...
... .. ..
... ....... .
... .. ....
... ...... .. .. ..
... . ... ...
... ..... .. ..
... .....
..... .. .. ....
... ..... ... .. ..
... ..... .. ... ...
... ..... .. ... ....
... ..... .. ... ..
... ..... .. ..
... ....... ... .
.
. ...
... ...... ... ... ...
. ....... . .. ...
.... ..... ..
. .... ...
..... .. ..
.. ..... .
.. .. ...
... ..... .. ...
.... ..... .... .... ...
... ..... .. ... ...
..... ... ..
... ...... .. ..
.............................................................................................................................................................................................................................................................................................................
A P B

9. Two bus routes stop at a certain bus stop. The A bus comes in one-hour intervals while the
B bus also comes in regular intervals of a different length. When grandma rests on the bench
by the bus stop, one A bus and two B buses come by. Later, grandpa rests on the same bench
and eight A buses come by. The minimum number of B buses that must have come by during
that time is

(a) less than 4 (b) 4 or 5 (c) 6 or 7 (d) 8 or 9 (e) more than 9

Solution:
To minimize the number of B buses coming by, we stretch the length of their intervals as far
as possible. Suppose grandma sees the A bus that comes at 10:00. Then she does not see
the ones that come at 9:00 or 11:00. Thus the length of the interval between two B buses is
strictly less than 2 hours. The two B buses grandma sees may have come at 9:01 and 10:59,
in which case the interval is of length 118 minutes. Grandpa is on the bench for at least 7
hours. This is longer than three intervals for the B buses, so he must have seen at least 3 of
them. Suppose he sees the A buses at 11:00, 12:00, 13:00, 14:00, 15:00, 16:00, 1700 and 18:00.
Then he will see only the B buses at 12:57, 14:55 and 16:53. The answer is (a).
10. Suppose that 162013 = ab , where a and b are positive integers. The number of possible values
of a is

(a) 2 (b) 8 (c) 11 (d) 16 (e) 24

Solution:
Since 162013 = 24×2013, a must be of the form 2k where k is a positive integer divisor of
4 × 2013 = 22 × 3 × 11 × 61. The prime factorization of k may contain up to two 2s, one 3,
one 11 and one 61, so that there are (2 + 1)(1 + 1)3 = 24 possible values of k, and therefore
of a. The answer is (e).

11. The following five statements are made about the integers a, b, c, d and e: (i) ab is even and
c is odd; (ii) bc is even and d is odd; (iii) cd is even and e is odd; (iv) de is even and a is odd;
(v) ea is even and b is odd. The maximum number of these statements which may be correct
is

(a) 1 (b) 2 (c) 3 (d) 4 (e) 5

Solution:
If a, c and d are odd while b and e are even, then (i), (ii) and (iv) are all correct. Suppose
at least four statements are correct. By symmetry, we may assume that they are (i), (ii), (iii)
and (iv). However, by (i) and (ii), c and d are both odd, and yet by (iii), cd is even. This is
a contradiction. The answer is (c).

12. A very small cinema has only one row of five seats numbered 1 to 5. Five movie-goers arrive
one at a time. Each takes a seat not next to any occupied seat if this is possible. If not, then
any seat will do. The number of different orders in which the seats may be taken is

(a) 24 (b) 32 (c) 48 (d) 64 (e) 72

Solution:
The first two movie-goers to arrive may take the pair (1,5), (1,4), (2,5), (1,3), (2,4) or (3,5)
of seats. In each case, there are 2!=2 ways for them to do so. If they take (1,5), (1,3) or
(3,5), then the third movie-goer has only one choice of seat. The remaining two seats may be
occupied in 2!=2 ways. Otherwise, the last three movie-goers may take any vacant seats, and
this can be done in 3!=6 ways. Hence the total number of orders in which the seats may be
taken is 2 × 3 × (2 + 6) = 48. The answer is (c).

13. Let f (x) = x2 + x + 1. Let n be the positive number such that f (n) = f (20)f (21). Then the
number of distinct prime divisors of n is

(a) 1 (b) 2 (c) 3 (d) 4 (e) more than 4

Solution:
Note that f (m − 1)f (m) = (m2 − m + 1)(m2 + m + 1) = m4 + m2 + 1 = f (m2 ). Substituting
m = 21 yields f (20)f (21) = f (441). Therefore, n = 441 = 32 72 . The answer is (b).
14. The number of pairs (x, y) of integers satisfying the equation x2 + y 2 + xy − x + y = 2 is

(a) 3 (b) 4 (c) 5 (d) 6 (e) none of these

Solution:
The equation can be written as x2 + (y − 1)x + y 2 + y − 2 = 0. By the Quadratic Formula,
the solutions are
q q
−(y − 1) ± (y − 1)2 − 4(y 2 + y − 2) −(y − 1) ± −3(y − 1)(y + 3)
x= = .
2 2
These are real if and only if (y + 3)(y − 1) ≤ 0. Since y is an integer, it must be one of −3,
−2, −1,0 and 1. If y = −3, then x = 2. If y = −2, then x = 0 or 3. If y = −1, x is not an
integer. If y = 0, then x = −1 or 2. Finally, if y = 1, then x = 0. The answer is (d).

15. A triangle ABC with AB = 7, BC = 8 and CA = 10 has an interior point P such that
6 AP B = 6 BP C = 6 CP A = 120◦ . Let r1 , r2 and r3 be the radii of the circles passing

through the vertices of triangles P AB, P BC and P CA respectively. The value of r12 + r22 + r32
is

(a) 71 (b) 72 (c) 73 (d) 74 (e) 75

Solution:

Let O be the centre of the circle passing through the vertices of tiangle P AB. Note that since
6 AP B > 90◦ , O lies on the perpendicular bisector of AB outside of triangle P AB. Since

OA = OP = P B,

6AOB = 6 AOP + 6 P OB = 180◦ − 26 AP O + 180◦ − 26 BP O = 360◦ − 26 AP B = 120◦ .

√ 2 2
Hence AB = 3OA so that OA = AB √ . It follows that r2 = AB . Similarly, r2 = BC and
3 1 3 2 3
CA2 72 +82 +102
r32 = 3
. Hence r12 + r22 + r32 = 3
= 71. The answer is (a).
...............................................................
............. ..........
........
........ ........
...... ......
.. .........
. .....
A ....
.....
...... ...............
. . ... .... .......
......
.. ... .... .... .............
.. .. . . .. ..
... ... ... ... ...........
... ... .. ... ... ...
.. ... ... ... ... ..
... .... ... ..
..
.
.... ... ....
.
.....
......
. . ..
... .... ..
.
. .
.. .....
...
... .. . . .
..
. .. .. ....
. .. . .
.. .....
..
. ...
. ..
. .
. .. ...
.. ...
.... ..
.
... ..
.
..
.
.. .. ...
... .... .
.
.. .
..
.. ...
.. ...
... .... .. .
... .. ...
.. .
. .
. ... .. ...
.. ... .. .. ...
...
...
...
O .. ..
......
.
.
.. .......
.
.
.
.
..
.
.
..
...
..
....
.
.
.
...
...
...
...
.. ..... . . .
.. . . ...
... . .
. . . .
.. ...
... .. . ...... .
. .
... . ...
... .
.. ...... .
.. .
. ...
. .. . .. .
.
. ...
... .. ..... .
. .
.. .
... ... .. ...... . . .
.. .. ...
. .. . .. ...
.. .. ...... .
.. .. ...
.. .
.. .. .. ... ... . . ...
.. ... .......... ... .. ...
.. .. . . ...
.. .. . .
. .
..... . . ...
... .. ..
. . . ...... .. .
.. ..
. ...
... .. .. . ...
... ... .
. .. .
....... . ...
.. .
. .
....... ...
...
... .. .... ......... ................. ...
....... .......... ...
....
....
...
....
..
..
.
..
.. ...
.
.
.
..
...
...
.. .....
. P
.
.. ....
. ..
............
..
. ...........
. ..
..
...
...
...
...
..... .. .. ... ..... ..........
.
. . ..
. ..
. .......... ...
..... .. ..
. ..... ...... .
...
..........
...
..... .. .. ........... ..
......
....... . . .
.................. . .
. ..
. .......... .....
... .
.........
........... ......
.. ......................................................................................................................................................................................
. . .
.................... .........
......................... ........... .
B C
16. The list 1, 3, 4, 9, 10, 12, 13, . . . contains in increasing order all positive integers which can
be expressed as sums of one or more distinct integral powers of 3. The 100-th number in this
list is

(a) 981 (b) 982 (c) 984 (d) 985 (e) 999

Solution:
If we switch the powers of 3 to the powers to 2, then we get all the positive integers. Hence
we convert 100 into base 2, obtaining 100 = 26 + 25 + 22 . It follows that the 100-th number
on the list is 36 + 35 + 32 = 981. The answer is (a).
 The Alberta High School Mathematics Competition
Solution to Part II, 2014

Problem 1.
ABCD is a square. The circle with centre C and radius CB intersects the circle with diameter AB
at E 6= B. If AB = 2, determine AE.
Solution: √ √
Let O be the midpoint of AB. Then OB = 1 and OC = 12 + 22 = 5. The line OC of centres of
the two circles is perpendicular to the common chord BE. AE is also perpendicular to BE since E
lies on the circle with diameter AB. Now 6 ABE = 90√◦ − 6 COB = 6 OCB. Hence triangles ABE
and OCB are similar, so that AE = AB·OB OC
= 2×1
√ = 2 5.
5 5

D.............................................................................................................................................................
C
... .
.. ...
..... .. ..
.. ..
........ ... ..
....... ... ....
...... . ...
..
... ... .. ...
.... .... .. ....
... .. ... ..
... ... ... ...
.... .... ... ...
... .. ... ...
... ... ... ...
.. ... ..
. ...
... ... ..
....................................... .. . ...
....
...E ...
...
... ..... .....
........
.......
. ... ..........
...........
...
...
... .
. . .
.................. .. .. .... ...
... .. .. ...... ...
... .
....... .................. ... ..... ...
.. ... .... ........ ..
... .
.. .... .. . . ...
... .... .... .... .............. .
..
.
...
... ...
.
... ... ... .... ....... . .
..... ........ .. . .. ...
... .. ... ..... .......
. ........
.. ..
.. ... .. .....
...... ... . ....... .. ...
.
....... ... ....... .. .......
........
.. ...
....
..... ... .
.... .. .................
. ........
.........
....
...
.. ..........
....... .. .... ....... ...... ...
................................................................................................................................................................................

A O B
Problem 2.
A family consists of two parents of the same age, and a number of children all of different ages.
The average age of the children is 15, and the average age of the whole family is 21. When each
child was born, the parents were at least 25 and at most 35 years old. All ages are given in whole
numbers of years. Find all possible values of the number of children in this family.
Solution:
Let the age of the parents be p and let the number of children be n. Then the total age of the
family is 15n + 2p = 21(n + 2), which simplifies to p = 3n + 21. Since the children are of different
ages and their average age is 15, the age of the eldest one is at least 15 + n−12
= n+29
2
. It follows
that 3n + 21 = p ≥ n+29 2
+ 25 = n+79
2
. This simplifies to 5n ≥ 37 which implies that n ≥ 8.
n−1 31−n
On the other hand, the age of the youngest child is at most 15 − 2 = 2 . It follows that
3n + 21 = p ≤ 31−n
2
+ 35 = 101−n
2
. This simplifies to 7n ≤ 59 which implies that n ≤ 8. It follows
that the only possible value is n = 8. This may be realized if the children are of ages 11, 12, 13, 14,
16, 17, 18 and 19, and both parents are of age 45.
Problem 3.
Two cars 100 metres apart are travelling in the same direction along a highway at the speed limit
of 60 kph. At one point on the highway, the speed limit increases to 80 kph. Then a little later,
it increases to 100 kph. Still later, it increases to 120 kph. Whenever a car passes a point where
the speed limit increases, it instantaneously increases its speed to the new speed limit. When both
cars are travelling at 120 kph, how far apart are they?
Solution:
Let the first car be at a point B while the second car is at a point A, both in the 60 kph zone. Then
AB = 100 metres. Let the first car be at a point D while the second car is at a point C, both in
the 120 kph zone. Now the amount of time the second car takes to go from A to C is the same as
the amount of time the first car takes to go from B to D. Both cars take the same amount of time
going from B to C. Hence the amount of time the second car takes to go from A to B at 60 kph
is the same as the amount of time the first car takes to go from C to D at 120 kph. It follows that
CD = 2AB = 200 metres.
Problem 4.
Let p(x)be a polynomial with integer coefficients such that p(1) = 5 and p(−1) = 11.
(a) Give an example of p(x) which has an integral root.
(b) Prove that if p(0) = 8, then p(x) does not have an integrral root.
Solution:
(a) We are given two pieces of information. So we seek a polynomial with two undetermined
coefficients. The first attempt is p(x) = ax+b. Then 5 = p(1) = a+b and 11 = p(−1) = −a+b.
Hence a = −3 and b = 8, but the only root of −3x + 8 = 0 is x = 83 , which is not integral.
However, it is easy to modify our polynomial to p(x) = 8x2 − 3x. We have p(1) = 5 and
p(−1) = 11, but this time, we have an integral root x = 0 in addition to x = 38 .

(b) Suppose p(x) has an integral root x = r. Then r − 1 divides p(r) − p(1) = −5, so that r is
one of −4, 0, 2 or 6. Also, r + 1 = r − (−1) divides p(r) − p(−1) = −11, so that r is one
of −12, −2, 0 and 10. The only common value between the two lists is r = 0, but p(0) = 8.
This is a contradiction.

Problem 5.
On a 2 × n board, you start from the square at the bottom left corner. You are allowed to move
from square to adjacent square, with no diagonal moves, and each square must be visited at most
once. Moreover, two squares visited on the path may not share a common edge unless you move
directly from one of them to the other. We consider two types of paths, those ending on the square
at the top right corner and those ending on the square at the bottom right corner. The diagram
below shows that there are 4 paths of each type when n = 4. Prove that the numbers of these two
types of paths are the same for n = 2014.

Solution:
The path of the marker is uniquely determined by its vertical moves. The only condition is that
no two vertical moves can be made in adjacent columns. Whether the path ends in the upper or
lower right corner is determined by the parity of the number of vertical moves. Let the columns
be represented by elements in the set {1, 2, . . . , n}. Consider all subsets which do not contain two
consecutive numbers. Let an be the number of such subsets of even size, and bn be the number of
such subsets of odd size. Then a0 = a1 = a2 = 1 because of the empty subset, b0 = 0, b1 = 1 and
b2 = 2.
For n ≥ 3, classify the subsets of {1, 2, . . . , n} into two types, those containing n − 1 and those not
containing n − 1. A subset of the first type cannot contain either n − 2 or n. Hence the number of
such subsets of even size is bn−3 and the number of such subsets of odd size is an−3 . The subsets
of the second type may be divided into pairs such that in each pair, the two subsets are identical
except that one contains n and the other does not. Hence the number of such subsets of even size
is equal to the number of such subsets of odd size. It follows that an − bn = bn−3 − an−3 . Hence

a3k − b3k = (−1)k (a0 − b0 ) = (−1)k ,

a3k+1 − b3k+1 = (−1)k (a1 − b1 ) = 0,
a3k+2 − b3k+2 = (−1)k (a2 − b2 ) = (−1)k+1 .

In particular, a2014 = b2014.

Bulgarian Mathematical
Olympiad
(Third and Fourth Round)

1995-2001
 BULGARIAN NATIONAL OLYMPIAD IN MATHEMATICS
Third round

1995

Problem 1. Let p and q be positive numbers such that the parabola y = x2 ; 2px + q has
no common point with the x-axis. Prove that there exist points A and B on the parabola such
that the segment AB is parallel to the x-axis and 6 AOB = 90 (O is the coordinate origin) if
and only if p2 < q  14 . Find the values of p and q for which the points A and B are de ned in
an unique way.
Solution. Since the parabola has no common point with the x-axis, then the roots of the
equation x2 ; 2px + q = 0 are not real and hence p2 < q . Let the points A(x1 y0 ) and B (x2  y0)
(Figure 1) be with the required properties. Then x1 and x2 are the roots of the equation
x2 ; 2px + q ; y0 = 0 and y0 > q ; p2, because the vertex of the parabola has coordinates
(p q ; p2 ). On the other hand OA2 = x21 + y02 , OB 2 = x22 + y02 , AB 2 = (x1 ; x2 )2 and it follows
from the Pythagorean theorem that y02 + x1x2 = 0. But x1 x2 = q ; y0 and thus y02 ; y0 + q = 0.
Consequently the existence of the points A and B is equivalent to the assertion that the equation
f (y) = y2 ; y + q = 0 has a solution y0 > q ; p2. (A and B are de ned in an unique way if
this is the only solution.) A necessary condition is that the discriminant of the equation is not
negative, i.e. q  41 . The last condition is sucient because f (q ; p2) = (q ; p2 ) + p2 > 0 and
1 > 1  q ; p2 . The corresponding solution y is unique i q = 1 .
2 4 0
4
Figure 1. Figure 2.

1
 Problem 2. Let A1A2A3A4A5A6A7, B1B2B3 B4B5B6B7, C1C2C3C4C5C6C7 be regular
heptagons with areas SA , SB and SC , respectively. Let A1 A2 = B1 B3 = C1C4. Prove that
1 < SB + SC < 2 ; p2:
2 SA
Solution. Let A1 A2 = a, A1 A3 = b, A1 A4 = c (Figure 2). By the Ptolomeus theorem for the
quadrangle A1 A3 A4 A5 it follows that ab + ac = bc, i.e. a + a = 1. Since 4A1A2 A3 
= 4B1 B2 B3 ,
b c
then B1 B2 = a and hence B B = a . Analogously C C = a . Therefore SB + SC = a + a .
2 2 2 2

B12B3 2 b 1 2
b 1 2
c SA b2 c2
a a 1 a a 1 a a
Then + > ( + )2 = (equality is not possible because 6= ). On the other hand
b c 2 b c 2 b c

a2 + a2 = a + a 2 ; 2a2 = 1 ; 2a2 :
b2 c2 b c bc bc (1)

a 2 sin2 7 1 2 <
By the sine theorem we get bc = = . Since cos
sin2 7 sin 47 4 cos 27 (1 + cos 27 ) 7
p p
cos 4 = 22 , then abc > p 1 p = 2 ; 1. From here and from (1) we get the right hand
 2

2 2
4 2 (1 + 2 )
side inequality of the problem.
Problem 3. Let n > 1 be an integer. Find the number of the permutations (a1 a2 : : : an) of
the numbers 1 2 : : : n with the following property: there exists only one index i 2 f1 2 : : : n ;
1g such that ai > ai+1 .
Solution. Denote by pn the number of the permutations with the given properties. Obviously,
p1 = 0 and p2 = 1. Let n  2. The number of the permutations with an = n is equal to pn;1.
Consider all the permutations
! (a1  a2 : : : an ) with ai = n, where 1  i  n ; 1 is xed. Their
number is ni ;; 11 . Consequently
nX  !
;1
n;1 n;1
pn = pn;1 +
i=1 i ; 1 = pn;1 + 2 ; 1:
From here
pn = (2n;1 ; 1) + (2n;2 ; 1) +    + (2 ; 1)
= 2n ; n ; 1:
Problem 4. Let n  2 and 0  xi  1 for i = 1 2 : : : n. Prove the inequality
n
(x + x +    + x ) ; (x x + x x +    + x x + x x ) 
1 2 n 1 2 2 3 n;1 n n 1 : 2
When is there an equality?
Solution. Denote by S (x1 x2 : : : xn ) the left hand side of the inequality. This function is
linear with respect to each of the variables xi . Particularly,
S (x1 x2 : : : xn)  max (S (0 x2 : : : xn) S (1 x2 : : : xn)) :
2
 From here it follows by induction that it is enough to prove the inequality when all xi are
equal to 0 or 1. On the other hand for arbitrary xi we have
2S (x1 x2 : : : xn) = n ; (1 ; x1 )(1 ; x2) ; (1 ; x2)(1 ; x3 ) ;   ; ( )
;(1 ; xn)(1 ; x1) ; x1x2 ; x2x3 ;    ; xn x1
i.e. S (x1 x2 : : : xn )  n , when xi 2 0 1]. In the case when xi are equal to 0 or 1, the left
2 n
hand side of the last inequality is an integer. Consequently S (x1 x2 : : : xn)  2 . It follows
from ( ) that

when n is even, the equality is satis ed i (x1 x2 : : : xn) = (0 1 0 1 : : : 0 1)

when n is odd, the equality is satis ed i (x1 x2 : : : xn) = (x 0 1 0 1 : : : 0 1), where
x 2 0 1] is arbitrary.
Problem 5. The points A1, B1, C1 lie on the sides BC , CA, AB of the triangle ABC
respectively and the lines AA1 , BB1 , CC1 have a common point M . Prove that if the point M
is center of gravity of 4A1B1 C1, then M is the center of gravity of 4ABC .
Solution. Let M be the center of gravity of 4A1B1 C1. Let A2 be a point on MA! such
that B1 A1 C1A2 is a parallelogram. The points B2 and C2 are constructed analogously. Since
A1C1kA1B1kC1B2, then the points A2, C1, B2 are colinear and C1 is the midpoint of A2B2.
The same is true for the points A2 , B1 , C2 and C2 , A1 , B2 . We shall prove that A2 = A, B2 = B
and C2 = C , which will solve the problem.
Assume that A2 6= A and let A be between A2 and M . Then C2 is between C and M , B is
between B2 and M and consequently A2 is between A and M , which is a contradiction.
Problem 6. Find all pairs of positive integers (x y) for which xx ;+ yy is an integer which
2 2

is a divisor of 1995.
Solution. It is enough to nd all pairs (x y ) for which x > y and x2 + y 2 = k(x ; y ), where
k divides 1995 = 3:5:7:19. We shall use the following well-known fact: if p is a prime number
of the type 4q + 3 and if it divides x2 + y 2 , then p divides x and y . (For p = 3 7 19 this can
be proved directly.) If k is divisible by 3 then x and y are divisible by 3 too. Simplifying by
9 we get an equality x21 + x21 = k1(x1 ; y1 ), where k1 divides 5:7:19. Considering 7 and 19 in
an analogous way we get an equality a2 + b2 = 5(a ; b) (it is not possible to get an equality
a2 + b2 = (a ; b)), where a > b. From here (2a ; 5)2 + (2b ; 5)2 = 50, i.e. a = 3, b = 1 or a = 2,
b = 1.
The above considerations imply that the pairs we are looking for are of the type (3c c),
(2c c), (c 3c), (c 2c), where c = 1 3 7 19 3:7 3:19 7:19 3:7:19.

3
 BULGARIAN NATIONAL OLYMPIAD IN MATHEMATICS
Fourth round

1995

Problem 1. Find the number of all integers n > 1, for which the number a25 ; a is divisible
by n for every integer a.
Solution. Let n be with the required property. Then p2 (p prime) does not divide n since
p2 does not divide p25 ; p. Hence n is a product of pairwise dierent prime numbers. On the
other hand 225 ; 2 = 2:32:5:7:13:17:241. But n is not divisible by 17 and 241 because 325  ;3
(mod 17) and 325  32 (mod 241). The Fermat theorem implies that a25  a (mod p)
when p = 2 3 5 7 13. Thus n should be equal to the divisor of 2  3  5  7  13, dierent from 1.
Therefore the number we are looking for is 25 ; 1 = 31.
Problem 2. A triangle ABC with semiperimeter p is given. Points E and F lie on the
line AB and CE = CF = p. Prove that the excircle k1 of 4ABC to the side AB touches the
circumcircle k of 4EFC .
Solution. Let P and Q be the tangent points of k1 with the lines CA and CB, respectively.
Since CP = CQ = p, then the points E , P , Q and F lie on the circle with center C and radius p.
We denote by i the inversion dened by this circle. Since i(P ) = P , i(Q) = Q, then i(k1) = k1.
On the other hand i(E ) = E and i(F ) = F . Hence i(k) is the line AB . But k1 touches AB and
thus k touches k1.
Problem 3. Two players A and B take stones one after the other from a heap with n  2
stones. A begins the game and takes at least 1 stone but no more then n ; 1 stones. Each
player on his turn must take at least 1 stone but no more than the other player has taken before
him. The player who takes the last stone is the winner. Find who of the players has a winning
strategy.
Solution. Consider the pair (m l), where m is the number of the stones in the heap and l
is the maximal number of stones that could be taken by the player on turn. We must nd for
which n the position (n n ; 1) is winning (i.e. A wins) and for which n it is losing (B wins).
We shall apply the following assertion several times: If (m l) is a losing position and l1 < l,
then (m l1) is losing too.
Now we shall prove that (n n ; 1) is a losing position i n is a power of 2.
Suciency : Let n = 2k , k  1. If k = 1 then B wins on his rst move. Assume that
(2k  2k ; 1) is a losing position and let consider the position (2k+1  2k+1 ; 1). If A takes at least
2k stones on his rst move, then B wins at ones. Let A take l stones, where 1  l < 2k . By
the inductive assumption B could play in such a way that he could win the game (2k  l) since
l  2k ; 1 the last move will be the move of B. After this move we get the position (2k  m) with
m  l, which is losing for A, according to the inductive assumption.
Necessity: It is enough to prove that if n is not a power of 2, then (n n ; 1) is a winning
position. Let n = 2 + r, where 1  r  2k ; 1. On his rst move A takes r stones and B is
k
faced to the position (2k  r), which is losing for B.

1
 Problem 4. The points C1, A1 and B1 lie on the sides AB, BC and CA of the equilateral
triangle ABC respectively in such a way that the inradii of the triangles C1AB1 , B1 CA1, A1 BC1
and A1 B1 C1 are equal. Prove that A1 , B1 and C1 are the midpoints of the corresponding sides.
Solution. We shall prove that BA1 = CB1 = AC1 (Figure Figure 1.
1). Assume the contrary and let BA1  CB1 > AC1 . Let 
be the rotation at 1200 which center coincides with the in center
of the incircle of 4ABC . This rotation transforms the incircles
of the triangles C1BA1 , A1CB1 and B1 AC1 to the incircles of
the triangles A1CB1 , B1 AC1 and C1BA1 , respectively. Let A2 =
(A1), B2 = (B1) and C2 = (C1). It follows that BB2 < BC1
and BC2 < BA1 . But the incircles of the triangles BC1 A1 and
BC2B2 have equal radii (because (4AC1B1 ) = 4BC2B2 ), which
is a contradiction.
Let r be the radius of the incircles of the triangles C1 AB1 ,
B1CA1, A1 BC1pand A1 B1C1. From the triangle B1 AC1 we have
r = 1 ; B2 1 C1 : 33 , and from 4A1B1C1 which is equilateral we
p
have r = B1 C1 : 63 . From here B1 C1 = 12 and consequently A1, B1 , C1 are midpoints of the
corresponding sides.
Problem 5. Let A = f1 2 : : : m + ng, where m and n are positive integers and let the
function f : A ! A be dened by the equations:
f (i) = i + 1 for i = 1 2 : : : m ; 1 m + 1 : : : m + n ; 1
f (m) = 1 and f (m + n) = m + 1:
a) Prove that if m and n are odd then there exists a function g : A ! A such that g (g (a)) =
f (a) for all a 2 A.
b) Prove that if m is even then m = n i there exists a function g : A ! A such that
g (g (a)) = f (a) for all a 2 A.
Solution. a) Let m = 2p +1, n = 2q +1 and g(i) = p + i +1 for i = 1 2 : : : p g(i) = q + i +1
for i = m + 1 m + 2 : : : m + q  g (2p + 1) = p + 1 g (p + 1) = 1 g (m + 2q + 1) = m + q + 1
g (m + q + 1) = m + 1. It is easy to check that g (g (a)) = f (a) for all a 2 A.
b) Let m = n and g (i) = m + i for i = 1 2 : : : m g (m + i) = i + 1 for i = 1 2 : : : m ; 1
g(2m) = 1.
For the converse let M = f1 2 : : : mg. It follows by the denition of f that the elements
of M remain in M after applying the powers of f with respect to superposition. Moreover,
these powers scoop out the whole M . The same is true for the set A n M . The function f is
bijective in TA and if there exists g verifying the condition, then g is bijective too. We shall prove
that g (M ) M = . It follows from the contrary that there exists i 2 M such that g (i) 2 M .
Consider the sequence i g (i) g 2(i) : : : and the subsequence i f (i) f 2(i) : : :. It is easy to see that
g(M ) = M . We deduce that there exists a permutation a1 a2 : : : am of elements of M , such
that g (ai) = ai+1 for i = 1 2 : : : m ; 1 g (am ) = a1 and f (a2i;1) = a2i+1 for i = 1 2 : : : s ; 1
f (a2s;1 ) = a1 , whereTm = 2s. The last contradicts to the properties of f which were mentioned
already. Thus g (M ) M = . Analogously g (A n M ) = A n M , if g (i) 2 A n M for i 2 A n M .
At last let us observe that when starting from an element of M and applying g we go to A n M ,
but when applying g for a second time we go back to M . The same is true for the set A n M .
2
From here and from the bijectivity of g it follows that M and A n M have one and the same
number of elements, i.e. n = m.
Problem 6. Let x and y be dierent real numbers such that xx ;; yy is an integer for some
n n

n ; yn
four consecutive positive integers n. Prove that xx ; y is integer for all positive integers n.
Solution. Let tn = xx ;; yy . Then tn+2 + b:tn+1 + c:tn = 0 for b = ;(x + y), c = xy, where
n n

t0 = 0, t1 = 1. We shall show that b c 2 Z. Let tn 2 Zfor n = m m + 1 m + 2 m + 3. Since

cn = (xy)n = t2n+1 ; tn :tn+2 2 Zwhen n = m m + 1, then cm  cm+1 2 Z. Therefore c is rational
and from cm+1 2 Zit follows that c 2 Z. On the other hand
b = tm tm+3 ;cmtm+1 tm+2 
i.e. b is rational. From the recurrence equation it follows by induction that tn could be rep-
resented in the following way tn = fn;1 (b), where fn;1 (X ) is a monic polynomial with integer
coecients and degfn;1 = n ; 1. Since b is a root of the equation fm (X ) = tm+1 , then b 2 Z.
Now from the recurrence equation it follows that tn 2 Zfor all n.

3
 BULGARIAN NATIONAL OLYMPIAD IN MATHEMATICS
Third round

1996

Problem 1. Prove that for all positive integers n 3 there exist an odd positive integers
x and y , such that
n n

7x2 + y 2 = 2 :
n n
n

Solution. If n = 3 we have x3 = y3 = 1.
Suppose that for an integer n 3 there are odd positive integers x , y , such that 7x2 + y 2 =
n n n n
2 . We shall prove that for each pair
n

 
X = x +2 y  Y = j7x 2; y j and X = jx ;2 y j  Y = 7x 2+ y
n n n n n n n n

we have 7X 2 + Y 2 = 2 +1 . Indeed,
n

2 2  
7 x 2 y + 7x 2 y = 2 7x2 + y 2 = 2  2 = 2 +1 :
n n n n
n n
n n

Since x and y are odd, i.e. x = 2k + 1 and y = 2l + 1 (k, l are integers), then
x + y = k + l + 1 and jx ; y j = jk ; lj, which shows that one of the numbers x + y
n n n n

n n n n n n

2 2 2
and j x
n ; y jn
is odd. Thus, for n + 1 there are odd natural numbers x +1 and y +1 with the
2 n n

required property.
Problem 2. The circles k1 and k2 with centers O1 and O2 respectively are externally tangent
at the point C , while the circle k with center O is externally tangent to k1 and k2 . Let ` be
the common tangent of k1 and k2 at the point C and let AB be the diameter of k, which is
perpendicular to `, the points A and O1 lie in one and the same semiplane with respect to the
line `. Prove that the lines AO2, BO1 and ` have a common point.
Figure 1. Figure 2.

1
 Solution. Denote by r, r1 and r2 the radii of k, k1 and k2, by M and N the tangent points
of k with k1 and k2, respectively and by P the common point of ` and AB (Figure 1).
It follows from O1 O2 ? ` and AB ? ` that 4BON  4CO2N . Then CNO2 = ONB , 6 6

and consequently the points C , N and B are colinear. Also, BN =

CN CO r
BO = r . Analogously,
2 2
A, M and C are colinear and MC AM = r .
r1
The lines AN , BM and ` have a common point H , which is the altitude center of 4ABC .
By Ceva's theorem we have: AP  BN  CM = AP  r  r1 = 1, from where:
PB NC MA PB r2 r
r1 = r2 : (1)
PB AP
Let now D1 and D2 be the common points of the line ` with the lines BO1 and AO2,
respectively. Obviously, 4O1CD1  4BPD1 (Figure 2), from where D CD1 = r1 . Analogously,
1P PB
CD2 = r2 and according to (1) we have CD1 = CD2 , which shows that D D . Thus, the
D2P AP D1 P D2P 1 2
lines AO2, BO1 and ` have a common point.
Problem 3. a) Find the maximal value of the function y = 4x3 ; 3x in the interval ;1 1].
 3b)Let 2a, b and c be real numbers and M be the maximal value of the function y =
4x + ax + bx + c in the interval ;1 1]. Prove that M 1. For which a, b, c is the equality
reached?  
Solution.a) Using that 4x3 ; 3x = 12x2 ; 3 =
0

Figure 3.  
12 x ; 12 x + 21 , we nd that the function
 
y = 4x3 ; 3x
has a local maximums when x =  12 . Then its maximal
value in the interval ;1 1]is the biggest
 among the
1 1
r r r r

-1 -1/2 0 1/2 1 numbers y (;1), y (1), y ; and y (Figure 3).

 2  2
But y (;1) = y ; 12 = y 12 = y (1) = 1, thus the
maximal value is equal to 1.
b) Let f (x) = 4x3 + ax2 + bx + c. Assume that there exist numbers a, b, c, for which
the maximal value M of the function y = jf (x)j in ;1 1] is less than 1, i.e. M < 1. Then
;1 < f (x) < 1 for all x 2 ;1 1]. ; 
Consider
 the function
 g(x) = f (x) ; 4x3 ; 3x = ax2 + (b + 3)x + c. We have g (;1) > 0,
g ; 12 < 0, g 21 > 0 and g (1) < 0. Consequently g(x) changes its sign at least 3 times,
which means that the quadratic equation ax2 + (b + 3)x + c = 0 has at least 3 dierent roots.
 3only if a = b + 3 = c = 0, i.e. if f (x) = 4x ; 3x. According to a) the maximal
This is possible 3

value of y = 4x ; 3x is  1.
The equality M = 1 is reached only when a = 0, b = ;3 and c = 0.
Problem 4. The real numbers a1 a2 : : : a (n 3) form an arithmetic progression. There
n

exists a permutation a 1  a 2  : : : a n of the same numbers, which is a geometric progression.

i i i

2
Find the numbers a1  a2 : : : a , if they are pairwise dierent and the biggest among them is
n

equal to 1996.
Solution. Let a1 < a2 <    < a = 1996 and q be the quotient of the geometric progression
n

a 1  a 2  : : : a n . We have q 6= 0 and q 6= 1. The numbers a n  a n;1  : : : a 1 form also a geometric

i i i i i i

progression which quotient is q1 . Thus, we can assume that jq j > 1, i.e. q > 1 or q < ;1. Then
ja 1 j < ja 2 j <    < ja n j, from where a 6= 0 for all i.
i i i i

More exactly, either all numbers are positive (q > 1) and then a 1 < a 2 <    < a n , which
i i i

together with a1 < a2 <    < a shows that a k = a , i.e. the numbers a1  a2 : : : a form
n i k n

an arithmetic as well as a geometric progression, or the numbers a 1  a 2  : : : a n change their

i i i

signs alternatively (q < ;1) and then the positive ones form an increasing geometric progression
with quotient q 2 , and the order is the same as in the arithmetic progression. (The numbers
a1 a2 : : : a could not be all negative, because a = 1996 > 0.)
n n

Assume now that 3 among the numbers a1 a2 : : : a are positive. Then 0 < a 2 < a 1 <
n n; n;

a and they form a geometric as well as an arithmetic progression. Therefore 2a 1 = a 2 + a

n n; n; n

and a2 1 = a 2 a . From here a 2 = a 1 = a , which is a contradiction.

n; n; n n; n; n

Thus at most two among the numbers are positive. Analogously, at most two among the
numbers are negative. Consequently, n  4.
Let n = 4. Then a1 < a2 < 0 < a3 < a4 and 2a2 = a1 + a3 , 2a3 = a2 + a4 . But q < ;1 and
the geometric progression is either a3  a2 a4 a1 or a2  a3 a1 a4. Let it be a3  a2 a4 a1. Then
a2 = a3q, a4 = a3q2 and a1 = a3 q 3 . Thus, 2a3q = a3q3 + a3 and 2a3 = a3q + a3q2 . From here
q = 1, which contradicts to q < ;1.
So n = 3. There are two possibilities:
I. a1 < a2 < 0 < a3 = 1996. Then the geometric progression is a2, a3 = a2q, a1 = a2q2. It
follows from 2a2 = a1 + a3 that 2a2 = a2q 2 + a2 q , i.e. q 2 + q ; 2 = 0. Thus, q = ;2, ;2a2 = 1996,
a2 = ;998 and the numbers are (;3992 ;998 1996).
II. a1 < 0 < a2 < a3 = 1996. Now the geometric progression is a2, a1 = a2q, a3 = a2q2.
From 2a2 = a1 + a3 we obtain 2a2 = a2 q + a2 q 2, i.e. again q = ;2. Therefore, a3 = 4a2 = 1996
and a2 = 499. The numbers are (;998 499 1996).
Problem 5. A convex quadrilateral ABCD, for which ABC + BCD < 180 , is given. 6 6 

The common point of the lines AB and CD is E . Prove that ABC = ADC if and only if 6 6

AC 2 = CD  CE ; AB  AE:
Solution. Let ABC =  , ADC =  , BAC = ' and CAD =
(Figure 4). The point A
6 6 6 6

is between E and B and the point D is between E and C . Also, AEC =  + ' +
; 180 . 6 

Applying the sine theorem to the trian- Figure 4.

gles ACD, ACE , ABC and again to 4ACE ,
we obtain:
CD = sin

AC sin 
CE = ; sin '
AC sin ( + ' +
) 
AB = sin ( + ') 
AC sin 
AE = ; sin ( +
) :
AC sin ( + ' +
)
3
 From here the equation AC 2 = CD  CE ; AB  AE is equivalent to the equations:
CD  CE ; AB  AE ; 1 = 0
AC AC AC AC
sin ( + ') sin ( +
) sin  ; sin
sin ' sin 
; sin  sin  sin (' +
+ ) = 0
(cos ( + ' ;  ;
) ; cos ( + ' +  +
)) sin 
; (cos (' ;
) ; cos (' +
)) sin 
; (cos ( ; ) ; cos ( + )) sin (' +
+ ) = 0
sin ( + ' +
) ; sin (2 + ' +
;  )
+ sin (2 +
;  ; ') ; sin ( +
; ') = 0
sin ( ;  ) cos ( + ' +
) + sin ( ;  ) cos ( +
; ') = 0
sin ( ;  ) sin ( +
)sin ' = 0
But sin ' 6= 0 and sin ( +
) =6 0. Consequently, sin ( ; ) = 0 and  = .
Problem 6. A rectangle m n (m > 1, n > 1) is divided into mn squares 1 1 with lines,
parallel to its sides. In how many ways could two of the squares be canceled and the remaining
part be covered with dominoes 2 1?
Solution. Denote by F (m n) the number we are looking for. Since every domino covers
exactly two squares, then F (m n) = 0 if m and n are odd.
Figure 5. Let at least one of the numbers m and n be even. We color the
squares in two colors | white and black in such a way that every
two neighbor squares (with common side) are of dierent colors
(Figure 5). The number S0 of the white squares is equal to the
number S1 of the black ones and S0 = S1 = mn 2 .
Each domino covers one white and one black square. If two
white or two black squares are canceled, then it is impossible to
cover by dominoes the remaining part of the rectangle. Now we
shall show that if one white and one black squares are canceled
then the remaining part can be covered by dominoes.
Since mn is even, then mn = 2t, where t 2. We make induction with respect to t. The case
t = 2 is obvious. Let t0 > 2 and the proposition is true for all 2  t  t0. Let mn = 2 (t0 + 1).
Denote by T1 the rectangle consisted of the rst two rows of the considered rectangle, and by T2
| the rectangle consisted of the remaining rows. If the two canceled squares are in T1 or in T2,
then we can cover each of the rectangles by dominoes and consequently we can cover the given
rectangle.
Let one of the canceled squares be in T1 , and the other | in T2. We place a domino in
such a way that it covers one square from T1 and one square from T2. It is also possible that
the canceled square from T1 and the covered square from T1 are of dierent colors. Thus, the
canceled square from T2 and the covered square from T2 are of dierent colors. According to
the inductive assumption the remaining part of T1 and the remaining part of T2 can be covered.
Thus, the rectangle can be covered too. 2
Finally one white and one black squares can be chosen in S0  S1 = mn ways. Therefore,
2
m
in this case F (m n) = 4 .
2 n2

4
 BULGARIAN NATIONAL OLYMPIAD IN MATHEMATICS
Fourth round

1996

Problem 1. Find all primary numbers p and q, for which (5 ; 2 pq

p p)(5q ; 2q )
is an integer.
Solution. Let p be prime number and pj (5p ; 2p ). It follows by the Fermat theorem that
5 ; 2p  3 (mod p). Consequently p = 3.
p

Let now p and q be such prime numbers that (5 ; 2 )(5 ; 2 ) is an integer. If pj (5p ; 2p ),
p p q q

pq
then p = 3. Since 53 ; 23 = 3  3  13, then either q j (5q ; 2q ), i.e. q = 3, or q = 13. Therefore the
pairs (3,3), (3,13), (13,3) satisfy the problem condition. It remains the case when p 6= 3 q 6= 3.
Now pj (5q ; 2q ) and q j (5p ; 2p ). We can assume that p > q . It is clear that (p q ; 1) = 1 and
consequently, there are positive integers a and b, for which ap ; b(q ; 1) = 1 (Bezou theorem).
Since (q 5) = (q 2) = 1, it follows by the Fermat theorem that 5q;1  2q;1 (mod q ). From
5p  2p( (mod q ) we deduce that 5ap  2ap (mod q ) and therefore, 5b(q;1)+1  2b(q;1)+1
(mod q ). But 5b(q;1)+1  5 (mod q ) and 2b(q;1)+1  2 (mod q ). Thus, q = 3, which is a
contradiction. Finally, (p q ) = (3 3) (3 13) (13 3).
Problem 2. Find the side length of the smallest equilateral triangle in which three disks
with radii 2, 3 and 4 without common inner points can be placed.
Solution. Let in a equilateral 4ABC two disks with radii 3 and 4 without common inner
points be placed. It is clear that a line ` exists, which separates them, i.e. the disks are in
dierent semiplanes with respect to ` (Figure 1).
Figure 1. Figure 2.

This line divides the triangle into a triangle and a quadrilateral or into two triangles. In
both cases the disks can be replaced in the gure in a way that each of them is tangent to two
of the sides of 4ABC . It is clear that the new disks have no common inner point. Let the disks
be inscribed in 6 A and 6 B of 4ABC , respectively. We translate the side BC parallelly to itself
1
 towards the point A, till the disk which is inscribed in 6 B touches the disk which is inscribed in
6 A (Figure 2). Thus, we get an equilateral 4A1B1 C1 with a smaller sides, in which two disks
with radii 3 and 4 and without common inner points are placed.
Figure 3. Let A1 B1 = x, I be the incenter of 4A1B1 C1, while O1
and
x O2 be the centers of the two disks. Then A1I = B1 I =
p  A1O1 = 6 B1O2 = 8. Since the disk with radius 4 is
3
inside thep4A1B1 C1, then O2 2 IB1 . Thus, B1 O2  B1 I ,
i.e. x  8 3. On the other hand O1I = px ; 6 O2I = px ;
3 3
8 O1O2 = 7 and 2 by the cosine
2 theorem for 4
 x O 1O 2 I we
 nd
x x x
that p ; 6 + p ; 8 + p ; 6 p ; 8 = 49.
3p 3 p3 3
Butpx  8 3, and from here x = 11 3. Consequently, AB 
11 3. On thep other hand in the equilateral 4ABC with
side length 11 3 three disks with radii 2, 3 and 4 (without
common inner points) can be placed inscribing circles with
these radii in the angles of the triangle (Figure 3).
Note that the disks wit radii 3 and 4 are tangentpto each other. It follows from the above
considerations that the solution of the problem is 11 3.
Problem 3. The quadratic functions f (x) and g(x) are with real coecients and have the
following property: if the number g (x) is integer for a positive x, then the number f (x) is integer
too. Prove that there are such integers m and n, that f (x) = mg (x) + n for all real x.
q
Solution. Let g (x) = px2 + qx + r. We can assume that p > 0. Since g (x) = p(x + )2 + r ;
q2
2p 4p
q
after the variable change of x by x + 2p we reduce the problem for the following quadratic
functions
s f (x) = ax2 + bxs+ c and!g(x) = px2 + s p > 0. Letsk be such
! an integer thatsk > s and
k ; s > q . Since g k ; s = k is integer, then f k ; s = a(k ; s) + b k ; s + c
p 2p p p p p
is an integer too. Consequently, the number
s ! s !
f k + 1 ; s ; f k ; s = pb : p 1 p +a (1)
p p p k+1;s+ k;s p
is an integer for all k which are suciently big. It follows from here that ap is an integer. Indeed,
suppose that a is not an integer. If b > 0, we chose k in a way that
p
a
pp k + 1 ; s + k ; 1 < p + 1 ; ap 
b :p 1 p

and if b < 0, we chose k in a way that

 
pbp : pk + 1 ; 1s + pk ; 1 > ap ; ap :
In both cases there is a contradiction with the fact that (1) is an integer. Now, it follows
that
pbp : pk + 1 ; 1s + pk ; s

2
for all k, which are suciently
s big.! This is possible only when b = 0.
Let = m. Then, f k ; s = m(k ; s) + c is an integer (when k is suciently big), i.e.
a
p p
c ; ms is an integer. Let n = c ; ms. Now it is clear that f (x) = mg(x) + n for all x.
Problem 4. The sequence fan g1n=1 is de ned by
a1 = 1 an+1 = ann + an  n  1:
n

Prove that ba2n c = n when n  4 (it is denoted by bxc the integer part of the number x).
x n
Solution. Let f (x) = + . Since f (a) ; f (b) =
(a ; b)(ab ; n2 ) , it follows that the
n x abn
function f (x) is decreasing in the interval (0 n).p
Firstly, by induction we shall prove that n  an  p n when n  3. We have
p p n ; 1n
a1 = 1 a2 = 2 and a3 = 2, i.e. 3  a3  p2 . Let n  an  p
3
for an integer n  3.
n ; 1
p p
Then, an+1 = f (an )  f ( n) = np+n1 and an+1 = f (an )  f p n = p n > n+1
n;1 n;1
and thus the induction
p nishes. p
Since an  n, it remains to prove, that an < n + 1. We have an+1 = f (an ) 
f pn = p n when n  3. Consequently, an  pn ; 1 when n  4. Then,
n;1 n;1 n;2
n ; 1  (n ; 1)2 + n2 (n ; 2) p
an+1 = f (an ) < f p = p < n+2
n;2 (n ; 1)n n ; 2
pn  an < p4.n +(The
when last inequality is equivalent to 2n2(n ; 3) + 4n ; 1 > 0.) Therefore,
1 when n  4, i.e. ba2n c = n.
n
Problem 5. The quadrilateral ABCD is inscribed in a circle. The lines AB and CD meet
each other in the point E , while the diagonals AC and BD | in the point F . The circumcircles
of the triangles AFD and BFC have a second common point, which is denoted by H . Prove
that 6 EHF = 90.
Solution. Let O be the circumcenter of ABCD. We shall prove that O is the second common
point of the circumcircles of 4AHB and 4CHD. (Since AB and CD are not parallel, then
O 6= H .) After that we shall prove that the points E , H and O are colinear and 6 OHF = 90.
We shall consider the possible positions of H .
Let G be the common point of AD and BC (these lines are not parallel because the cir-
cumcircles of 4AFD and 4BFC are not tangent). It is clear that H is in the interior of
6 AGB .
1) H is in 4CGD. Then, 6 CHD = 6 CHF + 6 DHF = 180 ; 6 CBF + 180 ; 6 DAF =

360 ; CD d > 180 (from CD d < AB d and AB d + CDd < 360), which is impossible.

2) H is in 4CFD. Then, 6 CHD = 360 ; 6 CHF ; 6 DHF = 6 CBF + 6 DAF = CD d=
6 COD and 6 AHB = 6 AHF + 6 BHF = 6 ADF + 6 BCF = AB d = 6 AOB .
3) H is in 4ABF . Analogously, 6 CHD = 6 COD and 6 AHB = 6 AOB .
4) H is on AB . Again 6 CHD = 6 COD and 180 = 6 AHB = 6 AOB . Consequently, O is
the midpoint of AB .
5) H is not in 4ABG. Then 6 CHD = 6 COD and 6 AHB = 360 ; 6 AOB .

3
 Note that in the cases 2), 3) and 5) the points O and H are in one and the same semiplane
with respect to the line AB and with respect to the line CD. Consequently, the points A B H
and O are concyclic. We have the same for the points C D H and O.
Now we shall prove that the line EH passes through the point O. Let this line meet the
circumcircles of 4AHB and 4CHD at the points O1 and O2 , respectively. Since these points
lie on the ray EH ! , it follows from the equalities EH:EO1 = EA:EB = EC:ED = EH:EO2
that O1  O2  O.
The points O and H are in one and the same semiplane with respect to CD and we can
assume that the quadrilateral COHD is convex. Let H be in 4CFD. Then, 6 OHF = 6 FHC ;
6 OHC = 180 ; 6 FBC ; 6 ODC = 180 ; 21 :6 COD ; (90 ; 12 :6 COD) = 90. The other
possibility for H is to be inside 6 AGB and outside the non-convex quadrilateral AFBG.
Hence 6 OHF = 6 OHC + 6 FHC = 6 ODC + 6 FBC = 90 .
Problem 6. A square table of size 7 7 with the four corner squares deleted is given.
Figure 4. a) What is the smallest number of squares which need to be colored black
so that a 5-square entirely uncolored Greek cross (Figure 4) cannot be found
on the table?
b) Prove that it is possible to writte integers in each square in a way that
the sum of the integers in each Greek cross is negative while the sum of all
integers in the square table is positive.
Solution. Denote the square in row i and column j by (i j ). Note that
a cross is uniquely determined by its central cell. The cross with central cell
(i j ) is denoted by Cij , 2  i j  6. The number of all crosses is 25.
a) The squares (1 i), (i 1), (7 i), (i 7), i = 2 : : : 6, are included in exactly one cross the
squares (2,2), (2,6), (6,2), (6,6) are included in exactly 3 crosses the squares (2 i), (i 2), (6 i),
(i 6), i = 3 4 5 | in exactly 4 crosses and nally, the squares (i j ) 3  i j  5 are included
in exactly 5 crosses. Since C22 contains a colored square, then at least one of (1,2), (2,1), (2,2),
(2,3), (3,2) should be colored. Similarly, at least one of (2,7), (1,6), (2,6), (2,5), (3,6), at least
one of (7,2), (6,1), (6,2), (5,2), (6,3) and at least one of (6,7), (7,6), (6,6), (6,5), (5,6) are colored.
Any of them is contained in at most four crosses (the rst two in each quintuple in 1 cross, the
third one in 3 crosses and the remaining two | in 4 crosses). Denote by x the number of the
colored squares among (i j ) 3  i j  5. The number of crosses containing a colored square is
not greater than 4:4 + 5x, whence
16 + 5x  25: (1)
Thus, x  2 and the number of the colored cells is at least 6.
Suppose that the number of the colored squares is 6. Then x = 2. Moreover, acording (1)
there exists at most one Greek cross containing more than one colored square. If two squares
with a common side or a common vertex are colored, then it is easy to check that there are two
crosses with at least two colored squares each. Therefore, the squares (4,4), (3,3), (3,5), (5,3),
(5,5) should be uncolored. With no loss of generality, let the two colored squares be (4,3) and
(4,5). But now the crosses C34 and C54 do not contain colored squares.
To prove that the minimal number of colored squares is 7, color for example the squares
(2,5), (3,2), (3,3), (4,6), (5,4), (6,2), (6,5).
b) Let us write ;5 in each of the colored squares from a) and 1 in the remaining squares.
Since every Greek cross contains a colored square, the sum of the numbers in its squares does not
exceed 1+1+1+1;5 = ;1 < 0. The sum of all numbers in the table is 7:(;5)+(45;7):1 = 3 > 0
and we are done.
4
XLV National Mathematics
Olympiad: 3rd round, April
1997
Problem 1. Find all natural numbers a, b and c such that the
roots of the equation
x2 ; 2ax + b = 0
x2 ; 2bx + c = 0
x2 ; 2cx + a = 0
are natural numbers.

Solution: Let fx1 x2g, fx3 x4g and fx5 x6g be the roots of the
rst, second and third equation respectively, and let all of them be
natural numbers.
Assume that x  2 for all i = 1 2 : : :  6. Then 2a = x1 + x2 
i

x1x2 = b, 2b = x3 + x4  x3x4 = c and 2c = x5 + x6  x5x6 = a.

Thus 2(a + b + c)  a + b + c, which is impossible, since a b c are
natural numbers.
1
 Therefore at least one of the numbers x equals 1. Without loss
i

of generality suppose x1 = 1, so 1 ; 2a + b = 0.
If x  2 for i = 3 4 5 6, then
i

2(b + c) = (x3 + x4) + (x5 + x6)  x3x4 + x5x6 = c + a

whence 2(2a ; 1+ c)  c + a ) c  2 ; 3a, which is impossible when
a b c are natural numbers.
So at least one of x3 x4 x5 x6 equals 1. Let x3 = 1. Now 1 ; 2b +
c = 0. Assuming that x5  2 and x6  2, we get 2c = x5 + x6  a,
so 2(2b ; 1)  b +2 1 . Thus 7b  5, a contradiction.
Therefore at least one of the numbers x5 x6 is 1 and it follows
that 1 ; 2c + a = 0. Further
0 = (1 ; 2a + b) + (1 ; 2b + c) + (1 ; 2c + a) = 3 ; (a + b + c)
and since a b c are natural numbers, it follows that a = b = c = 1.
Direct verication shows that a = b = c = 1 satisfy the conditions
of the problem.

Problem 2. Given a convex quadrilateral ABCD which can be

inscribed in a circle. Let F be the intersecting point of diagonals
AC and BD and E be the intersecting point of the lines AD and
BC . If M and N are the midpoints of AB and CD, prove that
MN = 1  AB ; CD :
EF 2 CD AB

2
Solution: Let 6 AEB =  , EC = c, ED = d, ;!i = 1c  ;! EC and
j = d1  ;!
;! ED. Since ABCD is an inscribed quadrilateral, CDAB =
AE = BE = k. Therefore ;! EA = kc;!
j and ;!
EB = kd ;!i . Since
CE DE
F 2 AC and F 2 BD there exist x and y such that
;!
EF = x;!
EA + (1 ; x);!
EC = xkc;!j + (1 ; x)c;!
i
and ;!
EF = y;!EB + (1 ; y);!
ED = ykd;! i + (1 ; y)d;!
j:
Comparing the coecients of ;!
i and ;!
j in these equalities gives
xkc = (1 ; y)d and ykd = (1 ; x)c. This implies x = (kkd2 ;;1)c c .
Therefore
;!  
EF = k2 k; 1 (kd ; c);!
j + (kc ; d);!
i
and thus
 !2  
k
EF = k2 ; 1 (kd ; c)2 + (kc ; d)2 + 2(kd ; c)(kc ; d) cos  :
2

On the other hand

;;! = 1  ;!
MN AD + ;!
BC

= 1  ;!
ED ; ;!
EA + ;!
EC ; ;!
EB

2 2  
= 12  (d ; kc);!j + (c ; kd);!i
and it follows that
 
MN 2 = 14  (d ; kc)2 + (c ; kd)2 + 2(d ; kc)(c ; kd) cos  :
3
  2 !2  2
Therefore EF 2 = 4 k k; 1 = 41 k ; k1 and so
MN 1
2

MN = 1  AB ; CD
EF 2 CD AB
Problem 3. Prove that the equation
x2 + y2 + z2 + 3(x + y + z) + 5 = 0
has no solution in rational numbers.

Solution: It is easy to see that the equation is equivalent to

(2x + 3)2 + (2y + 3)2 + (2z + 3)2 = 7:
It has a solution in rational numbers if and only if there exist
integer numbers a, b, c and a natural number m such that
() a2 + b2 + c2 = 7m2:
Suppose such numbers exist and give m its smallest possible
value. There are two cases:
(I) m = 2n is an even number. Now a2 + b2 + c2 is divisible by 4.
This implies that all numbers a, b, c are even ones, so a = 2a1,
b = 2b1, c = c1. It follows now that a21 + b21 + c21 = 7n2, which
contradicts the way m have been chosen.
(II) m = 2n + 1 is an odd number. Now m2 1 (mod 8) and
therefore a2 + b2 + c2 7 (mod 8), which is impossible when
a, b, c are integer.
4
Problem 4. Find all continuous functions f (x) dened in the set
of real numbers and such that
 1 
f (x) = f x + 4
2

for all real x.

Solution: Let f (x) be a function satisfying the conditions of the

problem. Obviously f (x) is an even function.
Let x0  0. There are two cases:
(I) 0  x0  12 . Consider the sequence
(1) x0 x1 : : :  x  : : :
n

dened by the equalities x +1 = x2 + 41 .

n n

It is easy to see by induction that 0  x  1 for all n. Moreover

n
2
 2
x +1 ; x = x2 ; x + 41 = x ; 12  0
n n n n n

which implies that (1) is a monotone increasing function. Since it is

!1 x = .
bounded, it follows that it is a convergent function. Let lim n

Now 2 ;  + 14 = 0, so  = 21 .
n

On the other hand, since f (x) is a continuous function,

1
lim
n!1 f ( xn ) = f 2 :
5
 But  
f (x +1) = f x2 + 41 = f (x )
n n n

 
for all n. Thus f (x0) = f (x1) =    which means that f (x0) = f 21
 1
for all x0 2 0 2 .
(II) x0 > 12 . Consider the following sequence:
(2) x0 x1 : : :  x  : : :
n

s
dened by x +1 = x ; 14 .
n n

As in the previous case, we show that (2)  1is a convergent function

1
!1 x = 2 . Further lim
and lim
n
n
!1 f (x ) = f 2 and since
n
n

 1 
f (x +1 ) = f x +1 + 4 = f (x )
n
2
n n

 
for all n we get that f (x0) = f 21 .
Therefore f (x) is a constant function in the interval 0 +1), and
since it is even, it is a constant for all x.
Conversely, any constant function satises the conditions of the
problem.

Problem 5. Two squares K1 and K2 with centres M and N and

sides of length 1 are placed in the plane in a way that MN = 4, two
of the sides of K1 are parallel to MN and one of the diagonals of
6
K2 lies on the line MN . Find the locus of midpoints of segments
XY where X is an interior point for K1 and Y is an interior point
for K2.

Solution: The locus is the interior of a regular hexagon centred at

the midpoint of the segment MN and with a side length of 21 .

To prove this, proceed as follows.

Fix the point Y in the interior of K2. When X varies in the
interior of K1 the locus of the midpoints of XY is a square K10 which
is homothetic to K1 by homothecy with centre Y and coecient 21 .
Obviously the side of this square is 21 and its centre is the midpoint
of MY . When Y varies in the interior of K2 then the locus of
the midpoints of MY is a square K20 which is homothetic to K2 by
homothecy with centre M and coecient 21 . The side of this square
equals 1 and its centre Q is the midpoint of MN . Finally, when
2
the centres of K10 vary in the interior of K20 , the squares vary in the
interior of a regular hexagon centred at Q and with a side length
of 12 .
7
Problem 6. Find the number of non-empty sets of S = f1 2 : : :  ng
n

such that there are no two consecutive numbers in one and the same
set.

Solution: Denote the required number by f . It is easy to see that

n

f1 = 1, f2 = 2, f3 = 4.
Divide the subsets of S having no two consecutive numbers into
n

two groups|those that do not contain the element n and those

that do. Obviously the number of subsets in the rst group is f ;1. n

Let T be a set of the second group. Therefore either T = fng or

T = fa1 : : : a ;1 ng, where k > 1. It is clear that a ;1 6= n ; 1,
k k

so fa1 : : :a ;1g  S ;2 , whence the number of sets in the second

k n

group is f ;2 + 1. Therefore
n

f = f ;1 + f ;2 + 1:
n n n

After substituting u = f + 1 we get

n n

u1 = 2 u2 = 3 u = u ;1 + u ;2 :
n n n

Therefore the sequence fu g1=1 coincides with the Fibonacci se-

n n
quence from its third number onwards. Thus we obtain
0 p ! +2  p ! +21
f = p @ 1 +2 5
1 ; 1 ;2 5 A ; 1:
n n

n
5

8
XLV National Mathematics
Olympiad: 4th round, May
1997
Problem 1. Consider
!
the polynomial
! ! !
Pn (x) = 2 + 5 x + 8 x + + 3k n+ 2 xk 
n n n 2

n 2 
where n 2 is a natural number and k = 3 .

;

(a) Prove that Pn+3 (x) = 3Pn+2 (x) 3Pn+1 (x) + (x + 1)Pn (x)
;

(b) Find
;
all integer numbers a such that Pn (a3) is divisible by
n
3 2 ] for all n 2.
1


Solution:
  (a) Compare the coe cients in front of xm, 0 m  
n + 1 . It su ces to show that
3 ! ! ! ! !
n+3 =3 n+2 n + 1 n n
3m + 2 3m + 2 3 3m + 2 + 3m + 2 + 3m 1 :
;
;

1
 ! ! !
Using the identity a +b 1 = ab + b a 1 we get that
;

! !! ! !!
n+3 n+2 2 n + 2 n + 1 +
3m + 2 ! 3m + 2 !! 3m + 2! 3m + 2
; ; ;

+ 3nm++12 ;
3 m
n
+ 2 ;
3m
n
1 =
! ! ! !
;

n + 2 n + 1
= 3m + 1 2 3m + 1 + 3m + 1 n n =
;

! !!
;
3 m 1 ;

= n + 2 n + 1 +
3m + 1 ! 3m + 1 !!
;

!
n+1 n n
;
3!m + 1 ! 3m + 1 ! 3m 1 =
; ;
;

= n3+m1 ;
n
3m ;
n
3m 1 =
;

= 0:
(b) Suppose a satises the condition of (b). Then P5(a3) =
10 + a3 is divisible by 9 and so a  ; 1 (mod 3). On the contrary,
let a  ; 1 (mod 3). Now a3 + 1 0 (mod 9). Since P2(a3) =


1 P3(a3) = 3 and P4(a3) = 6 it follows by induction from (a) that

Pn (a3) is divisible by 3 2 ] for any n. Therefore the required values
n;1

of a are all integer numbers congruent to 2 modulo 3.

Problem 2. Let M be the centroid of ABC . Prove the inequal-

4

ity
sin 6 CAM + sin 6 CBM  p
2
3
2
 (a) if the circumscribed circle of AMC is tangent to the line AB 
4

(b) for any ABC .

4

Solution: We use the standard notation for the elements of ABC .

4

Let G be the midpoint of AB .

(a) It follows from the conditions of the problem that
 c 2
= GA2 = GM GC = 1 m2c = 1 (2a2 + 2b2 c2)
2 3 12
;

and therefore a2 + b2 = 2c2. Using the median formula we get ma =

p p
3 b and m = 3 a. Further
b
2 2
 1  = (a2 + b2) sin  :
A = sin 6 CAM + sin 6 CBM = S bm1 + am
3ab
p
a b

From the Cosine Law a2 + b2 2ab cos  = c2 = a +2 b , so a2 + b2 =

2 2
;

4ab cos  . Therefore A = 2 sin 2 2 .

p  p
3 3
(b) There are two circles through C and M tangent to the line
AB . Denote the contact points by A1 and B1 and let A1 GA!, 2

B1 GB !. Since G is the midpoint of A1B1 and CM : MG = 2 : 1,

2

M must be the centroid of A1B1C . Furthermore it is clear that

4
6 CAM 6 CA1M and 6 CBM 6 CB1M . Suppose 6 CA1M 90
  

and 6 CB1M 90. Now sin 6 CAM + sin 6 CBM sin 6 CA1M +
 

sin 6 CB1M 2 .
 p
3

3
 It remains to consider the case 6 CA1M > 90, 6 CB1M 90 

(the above angles could not both be obtuse). It follows from CA1M
4

that CM 2 > CA21 + A1M 2, so

1 (2b2 + 2a2 c2) > b2 + 1 (2b2 + 2c2 a2)
9 1 ; 1 1 1
9 1 ;
1 1

(a1 b1 c1 are sides of A1B1C ). We know from (a) that a21 + b21 = 2c21
4

and the above inequality becomes a21 > 7b21. Again from (a) we
obtain
v
u !
b sin  b u
1 t a 2 + b2 2
sin 6 CB1M = 1 1
= 1 4a1b1 :
1 1
a1 3 a1 3
p p ;

Substituting ab12 = x we get that

2

1
s
sin 6 CB1M = 1p
14x x2 1 < 1 2 49 1 1= 1
7
p p
4 3 4 3
; ; ; ;

since x < 17 . Therefore

sin 6 CAM + sin 6 CBM < 1 + sin 6 CB1M < 1 + 71 < 2 : p
3
Note: The inequality holds only for ABC with angles  =
4

22:5,  = 112:5 ,  = 45 or  = 112:5 ,  = 22:5 ,  = 45 .

Problem 3. Let n and m be natural numbers such that m + i =

aib2i for i = 1 2 : : :  n, where ai and bi are natural numbers and ai
is not divisible by a square of a prime number. Find all n for which
there exists an m such that a1 + a2 + + an = 12.
4
Solution: It is clear that n 12. Since ai = 1 if and only if m + i is


a perfect square, at most three of the numbers ai equal 1 (prove it!).

It follows now from a1 + a2 + + an = 12 that n 7. 

We show now that the numbers ai are pairwise distinct. Assume

the contrary and let m + i = ab2i and m + j = ab2j for some 1 i < 

j n. Therefore 6 n 1 (m + j ) (m + i) = a(b2j b2i ). It

  ;  ; ;

is easy to see that the former is true only if (bi bj  a) = (1 2 2) or

(2 3 1) and in either case a1 + a2 + + an > 12.
All possible values of ai are 1, 2, 3, 5, 6, 7, 10 and 11. There are
three possibilities for n: n = 2 and a1 a2 = 1 11  2 10  5 7 
f g f g f g f g

n = 3 and a1 a2 a3 = 1 5 6  2 3 7  n = 4 and a1 a2 a3 a4 =

f g f g f g f g

f 1 2 3 6 . Suppose n = 4 and a1 a2 a3 a4 = 1 2 3 6 . Now

g f g f g

(6b1b2b3b4)2 = (m + 1)(m + 2)(m + 3)(m + 4) = (m2 + 5m + 5)2 1, ;

which is impossible. Therefore n = 2 or n = 3.

If n = 3 and (a1 a2 a3) = (1 5 6), then m = 3 has the required
property, and if n = 2 and (a1 a2) = (11 1), then m = 98 has the
required property. (It is not di cult to see that the remaining cases
are not feasible.)

Problem 4. Let a, b and c be positive numbers such that abc = 1.

Prove the inequality
1 + 1 + 1 1 + 1 + 1 :
1+a+b 1+b+c 1+c+a 
2+a 2+b 2+c

Solution: Let x = a + b + c and y = a1 + 1b + 1c = ab + bc + ca

(abc = 1). It follows from Cauchy's Inequality that x 3 and y 3.
 

Since both sides of the given inequality are symmetric functions of

5
a, b and c, we transform the expression as a function of x, y and
abc = 1. After simple calculations we get
3 + 4x + y + x2 12 + 4x + y 
2x + y + x2 + xy 9 + 4x + 2y


which is equivalent to
3x2y + xy2 + 6xy 5x2 y2 24x 3y 27 0:
; ; ; ; ; 

Write the last inequality in the form

( 35 x2y 5x2) + ( xy3 y2) + ( xy3 3y) + ( 43 x2y 12x) +
2 2
; ; ; ;

+ ( xy3 3x) + (3xy 9x) + (3xy 27)

2
; ; ; 0:

When x 3, y 3, all terms in the left hand side are non-negative

 

and the inequality is true. Equality holds when x = 3, y = 3, which

implies a = b = c = 1.

Problem 5. Given a ABC with bisectors BM and CN (M

4 2

AC , N AB ). The ray MN ! intersects the circumcircle of ABC

2 4

at point D. Prove that

1 = 1 + 1 :
BD AD CD
Solution: Let A1, B1 and C1 be the orthogonal projections of D
on the lines BC , CA and AB , respectively. It follows from DAB1
4

and the Sine Law that DB1 = DA sin 6 DAB1 = DA sin 6 DAC =
DA DC (R is the circumradius of ABC ). Analogously DA =
2R 4 1

6
DB DC and DC = DA DB . Our equality is now equivalent to
2R 1
2R
AD CD = BD CD + AD BD and so it su ces to prove that
(1) DB1 = DA1 + DC1
Denote by m the distance from M to AB and BC and by n the
DM = x (x > 1). Further,
distance from N to AC and BC . Let MN
DB1 = x, DC1 = x 1 and DA1 m = x. Therefore DB = nx,
;

n m ;
n m ;
1

DC1 = m(x 1) and DA1 = nx m(x 1) = DB1 DC1 and (1)

; ; ; ;

holds.

Problem 6. Let X be a set of n + 1 elements, n 2. Ordered 

n-tuples (a1 a2 : : : an) and (b1 b2 : : : bn) formed from distinct ele-
ments of X are called `disjoint' if distinct indices i and j exist such
that ai = bj . Find the maximal number of n-tuples any two of which
are `disjoint'.

Solution: For n 2 denote by A(n + 1) the maximum number of



ordered n-tuples such that any two of them are `disjoint'. Also let
S (X ) be a set of such n-tuples for which S (X ) = A(n + 1). It is
j j

clear that for any  X the following holds:

2

(a1 a2 : : : an) S (X ) a1 =  A(n):

jf 2 j gj 

Thus A(n + 1) (n + 1)A(n). Therefore



A(n + 1) (n + 1)n : : : A(3):



Direct verication shows that A(3) = 3, so A(n + 1) (n +2 1)! . 

7
 We prove now that A(n + 1) = (n +2 1)! by constructing a set of
(n + 1)! ordered n-tuples, any two of which are `disjoint'. We may
2
assume that X = 1 2 : : :  n + 1 . Consider a set E of all even
f g

permutations of 1 2 : : :  n + 1. (A permutation (a1 a2 : : : an+1) is

called even if the number of pairs (i j ) such that i < j and ai > aj
is an even number.) The set
f(a1 a2 : : :  an) (a1 a2 : : : an an+1) E
j 2 g

has (n +2 1)! ordered n-tuples, any two of which are `disjoint'.

8
XLVII National
Mathematics Olympiad:
3rd round, 25{26 April 1998
Problem 1. Find the least positive integer number n (n 3)
with the following property: for any colouring of n dierent points
A1 A2 : : :  An on a line and such that A1A2 = A2A3 = : : : =
An;1An in two colours, there are three points Ai Aj  A2j;i(1  i <
2j ; i  n) which have the same colour.

Solution: Assume the two colours are white and black. Consider
8 points coloured as follows: A1 A2 A5 A6 (white), A3 A4 A7 A8
(black). Obviously no three points Ai Aj  A2j;i (1  i < 2j ; i  n)
have the same colour and therefore n 9.
If we can show that n = 9 has the required property, we will be
done. Suppose there are 9 points coloured black or white and no
three points Ai Aj  A2j;i (1  i < 2j ; i  n) have the same colour.
First assume that for i = 3, i = 4 or i = 5 points Ai and Ai+2
have the same colour (say white). Then the points Ai;2 Ai+1 Ai+4
1
should be black (note that i ; 2 1 and i + 4  9), which is a
contradiction.
Suppose now that for i = 3 4 5 the points Ai and Ai+2 have
dierent colours. Without loss of generality assume A5 is a white
point. Then A3 and A7 are black. Because of the symmetry we may
suppose that A4 is white and A6 is black. Consiquently A8 is white,
A2 is black (2 + 8 = 2  5) and A9 is white (7 + 9 = 2  8). Therefore
A1 should be both white (1 + 3 = 2  2) and black (1 + 9 = 2  5),
which is again a contradiction.
Consequently the assumption is not true and so n = 9.

Problem 2. Let ABCD be a quadrilateral such that AD = CD

and 6 DAB = 6 ABC < 90 . The line passing through D and the
midpoint of the segment BC intersects the line AB at the point E .
Prove that 6 BEC = 6 DAC .

Solution: Let M be the midpoint of BC and let AD and BC

meet at point N and AN and EC meet at point P . It follows from
Menelaus' Theorem aplied to 4DMN and 4DEN that DP  NC 
ME = PN  CM  ED and DA  NB  ME = AN  BM  ED.
Combining the above equalities with AN = BN , BE = CE and
AD = CD, we get DP  NC = DC  PN . Therefore CP is bisector
of 6 DCN .
Consequently 6 ACP = 6 ACD + 6 DCP = 12 (6 NDC + 6 DCN ) =
6 NAB so 6 DCP = 6 CAB and we obtain 6 BEC = 6 ABC ;
6 BCE = 6 BAD ; 6 DCP = 6 DAC , Q. E. D.

2
 Note: The assertion is also true if the condition 6 ABC < 90
is left out.

Problem 3. Let R+ be the set of all positive real numbers. Prove

that there is no function f : R+ ! R+ such that
(f (x))2 f (x + y)(f (x) + y)
for arbitrary positive real numbers x and y.

Solution: Suppose there exists a function satisfying the conditions

of the problem. Write the initial equality in the form
f (x) ; f (x + y) f (fx()x+)y :
y
First we prove that f (x) ; f (x + 1) 21 for x > 0. Obviously
f is a (strictly) monotone non-increasing function. Fix x > 0 and
choose a natural number n, such that n  F (x + 1) 1. When
k = 0 1 : : :  n ; 1, we obtain that
f (x + nk ) n1
f (x + nk ) ; f (x + k +n 1 ) 1:
f (x + n ) + n 2n
k 1

Adding the above inequalities gives f (x) ; f (x + 1) 21 .

Let the natural number m be such that m 2f (x). Therefore
X m;1
f (x) ; f (x + m) = (f (x + i) ; f (x + i + 1)) m2 f (x)
i=0
and so f (x + m)  0. But this contradicts the fact that f is strictly
positive.
3
Problem 4. Let f (x) = x3 ; 3x + 1. Find the number of dierent
real solutions of the equation f (f (x)) = 0.

Solution: Since f 0(x) = 3(x ; 1)(x + 1) it follows that f is strictly

monotone non-decreasing in the intervals (;1 ;1] and 1 1) and
strictly monotone non-increasing in the interval ;1 1]. Moreover
lim f (x) = 1 f (;1) = 3 f (1) = ;1 f (3) = 19 > 0 and so
x!1
the equation f (x) = 0 has three distinct roots x1 x2 x3 such that
x1 < ;1 < x2 < 1 < x3 < 3. Therefore f (x) = x1 has only one real
root (which is less than ;1) and f (x) = x2 and f (x) = x3 have three
distinct real roots each (one in each of the intervals (;1 ;1) (;1 1)
and (1 1)). Since the roots of f (f (x)) = 0 are exactly the roots
of these three equations, we conclude that it has seven distinct real
roots.

Problem 5. The convex pentagon ABCDE is inscribed in a circle

with radius R. The inradii of the triangles ABC , ABD, AEC and
AED are denoted by rABC , rABD , rAEC and rAED . Prove that

a.) cos 6 CAB + cos 6 ABC + cos 6 BCA = 1 + rABC

R 
b.) If rABC = rAED and rABD = rAEC , then 4ABC
= 4AED.

Solution: a.) Using the standard notation for the elements of

4ABC , we obtain that
r = S = 4S 2 = 4(p ; a)(p ; b)(p ; c) =
R pR pabc abc

4
 1 (a(b2 + c2 ; a2) + b(c2 + a2 ; b2) + c(a2 + b2 ; c2) ; 2abc) =
= 2abc
= cos 6 CAB + cos 6 ABC + cos 6 BCA ; 1:
b.) Applying continuously the equality from a.) and using the
fact that ABCDE is inscribed in a circle convex pentagon, it is easy
to see that
rABC + rAEC + rEDC = rAED + rABD + rBCD :
The condition of the problem and the above equality imply rBCD =
rEDC . Since 4BCD and 4EDC have a side in common and equal
circumradiuses we get that 4BCD = 4EDC (prove it using a.)).
In particular BC = ED and using rABC = rAED again, we get
4ABC = 4AED.

Problem 6. Show that the equation

x2y2 = z2(z2 ; x2 ; y2)
has no solution in positive integer numbers.

Solution: Suppose x y z is a solution for which xyz (which is a

natural number|why?) has its minimum value.
Write x y z in the form x = dx1 y = dy1 z = dz1, where d =
(x y z). Our equation is equivalent to x21y12 = z12(z12 ; x21 ; y12). Let
u = (x1 z1) v = (y1 z1) x1 = ut y1 = vw. Since z1 divides x1y1,
we get z1 = uv. A substitution in the last equality gives

5
 (u2 + w2)(v2 + t2) = 2u2v2. Further, since (x1 y1 z1) = 1, it
follows that (u w) = 1 (v t) = 1. Therefore
u2 + w2 = v2 v2 + t2 = 2u2
or
u2 + w2 = 2v2 v2 + t2 = u2:
Without loss of generality we may assume that the rst pair of
equalities hold. It is easily seen that v and u are odd integer numbers.
It follows now from u2 + w2 = v2 that u = m2 ; n2 w = 2mn v =
m2 + n2, where m and n are coprime natural numbers (of distinct
parity). Substitution in v2 + t2 = 2u2 shows that t2 + (2mn)2 =
(m2 ; n2)2 and so t = p2 ; q2 mn = pq m2 ; n2 = p2 + q2 for some
natural numbers p and q. Therefore
p2 q2 = m2(m2 ; p2 ; q2)
which shows that p, q, m is a solution of the original equation. It re-
mains to be seen that pq
m = n < d ( p 2
; q 2
)2 mn = xy , which contra-
z
dicts the way we have chosen x y z.

6
XLVII National
Mathematics Olympiad: 4th
round, 16{17 May 1998
Problem 1. Let n be a natural number. Find the least natural
number k for which there exist k sequences of 0's and 1's of length
2n + 2 with the following property: any sequence of 0's and 1's of
length 2n + 2 coincides in at least n + 2 positions with some of these
k sequences.

Solution: We shall prove that k = 4. Assume that k 3 and let

the respective sequences be ai1 ai2 : : :  ai2n+2 for i = 1 : : :  k. Since
k 3 there is a sequence b1 b2 : : : b2n+2 such that (b2l+1 b2l+2) 6=
(ai2l+1 ai2l+2) for l = 0 1 : : :  n and i = 1 : : : k. This is a contra-
diction. For k = 4 it is easily seen that the sequences 000 : : : 0 011 : : : 1,
100 : : : 0 111 : : : 1 have the required property.

Problem 2. The polynomials Pn (x y) n = 1 2 : : : are dened by

P1(x y) = 1 Pn+1 (x y) = (x+y;1)(y+1)Pn(x y+2)+(y;y2)Pn (x y):
1
Prove that Pn (x y) = Pn (y x) for all x y and n.

Solution: We know that P1 (x y) = 1 and P2(x y) = xy + x +

y ; 1. Assume that Pn;1 (x y) and Pn (x y) (n  2) are symmetric
polynomials. Then
Pn+1 (x y ) = (x + y ; 1)( y + 1)Pn (x y + 2) + (y ; y
2
)Pn (x y )
= (x + y ; 1)( y + 1)Pn (y + 2 x) + (y ; y
2
)Pn (y x)
!
+ 1)(x + 1)Pn;1 (y + 2 x + 2)
= (x + y ; 1)(y + 1)
(x +
+ (x
y

; 2
x )Pn;1 (y + 2 x)
!
; 2) + ( (y + x ; 1)( + 1)x P n;1 (y x + 2)
+ (y y
x ) ; 2) n;1 (
x P y x

= ( + ; 1)( + 1)( + + 1)( + 1) n;1 ( + 2

x y y x y x P y x + 2)
+( ; )( ;
2 2
y y ) n;1 (
x )
x P y x

+ ( + ; 1)( + 1)( ;
2
x y y) n;1 ( + 2 )
x x P y x

+( ; )( + ; 1)( + 1) n;1 ( + 2 )
2
y y x y x P x y :

and by induction it follows that all the polynomials are symmetric.

Problem 3. On the sides of a non-obtuse triangle ABC a square,

a regular n-gon and a regular m-gon (n m > 5) are constructed
externally, so that their centres are vertices of a regular triangle.
Prove that m = n = 6 and nd the angles of ABC .

Solution: Let the square, the n-gon and the m-gon be constructed
on the sides AB , BC and CA, respectively. Denote their centres
by O1, O2 and O3 denote by A1, B1 and C1 the centres of the
equilateral triangles constructed externally on BC , CA and AB .

2
The lines O1C1 O2A1 and O3B1 intersect at the circumcentre O of
4ABC . Since 4A1A2A3 is equilateral, it follows straightforwardly
that 4O1O2O3 is equilateral if and only if C1A1jjO1O2 A1B1jjO2O3
and B1C1jjO1O3. This is equivalent to
OC1 = OA1 = OB1 = k:
C1O1 A1O2 B1O3
On the other hand,
OC1 = cot C + tan 30 
C 1 O1 cot 45 ; tan 30
OA1 = cot A + tan 30 
A1O2 cot 180n ; tan 30
OB1 = cot B + tan 30 :
B1O3 cot 180m ; tan 30
 180 = y. The above identities imply that
Set cot 180
n = x and cot m
cot A = kx ; kp+ 1  cot B = ky ; kp+ 1  cot C = k ; kp+ 1 :
3 3 3
From the identity cot A cot B + cot B cot C + cot C cot A = 1 we get
p
k = p 2(
p x + y + 1 ; 3) p 
3xy + ( 3 ; 2)(x + y) p + 3;2
cot C = p x +py ; xy + 3 ; 2 3p :
3xy + ( 3 ; 2)(x + y) + 3 ; 2
p p
Since m n  6 it follows that x y  3,i. e. xy  p3(x + y) ; 3.
The inequality cot C p0 implies p x + y ; xy + 3 ; 2 3  0. p and
therefore x + y + 3 ; 2 3 pxy  3(x + y) ; 3 ,i. e. x + y 2 3.
This shows that x = y = 3, i. e., m = n = 6. Hence cot C =
0 cot A = cot B = 1, so 6 C = 90 6 A = 6 B = 45 .
3
Problem 4. Let a1 a2 : : :  an be real numbers, not all of them
zero. Prove that the equation
p p p
1 + a1x + 1 + a2x + : : : + 1 + an x = n
has at most one nonzero real root.
X n
Solution: The given equation is equivalent to x p1 +ai a x = n.
i=1 i
a
Since the function p1 + a x (ai =
i
6 0) is strictly decreasing, it follows
i
that this equation has at most one nonzero root.

Problem 5. Let m and n be natural numbers such that A = ((m +

3)n + 1)=3m is integer. Prove that A is an odd integer number.

Solution: Assume that A is an even integer number, i. e., (m +

3)n + 1 = 6km. Then m is an even integer number. Moreover
mj3n +1, which shows that m = 3t +2 and n is odd. Let m = 2m1,
where   1 and m1 is odd. Then 2 j3n + 1 and therefore
n+1
 2
. Since m1j3 + 1, it follows that m1ja + 3, where a = 3 . It is
n 2 2
well-known that in this case m1 = 6t1 +1. Since m = 2 (6t1 +1) has
the form 3t +2 and 1  2 we see that  = 1. Then m = 12t1 +2
and from (m + 3)n + 1 = 6km it follows that 4j5n + 1, which is
impossible.

Problem 6. The sides and the diagonals of a regular n-gon X are

coloured in k colours so that:

4
 (i) for each colour a and any two vertices A and B of X , the
segment AB is coloured in colour a or there is a vertex C such
that AC and BC are coloured in colour a
(ii) the sides of any triangle with vertices among the vertices of X
are coloured in at most two colours.
Prove that k 2.

Solution: Assume that the colouring involves at least three dier-

ent colours a, b, c. We shall construct an innite subset of vertices
of X , which will imply a contradiction.
Let Z 2 X and A1 is a vertex such that the colour of A1Z
is a. From (i) it follows that there is a vertex B1, such that the
colour of B1Z and B1A1 is b. Analogously there is a vertex C1
such that the colour of C1Z and C1B1 is c. Considering the tri-
angles C1A1Z and C1A1B1 we see (using condition (ii)) that the
colour of C1A1 is c. Let A2 2 X be such that the colour of A2C1
and A2Z is a. It is easily seen that A2 6= A1 and the colour
A2A1 and A2B1 is a. Now we shall proceed by induction. Let the
vertices A2 B2 C2 : : :  Ak;1 Bk;1 Ck;1 be such that the colour of
AiAj  AiBj  AiCj is a, the colour of BiAj  BiBj  BiCj is b and the
colour CiAj  CiBj  CiCj is c (2 j < i < k).
Take a vertex Ak such that the colour of Ak Ck;1 and Ak Z is a.
Then considering the triangles ZAk Bj and Ck;1 AkBj (j < k), ZAkCj
and Bj+1 Ak Cj (j + 1 < k), Ak Aj Bj and Ak Aj Bj we see that the
colour of Ak Bj  Ak Cj and AkCj is a. The vertices Bk and Ck are
constructed in a similar way.

5
XLVIII National
Mathematics Olympiad:
3rd round, 17{18 April 1999
Problem 1. Find all triples (x y z) of natural numbers such that
y is a prime number, y and 3 do not divide z, and x3 ; y3 = z2.
Nikolay Nikolov

Solution: Since (x ; y)((x ; y)2 + 3xy)) = x3 ; y3 = z2, it follows

from the conditions of the problem that x ; y and (x ; y)2 +3xy are
relatively prime. Therefore x ; y = u2 and x2 + xy + y2 = v2. Thus
3y2 = (2v ; 2x ; y)(2v + 2x + y) and since y is a prime number,
there are three cases to consider:
1. 2v ; 2x ; y = y 2v + 2x + y = 3y. Now x = 0, which is
impossible.
2. 2v ; 2x ; y = 1 2v + 2x + y = 3y2. Now 3y2 ; 1 = 2(2x +
y) = 2(2u2 + 3y) and it follows that 3 divides u2 + 1, which is
impossible.
1
 3. 2v ; 2x ; y = 3 2v + 2x + y = y2. Now y2 ; 3 = 2(2x + y) =
2(2u2 +3y) and it follows that (y ; 3)2 ; (2u)2 = 12. Therefore
y = 7 u = 1 x = 8 z = 13. Direct verication shows that
(8 7 13) is a solution of the problem.

Problem 2. A convex quadrilateral of area S is inscribed in a

circle whose centre is a point interior to the quadrilateral. Prove
that the area of the quadrilateral whose vertices are the projections
of the point of intersection of the diagonals on the sides does not
exceed S2 . Christo Lesov

Solution: Let ABCD be a quadrilateral inscribed in a circle with

centre O and radius R. Denote by E the point of intersection of
AC and BD. Denote further by M , N , P , Q and F the projections
of E on AB , BC , CD, DA and MN , respectively. We know that
MN = BE sin 6 ABC = BE2R AC . Also,

EF = EM sin 6 EMN = AE  BEAB

sin 6 AEB sin 6 CBE:

Since BE sin 6 CBE = CE sin 6 BCE = CE AB 2 R and AE  CE =

R2 ; OE 2 , it follows that EF = R ;2ROE sin 6 AEB . Therefore
2 2

S4MEN = MN2 EF = AC  BE sin 8AEB (R2 ; OE 2) :

6
R2

2
Similar equalities hold for S4NEP  S4PEQ and S4QEM . By combin-
ing the above we obtain
SMNPQ = AC  BD sin 4AEB (R2 ; OE 2 )  SABCD 
6
R2 2
Q. E. D.

Problem 3. In a competition 8 judges marked the contestants by

yes or no. It is known that for any two contestants, two judges gave
both a yes two judges gave the rst one a yes and the second one
a no two judges gave the rst one a no and the second one a yes,
and nally, two judges gave both a no. What is the greatest possible
number of contestants? Emil Kolev

Solution: Denote the number of contestants by n. Consider a

table with 8 rows and n columns such that the cell in the ith row and
j th column contains 0 (1) if the ith judge gave the j th contestant
a no (a yes). The conditions of the problem now imply that the
table formed by any two columns contains among its rows each of
the pairs 00, 01, 10 and 11 twice. We shall prove that 8 columns
having this property do not exist. Assume the opposite. It is easily
seen that if in any column all 0s are replaced by 1s and vice versa,
the above property is retained. Therefore without loss of generality
suppose that the rst row consists of 0s. Denote the number of
0s in the ith row by ai. It is clear that the total number  of 0s is
8  4 = 32. Further, the number of occurrences
 !
of 00 is 2  2 = 56.
8

On the other hand the same number is

X
8
ai . Since a = 8, it
i=1 2 1

3
  !
follows that
X
8
ai = 24. It is easy to prove now that
X
8
ai
 30.
i=2 ! i =2 2
Therefore 56 =
X8
ai  58, which is false.
i=1 2
0 0 0 0 0 0 0
0 1 1 1 1 0 0
The diagram on the right shows 0 1 1 0 0 1 1
0 0 0 1 1 1 1
that it is possible to have exactly 1 0 1 0 1 0 1
7 contestants: 1 0 1 1 0 1 0
1 1 0 0 1 1 0
1 1 0 1 0 0 1

Problem 4. Find all pairs (x y) of integer numbers such that x3 =

y3 + 2y2 + 1. Nikolay Nikolov and Emil Kolev

Solution: It is obvious that x > y. On the other hand x < y +

1 () (y + 1)3 > y3 + 2y2 + 1 () y(y + 3) > 0. Therefore if
y > 0 or y < ;3 the problem has no solution. Direct verication
yields all pairs (x y) which satisfy the equality x3 = y3 + 2y2 + 1,
namely (;2 ;3), (1 ;2) and (1 0).

Problem 5. Let B1 and C1 be points on the sides AC and AB

of 4ABC . The straight lines BB1 and CC1 intersect at point D.
Prove that the quadrilateral AB1DC1 is circumscribed if and only if
the incircles of 4ABD and 4ACD are tangent.
Rumen Kozarev and Nikolay Nikolov

4
Solution: Note that the incircles of 4ABD and 4ACD are tan-
gent if and only if AB + AD ; BD = AC + AD ; CD, so AB + CD =
AC + BD.
Suppose AB1DC1 is circumscribed and the incircle touches AB1,
B1D, DC1 , C1A in the points M N P Q, respectively. Therefore
AB + CD = AQ + BQ + CP ; DP = AM + BN + CM ; DN =
AC + BD.
Conversely, let the incircles of 4ABD and 4ACD be tangent.
Denote the point of intersection of the tangent through C (dierent
from CA) with the incircle of 4ABB1 by D0 . It follows from the
above that BD0 ; CD0 = AB ; AC = BD ; CD, so DD0 = jCD ;
CD0j. Therefore D0  D, which completes the proof.

Problem 6. Each interior point of an equilateral triangle of pside 1

lies in one of six circles of the same radius r. Prove that r  103 .
Nikolay Nikolov and Emil Kolev

Solution: Divide each side of the triangle into ve equal parts
and draw lines parallel to the sides through these points. Thus the
triangle is divided into 25 equilateral triangles of side 51 . The total
number of vertices is 21 > 6  3. Therefore there exist 4 points which
are interior
p to one and the same circle. It is easy to see now that
r  103 , which solves the problem.

5
XLVIII National
Mathematics Olympiad: 4th
round, 18{19 May 1999
Problem 1. The faces of an orthogonal parallelepiped whose di-
mensions are natural numbers are painted green. The parallelepiped
is partitioned into unit cubes by planes parallel to its faces. Find
the dimensions of the parallelepiped if the number of cubes having
no green face is one third of the total number of cubes.
Sava Grozdev

Solution: Let x y z be the dimensions of the parallelepiped.

It follows from the conditions of the problem that x  3 and (x ;
2)(y ; 2)(z ; 2) = xyz . Since x ; 2 y ; 2 z ; 2 , when x  7,
3 x y z
we obtain that ( x ; 2)( y ; 2)(z ; 2) 5 1
 ( )3 > . Therefore x 6
xyz 7 3
and thus x = 3, x = 4, x = 5 or x = 6.
1. If x = 3, then (y ; 2)(z ; 2) = yz, which is impossible.
1
 2. If x = 4, then 2(y ; 2)(z ; 2) = 4yz 3 , so (y ; 6)(z ; 6) = 24.
In this case the only solutions are (4 7 30), (4 8 18), (4 9 14)
and (4 10 12).
3. If x = 5, then 3(y ; 2)(z ; 2) = 5yz
3 , so (2y ; 9)(2z ; 9) = 45.
Therefore the solutions are (5 5 27), (5 6 12) and (5 7 9).
4. If x = 6, then 4(y ; 2)(z ; 2) = 2yz, so (y ; 4)(z ; 4) = 8.
Thus there is an unique solution (6 6 8).
Answer: The problem has 8 solutions|(4 7 30), (4 8 18), (4 9 14),
(4 10 12), (5 5 27), (5 6 12), (5 7 9) and (6 6 8).

Problem 2. Let fang1n=1 be a sequence of integer numbers such

that
(n ; 1)an+1 = (n + 1)an ; 2(n ; 1)
for any n  1. If 2000 divides a1999, nd the smallest n  2 such
that 2000 divides an. Oleg Mushkarov, Nikolai Nikolov

Solution: It is obvious that a1 = 0 and an+1 = nn ;+ 11 an ; 2 when

n  2. Therefore the sequence is uniquely determined by its second
term. Furthermore the sequence an = (n ; 1)(cn + 2) (where c =
a2 ; 1 is an arbitrary real number) satises the equality from the
2
conditions of the problem. We conclude now that all sequences which
satisfy this equality are of the kind given above. Since all terms are
integer numbers and 2000 divides a1999, it is easy to see that c is
integer and c = 1000m + 2. Therefore 2000 divides an if and only if
2
1000 divides (n ; 1)(n +1). Thus n = 2k +1 and k(k +1) is divisible
by 250 = 53  2. Since k and k + 1 are relatively prime, we get that
the smallest n  2 equals 2  124 + 1 = 249.

Problem 3. The vertices of pa triangle have integer coordinates

and one of its sides is of length n, where n is a square-free natural
number. Prove that the ratio of the circumradius and the inradius
is an irrational number. Oleg Mushkarov, Nikolai Nikolov

Solution: Suppose that Rr = q, where q is a rational number.

Without lossp of generality assume that one of the ends of the side
of length n is at the origin of the coordinate system. Let the
remaining two vertices have coordinates (x y) and (z t), where p x y z
and
p t are integers.
p The sides of our triangle have lengths a = A b =
B and c = C , where n = A = x + y  B = z + t and C =
2 2 2 2

(x ; z)2 +(y ; t)2. It follows from the conditions of the problem that
q = Rr = abc
4S S p = abc(a + b + c) 
8S 2
where S is the p p Since S is2 rational (prove it!), it
p area ofpthe triangle.
follows that p ABC ( pA + B + C ) =p8S q is a rational number.
Therefore Ap BC + B AC = 8S 2q ; C AB and after squaring we
obtain that AB is a rational number. Thus AB is a perfect square.
By analogy both BC and CA are perfect squares. Let AB = E 2,
BC = F 2 and CA = G2, where E , F and G are integer. Write A B
and C in the following form: A = a1a22 B = b1b22 and C = c1c22,
where a1 b1 and c1 are square-free integers. So a1b1(a2b2)2 = m2
and therefore a1b1 is a perfect square, whence a1 = b1. By analogy

3
a1 = c1. Thus A = ma21 B = mb21 C = mc21, where m is square-free.
It follows from ma21 = n that m = n a1 = 1 and we obtain
x2 + y 2 = n
z2 + t2 = nb21
(x ; z)2 + (y ; t)2 = nc21
Since both b1 and c1 are integer, it follows from the Triangle Inequal-
ity that 1+b1 > c1 and 1+c1 > b1, whence b1 = c1. It is easy to deter-
mine now that x2 + y2 = 2(xz + yt) and consequently 2(xz + yt) = n.
Let 2(xt ; yz) = k. Then n2 + k2 = 4(x2 + y2)(z2 + t2) = 4n2b21,
so k2 = n2(4b21 ; 1). Therefore 4b21 ; 1 is a perfect square, which is
impossible.

Problem 4. Find the number of all natural numbers n 4 n

1023, such that their binary representations do not contain three
consecutive equal digits. Emil Kolev

Solution: Denote by an n  3, the number of sequences of zeroes

and ones of length n which begin with 1 and do not contain three
consecutive equal digits. Also, for any a b 2 f0 1g denote by xnab
the number of sequences of zeroes and ones of length n which begin
with 1 and do not contain three consecutive equal digits, such that
the last two terms are respectively a and b. It is easy to see that for
n5
xn00 = xn10;1  xn01 = xn00;1 + xn10;1 xn10 = xn11;1 + xn01;1 xn11 = xn01;1:
Adding up the above equalities, we obtain an = an;1 + xn10;1 + xn01;1 =
an;1 + an;2 . Since a3 = 3 and a4 = 5, it follows that a5 = 8, a6 = 13,
a7 = 21, a8 = 34, a9 = 55 and a10 = 89.
4
 Since the required number is equal to a3 + a4 + : : : + a10, it follows
from the above that the answer is 228.

Problem 5. The vertices A, B and C of an acute triangle ABC

lie on the sides B1C1 C1A1 and A1B1 of 4A1B1C1 and 6 ABC =
6 A1B1C1, 6 BCA = 6 B1C1A1, 6 CAB = 6 C1A1B1. Prove that the
orthocentres of 4ABC and 4A1B1C1 are equaly remote from the
circumcentre of 4ABC . Nikolai Nikolov

Solution: Denote by H the orthocentre of 4ABC . Since 6 CHB =

180 ; 6 CAB = 180 ; 6 C1A1B1, we have that A1 lies on the
circumcircle k1 of 4BHC . Similarly, B1 and C1 lie on circum-
circles k2 and k3 of 4CHA and 4AHB . Therefore 6 B1HC1 =
6 B1HA + 6 C1HA = 6 B1CA + 6 C1BA = 26 B1A1C1 and likewise
6 C1HA1 = 26 C1B1A1 and 6 A1HB1 = 26 A1C1B1, so H is the cir-
cumcentre of 4A1B1C1.
Let us draw straight lines passing through the vertices of 4ABC
and parallel to the corresponding sides and denote their points of in-
tersection by A0, B0 and C0. Since 6 A0B0C0 = 6 A1B1C1 6 B0C0A0 =
6 B1C1A1 and 6 C0A0B0 = 6 C1B1A1, it follows from the above that
the segments A0H B0H and C0H are of equal length and are diam-
eters of k1 k2 and k3. It is clear now that there exists a composition
of a rotation and a homothecy, both centred at H , such that the
image of 4A1B1C1 is 4A0B0C0. Therefore the image of ortho-
centre H1 of 4A1B1C1 is the orthocentre H0 of 4A0B0C0. Thus
6 HH1 H0 = 6 HA1A0 = 90, and to solve the problem we have to
show that the circumcentre O of 4ABC is the midpoint of HH0.
Indeed, the image of 4ABC by a homothecy centred at the cen-

5
 ;! of 4ABC
troid ;! and with ;! ;2 is 4A0B0C;!
coecient ;! 0 . Therefore
;!
MH0 = ;2 MH and since MH = ;2 MO, we obtain OH0 = ; OH .

Problem 6. Prove that the equation

x3 + y3 + z3 + t3 = 1999
has innitely many integer solutions. Grigor Grigorov

Solution: Since 103 + 103 + (;1)3 + 03 = 1999, we are looking for

solutions of the form x = 10 ; k y = 10 + k z = ;1 ; l t = l, where
k and l are integer. After simple calculations we obtain that our
equation is equivalent to l(l +1) = 20k2, whence (2l +1)2 ; 80k2 = 1.
The latter is Pell's equality. Since l = 4 k = 1 is a psolution, all
solutions are of the form (ln kn ), where 2ln + 1 + kn 80 = (9 +
p
80)n  n = 1 2 : : :. Therefore the original equation has innitely
many integer solutions.

6
 Union of Bulgarian Mathematicians

Sava Grozdev Emil Kolev

BULGARIAN
MATHEMATICAL COMPETITIONS
2000

Soa, 2000
 Winter Mathematical Competition
Russe, 4-6 February 2000
Problem 8.1 Given the inequality (n2 ; 1)x < ;3n3 ; 4n2 + n + 2
where n is an integer.
a) Factorize the expression ;3n3 ; 4n2 + n + 2:
b) Find all n for which the inequality holds true for any positive
number x:
Solution: a) ;3n3 ; 4n2 + n + 2 = (n + 1)2(2 ; 3n):
b) Since 0:x < 0 is not true for any x it follows that n 6= ;1: If
n > ;1 then the inequality is equivalent to (n;1)x < (n+1)(2;3n):
If n = 1 then 0:x < ;2 which is not true for any x: Let n 6= 1:
If n ; 1 > 0 then (n + n1)(2 ; 3n) < 0 and the inequality is not
;1
true for any positive x: If n ; 1 < 0 then the inequality becomes
x > (n + n1)(2 ; 3n) and n = 0 is the only solution. If n < ;1 then
;1
(n + 1)(2 ; 3n) > 0 and there exists x > 0 which is not a solution.
n;1
Therefore the inequality has no solution when n < ;1:
Problem 8.2 In an isosceles 4ABC (AC = BC ) the points A1 B1
and C1 are midpoints of BC AC and AB respectively. Points A2
and B2 are symmetric points of A1 and B1 with respect to AB: Let
M be the intersecting point of CA2 and A1C1 and let N be the
intersecting point of CB2 and B1C1: The intersecting point of AN
and BM is denoted by P: Prove that AP = BP:

2
Solution: Since CC1 k A1A2 and CC1 = A1A2 we have that
CC1A2A1 is a parallelogram. Thus, A1M = C1M: But A1B1C1B is
also a parallelogram and therefore the intersecting point of BM and
AC is B1: Hence P lies on the median BB1: Analogously P lies on
the median AA1: In the isosceles 4ABC the medians AA1 and BB1
are of the same length. Therefore AP = 32 AA1 = 23 BB1 = BP:
Problem 8.3 Find all pairs of prime numbers p and q such that
p2 + 3pq + q2 is:
a) a perfect square
b) a power of 5.
Solution: a) Let p2 + 3pq + q2 = r2 where p and q are prime
numbers. If p 6= 3 q 6= 3 then p2 + 3pq + q2  2(mod 3) and
r2  2(mod 3) a contradiction. Without loss of generality p = 3
and we get that q2 + 9q + 9 = r2 and 4q2 + 36q + 36 = (2r)2:
Therefore (2q ; 2r +9)(2q +2r +9) = 45: We may assume that r > 0
and so 2q +2r +9 = 15 or 2q +2r +9 = 45: In the rst case q + r = 3
which is impossible and in the second case solving the system
q+r = 18
2q ; 2r + 9 = 1
we nd q = 7: Because of the symmetry the only solutions are p =
3 q = 7 and p = 7 q = 3:
b) Let p2 + 3pq + q2 = 5n  where n is a natural number. Since
p  2 q  2 we have p2 + 3pq + q2  20 and so n  2: It follows
now that 25=(p2 + 3pq + q2) and 5=(p2 + 3pq + q2) = (p ; q)2 + 5pq:
Thus, 5=(p ; q)2 and 25=(p ; q)2: Therefore 25=5pq showing that
p = 5 or q = 5: But if p = 5 then q = 5 (and vice versa). We
obtain p2 + 3pq + q2 = 125 = 53: The only solution of the problem
is p = q = 5:
3
Problem 9a.1 Given the equation jx ;1 2j = jx ;152aj  where a is
a parameter.
a) Solve the equation.
b) If a is the square of a prime number prove that the equation
has a solution which is a compose integer.
Solution: After squaring and simple calculations we obtain (26a ;
1)x = (26a ; 1)(26a + 1):
a) If a = 261 then every x 6= 2 is a solution. If a 6= 1 then the
26
only solution is x = 26a + 1:
b) Let a = p2 where p is a prime number. If p = 3 then x = 235
is not a prime. If p 6= 3 then p = 3k  1 and x = 26(3k  1)2 + 1 =
3A + 27 which is divisible by 3:
Problem 9a.2 The quadrilateral ABCD is inscribed in a circle with
diameter BD: Let M be the symmetric point of A with respect to
BD and let N be the intersecting point of the straight lines AM and
BD: The line passing through N which is parallel to AC intersects
CD and BC in P and Q respectively. Prove that the points P C Q
and M are vertices of a rectangle.
Solution: It follows from the condition of the problem that _M lies
on the circumcircle of ABCD: Since <) MAC =<) MBC = MC 2 and
<) MNQ =<) MAC i.e. <) MNQ =<) MBC , we get that the points
M N B and Q lie on a circle. Since <) MNB = 90 we conclude that
<) BQM = 90 : Also, since 4BDC is a right angle triangle we _ have
that MQ k PC: From the other hand <) MDC =<) MAC = 2  and MC
therefore <) MDC =<) MNQ so the points N P M and D lie on a
circle. Thus, <) MPD =<) MND = 90 and MP k CQ: Therefore
P C Q and M are vertices of a rectangle.

4
Problem 9a.3 See Problem 8.3.
Problem 9b.1 Given the system:
1 +x = a;1
x+y
x
x + y = a ; 2
where a is a real parameter.
a) Solve the system if a = 0:
b) Find all values of a such that the system has an unique solu-
tion.
c) If a 2 (2 3) and (x y) is a solution of the system nd all values
of a such that the expression x + y takes its minimal value.
y x
Solution: It follows easily that x and x +1 y are roots of the equation
t2 ; (a ; 1)t + a ; 2 = 0: There are two cases to be considered:
x = 1 x = 1
(1) 1 = a ; 2 () 3 ; a when a 6= 2
x+y y = a ; 2
and
x = a;2 x = a;2
(2) 1 = 1 ()
x+y y = 3 ; a:

a) When a = 0 we have (x y) = (1 ; 32 ) or (x y) = (;2 3):

5
 b) The sytem has an unique solution (0 1) when a = 2 and (1 0)
when a = 3:
c) If (x y) is a solution and a 2 (2 3) then xy and xy are positive.
Further xy + xy  2 and equality occures when xy = xy : It follows
from (1) and (2) that xy = 3a ; ; 2 : Using the equality a ; 2 = 3 ; a
a 3;a a;2
we nd the only value of a = 2 : 5

Problem 9b.2 Given an acute 4ABC: The bisector of <) ACB

intersects AB at point L: The feet of the perpendiculars from L to
AC and BC are denoted by M and N respectively. Let P be the
intersecting point of AN and BM: Prove that CP ? AB:
Solution: Let l be the line through C which is parallel to AB: Let
F and E be respectively the intersecting points of AN and BM with
l: The intersecting point of CP and AB is denoted by D: We obtain
AD = PD = BD  so AD = CF : From the other hand AM =
CF PC CE BD CE CM
AB and BN = AB : But CM = CN and we get AM = CF :
CE CN CF BN CE
AD AM
Therefore BD = BN  which implies

(1) AM = BN
AD BD
Further if CH ? AB (H 2 AB ) then 4ALM  4AHC and so
AL = AM : In the same manner BL = BN : But CL is a bisec-
AC AH BC BH
tor and therefore AL = BL  so AM = BN : The last equation
AC BC AH BH
combined with (1) gives D  H which implies CP ? AB:
Problem 9b.3 Prove that the digit of the hundreds of 21999 +22000 +
6
22001 is even.
Solution: Write the number 21999 +22000 +22001 in the form 21999(1+
2 + 4) = 7:29:21990 = 7:29:210:21980 = 7:29:210:(220)99: Since 29 =
512 210 = 1024 and 220 = (210)2 we have that the last two digits of
220 coincide with the last two digits of 242  so the last two digits of
220 are 76: Moreover the last two digits of 76:76 are also 76: Therefore
the last two digits of the given number are the last two digits of the
product 7:12:24:76 which are 1 and 6: Since 21999 + 22000 + 22001 is
divisible by 8 and it ends by 16 the digit of the hundreds is even.
Problem 10.1 Find all values of the real parameter a such that the
nonnegative solutions of the equation (2a ; 1) sin x +(2 ; a) sin 2x =
sin 3x form an innite arithmetic progression.
Solution: Since sin 2x = 2 sin x cos x and sin 3x = sin x(4 cos2 x;1)
we may write the equation in the form
sin x (2 cos2 x ; (2 ; a) cos x ; a) = 0: Thus sin x = 0 cos x = 1 or
cos x = ; a2 : The nonnegative solutions of the equations sin x = 0
and cos x = 1 are x = k and x = 2k k = 0 1 2a : : : respectively.
Let jaj > 2: In this case the equation cos x = ; 2 has no solution
and therefore the nonnegative solutions of the initial equation are
0  2 : : : which form an arithmetic progression. Let now jaj 
2 and let x0 be the only solution of the equation cos x = ; a2 in
the interval 0 ]: In this case the nonnegative solutions of the last
equation are x = x0 + 2k and x = 2 ; x0 + 2k: It is clear now
that the nonnegative  solutions form an arithmetic progression only
when x0 = 0 x0 = 2 and x0 =  so giving a = ;2 a = 0 and a = 2:
The values of a are a = 0 and jaj > 2:
Problem 10.2 Let O I and H be respectively the circumcenter,
incenter and orthocenter for an acute nonequilateral 4ABC: Prove
7
that if the circumcircle of 4OIH passes through one of the vertices
of 4ABC then it passes through another vertex of 4ABC:
Solution: Assume that O I H and C lie on a circle. It is well known
that CI is the bisector of <) HCO: Thus <) IHO =<) ICO =<) ICH
=<) HOI and it follows from 4IHO that IH = IO = t:
We shall prove that if <) BAC 6= 60 then O I H and A lie on
a circle. Denote by M and N the projection points of I respec-
tively on AO and AH: Let O1 and O2 be such that IO1 = IO2 = t
and O1 lies between A and M and M lies between O1 and O2:
Analogously let H1 and H2 be such that IH1 = IH2 = t and H1
lies between A and N and N lies between H1 and H2: If O 
O1 H  H1 or O  O2 H  H2 then 4AIO  4AIH and there-
fore AO = AH: But AH = 2AO cos <) BAC and so <) BAC = 60 :
If O  O1 H  H2 or O  O2 H  H1  then it follows from
<) IO1O2 =<) IO2O1 =<) IH1H2 =<) IH2H1 that AOIH is inscribed.
Suppose now that A and B do not lie on the circumcircle of
4OIH: In this case <) BAC =<) ABC = 60 and therefore 4ABC
is equilateral which is a contradiction.
Problem 10.3 In each of the cells of a 3 3 table is written a real
number. The element in the i;th row and j ;th column equals to
the modulus of the dierence of the sum of the elements from the
i;th row and the sum of the elements from the j ;th column. Prove
that every element of the table equals either to the sum or to the
dierence of two other elements of the table.
Solution: Let p1  p2 p3 and q1 q2 q3 be the sum of the elements in
the rst, second and third row and in the rst, second and third
column respectively. It is clear that p1 + p2 + p3 = q1 + q2 + q3:
Therefore the element in the rst row and the rst column equals to
jp1 ; q1j: From the other hand jp1 ; q1j = jp2 + p3 ; q2 ; q3j which
implies jp1 ; q1j = 1jp2 ; q2j + 2jp3 ; q3j where 1 2 2 +1 ;1: Since
8
jp1 ; q1j  0 it is clear that 1 = 2 = ;1 is impossible. Therefore
jp1 ; q1j is either the sum or the dierence of jp2 ; q2j and jp3 ; q3j:
By analogy every element of the table is the sum or the dierence
of two other elements.
Problem 11.1. Prove that for every positive number a the se-
quence fxng1 n=1  such that x1 = 1 x2 = a xn+2 = xn+1 xn  n  1
q
3 2

is convergent and nd its limit.

Solution: It follows by induction that the terms of the sequence
fxng1n=1 can be expressed as xn = an  where f n g1n=1 is a siquence
dened by 1 = 0 2 = 1 n+2 = 2 n+13 + n  n  1: Thus, n+2 ;
n
n+1 = ; 31 ( n+1 ; n ) and therefore n+2 ; n+1 = ; 13 ( 2 ;
 

n k
1) = ; 31 : Adding the equalities k+2 ; k+1 = ; 13 for k =
   

0 1 : : :  n after simple calculations we obtain n+2 ; 0 = ; 13 +

 0

n+1
1 ; ; 13
 

1 1  
1 n 3 (1 ; ; 1 n+1 ): Since
 

;3 + : : : + ;3 =   = 4 3
1 + 13
0 = 0 and nlim
 

; 1 n = 0 it follows that lim = 3 : This shows

!1 3 n!1 n 4
that the sequence fxng1 n=1 is convergent and nlim x = a 4:
3
!1 n

Problem 11.2 Given a convex quadrilateral ABCD where M is

the intersecting point of its diagonals. It is known that DB =
3DM AM = MC:
a) Express BC and CD by the sides of 4ABD:
b) Prove that if 2 <) ADB ; <) ABD = 180  then <) DBC =
2 <) BDC:
Solution: a) Denote AB = c BC = p CD = q DA = b and
9
DB = a: It follows from the condition of the problem that DM =
1 a MB = 2 a: The formula for the median in a triangle applied for
43 ABC and34ACD gives
(1) 16 a2 = 2p2 + 2c2 ; 4AM 2 4 a2 = 2q2 + 2b2 ; 4AM 2
9 9
Let <) AMB = : From the Law of Cosines for 4AMB and 4AMD
we obtain
c2 = AM 2 + 49a ; 34 AM:a cos b2 = AM 2 + a9 + 23 AM:a cos
2 2

Thus 2b2 + c2 = 3AM 2 + 32 a2 and a substitution in (1) gives

p2 = 4a + 129b ; 3c q2 = 3b + 69c ; 2a
2 2 2 2 2 2
(2)

b) It follows from the condition of the problem that <) ADB >
90 and so c > a: Let D1 be a point on AB such that D1B = DB:
Further <) AD1B =<) ABD + 180 ; 2<) ABD = 180 + 2<) ABD =
<) ADB: Therefore 4AD1D  4ADB and thus b2 = c(c ; a): A
substitution in (2) implies p = 3c ;3 2a and q2 = 9c ; 3ac ; 2a2 :
2

9
Hence q2 = p(p ; a): Let B1 be a point on BC (B lies between C
and B1), such that BB1 = BD = a: Since qp = p +q a it follows that
4DBC  4B1DC: Therefore <) DBC =<) BB1D+ <) BDB1 =
2 <) BB1D = 2 <) BDC:
Problem 11.3 See problem 10.3.

10
 Spring Mathematical Tournament
Jambol, 24-26 March 2000
Problem 8.1. Let f (x) be a linear function such that f (0) = ;5
and f (f (0)) = ;15: Find all values of m for which the set of the
solutions of the inequality f (x):f (m ; x) > 0 is an interval of length
2:
Solution: Let f (x) = ax + b: It follows from f (0) = ;5 that b = ;5
and from f (f (0)) = ;15 that a = 2: Therefore the function is f (x) =
2x ; 5: Consider the inequality (2x ; 5)(2(m ; x) ; 5) > 0 ()
(2x ; 5)(2m ; 5 ; 2x) > 0: The solution of the last inequality is an
interval with end points 25 and 2m2; 5 : Therefore 25 ; 2m2; 5 = 2
so j5 ; mj = 2: Finally, we obtain that m = 3 and m = 7:
Problem 8.2 Given an isosceles right angle triangle ABC with
<) ACB = 90 : Point P lies on BC M is the midpoint of AB and
let L and N be points from the segment AP such that CN ? AP
and AL = CN:
a) Find <) LMN 
b) If the area of 4ABC is 4 times greater than the area of
4LMN nd <) CAP:
Solution: a) Let <) CAP = : We nd <) ACN = 90 ; and
<) MCN =<) ACN ; 45 = 45 ; =<) LAM: Since AM = CM and
AL = CN it follows that 4AML  = 4CMN: Therefore <) AML =
<) CMN and so <) LMN = 90 ; <) AML+ <) CMN = 90 :
b) It follows from 4AML 
= 4CMN that LM = MN and since

11
<) LMN = 90 we get SLMN = MN Also SABC = AC : Now
2 2

2
: 2
4MN 2 = AC 2 applies that MN = AC : Denote by Q the midpoint
2 2 2
AC
of AC: Thus QM = QN = 2 = MN: Therefore 4QMN is
equilateral and <) QNM = 60  which implies <) QNA = 60 ; 45 =
15 : But AQ = QN and so <) CAP =<) QNA = 15 :
Problem 8.3 There are 2000 white balls in a box. There are also
suciently many white, green and red balls. The following opera-
tions are allowed:
1) Replacement of two white balls with a green ball
2) Replacement of two red balls with a green ball
3) Replacement of two green balls with a white bal and a red
ball
4) Replacement of a white ball and a green ball with a red ball
5) Replacement of a green ball and a red ball with a white ball
a) After nitely many of the above opperations there are three
balls left in the box. Prove that at least one of them is a green ball.
b) Is it possible after nitely many opperations to have only one
ball left in the box?
Solution: Consider a box with x white, y green and z red balls.
Direct verication shows that after applying any of the allowed op-
erations the sum x + 2y + 3z does not change modulo 4. Since the
initial values are x = 2000 y = z = 0 we obtain that this sum is
congruent to 0 modulo 4.
a) There are 3 balls in the box and therefore x + y + z = 3:
Moreover x +2y +3z  0(mod4): If a green ball is not in the box then
y = 0 and so x + z = 3 and x +3z  x +3(3 ; x)  1 ; 2x  0(mod4)

12
which is impossible.
b) Suppose that there is only one ball left in the box. Therefore
x + y + z = 1 and x + 2y + 3z  0(mod4) which is impossible.
Problem 9a.1 Find all values of m such that the equation

1 + m ; 2m (jx ; mj ; m) = 0


x + m x ; m m2 ; x2
has exactly one nonnegative root.
Solution: The equation x +1 m + x ;mm ; m22;m x2 = 0 is equivalent
to (when x 6= m) to (m + 1)x = ;m(m + 1): When m = ;1 it has
innitely many roots { all numbers x 6= 1: If m 6= ;1 we obtain
x = ;m: The equation jx ; mj ; m = 0 has two roots: x = 0 and
x = 2m when m > 0 an unique root: x = 0 when m = 0 and has
no roots when m < 0:
Let m < 0 and m 6= ;1: In this case the equation has an unique
root x = ;m and it is nonnegative. Let m = 0: Then the equation
becomes x1 jxj = 0 which obviously has no roots. Let m > 0: The
equation has three roots x = ;m x = 0 x = 2m and two of them
are nonnegative.
Thus, the desired values of m are m < 0 and m 6= ;1:
Problem 9a.2 Given an acute-angled triangle ABC and let 

and  be its angles respectively to A B and C: For an arbitrary

interior point M denote by A1 B1 and C1 respectively the feet of
the perpendiculars from M to BC CA and AB: Find the locus of
M for which the triangle A1B1C1 is a right angle triangle.
Solution: Let M be a point such that <) A1C1B1 = 90 : Since the
quadrilateral AC1MB1 is inscribed we have <) B1C1M =<) B1AM =
x: Analogously <) A1C1M =<) A1BM = y: Denote z =<) AMB: It
follows now from the quadrilateral AMBC that x +  + y + (360 ;
13
z) = 360 : Since x + y =<) A1C1B1 = 90 we get z = 90 + :
Therefore <) AMB = 90 + : Thus, the locus of M such that
<) A1C1B1 = 90 is an arc G1 in the interior of 4ABC for which
the segment AB is seen under angle 90 + : Analogously one can
prove that the locus of M such that <) C1B1A1 = 90 is an arc G2
in the interior of 4ABC for which AC is seen under angle 90 +
:
Further, the locus of M such that <) B1A1C1 = 90 is an arc G3 in
the interior of 4ABC for which BC is seen under angle 90 + :
The desired locus is the union of the three arcs G1  G2  G3:
Problem 9a.3 See Problem 8.3
Problem 9b.1 The real numbers x and y are such that x2 +xy +y2 =
1: If F = x3y + xy3
a) prove that F  ;2
b) nd the greatest possible value of F:
Solution: a) It follows from the condition of the problem that (x +
y)2 = 1 + xy and therefore xy  ;1: From the other hand x2 + y2 =
1 ; xy: Thus, xy  1 so z = xy 2 ;1 1]: Hence F = xy(x2 + y2) =
z(1 ; z) and the inequality F  ;2 is equivalent to the inequality
z2 ; z ; 2  0: The latter one is true for any z 2 ;1 2]:
b) Assume that the greatest value of F exists and denote it by
A: This implies that the system
(1) xy(x2 + y2) = A
xy + (x2 + y2) = 1
has a solution. It is clear that t1 = xy and t2 = x2 + y2 are
roots of the quadratic equation t2 ; t + A = 0: But t2 ; t1 =
1 (x ; y)2 + 1 (x + y)2  0 and the equality is impossible since if
2 2
so then x = y = 0 which is a contradiction to x2 + xy + y2 = 1:
Therefore t2 > t1 and so D = 1 ; 4A > 0 i.e. A < 41 : The system
14
(1) is equivalent to x2 xy = t1 and the last is equivalent to
+ y = t2 
2

(x ; y) = t2 ; 2t1 The latter system has a solution i the in-

2

(x + y)2 = t2 + 2t1:
equalities t2 ; 2t1  0 and t2 + 2t1  0 hold true. It suces to prove
that when A < 14 the roots t1 < t2 of the equation g(t) = t2 ; t + A
satisfy the inequalities t2 ; 2t1  0 and t2 + 2t1  0: If A = 0
then t1 = 0 and t2 = 1 and the inequalities hold true. It is clear
that we can consider only the case A 2 (0 41 ): Now t1 > 0 t2 > 0
and the inequality t2 + 2t1  0 holds true. From the other hahd
t1 + t2 = 1 and the inequality t2 ; 2t1  0 is equivalent to t2  23 :
The same inequality is equivalent to t1  13 : For the roots t1 and
t2 we obtain t1  31 t2  23  which is equivalent to gg(( 32 ))  0
1

3p  0:

Therefore A  29 : Conversely, if A = 29 then x = y =  33 satisfy

the condition of the problem.
Problem 9b.2 A line l is drown through the orthocenter of an
acute-angled triangle ABC: Prove that the lines symmetric to l with
respect to the sides of the triangle intersect in a point.
Solution: Let the intersecting points of l with the sides AC and
BC be Q and P respectively and let the intersecting point of l with
the extention of AB be R: Without loss of generality A lies between
R and B: Denote the symmetric points of H with respect to AC BC
and AB by B1 A1 and C1 respectively. It is well known that A1 B1
and C1 lie on the circumcircle k of 4ABC: It is clear also that the
symmetric lines of l are B1Q A1P and C1R: Let B1Q \ A1P = S and
B1Q \ C1R = T: We obtain <) CB1Q+ <) CA1P =<) CHQ+ <) CHP
= 180 and therefore SA1CB1 is inscribed. Thus, S = k \B1Q: From

15
the other hand <) RC1A =<) RHA =<) AB1T and therefore B1ATC1
is inscribed. Hence, T = k \ B1Q: This implies that T  S which
shows that the three lines intersect in a point.
Problem 9b.3 See Problem 8.3.
Problem 10.1 Solve the equation px + p3 x + 7 = p4 x + 80:
Solution: It is clear that x  0: Note that x = 1 is a root of
the equation. We shall prove that there are no other p roots.
p Ras-
ingpthe p equation to the p forth
p power wepget x + 4( x) x + 7 +
2 3 3

6( x) ( x + 7) + 4( x)( x + 7)3 + ( 3 x + 7)4

2 3 2 3

= x + 80:
Let f ( x ) = 4( px)3 p3 x + 7 + 6(px)2(p3 x + 7)2
p p p
+4( x)( 3 x + 7)3 +( 3 x + 7)4: Obviously f (x) is an increasing func-
tion. If x > 1 then it follows from the inequalities x2 > x f (x) >
f (1) = 80 that x2 + f (x) > x + 80 a contradiction. If x < 1
then it follows from the inequalities x2 < x f (x) < f (1) = 80 that
x2 + f (x) < x + 80 a contradiction. This completes the prove.
Problem 10.2 The incircle of an isosceles 4ABC touches the legs
AC and BC at points M and _
N respectively. A tangent t is drawn
to the smaller of the arcs MN and let t intersects NC and MC at
points P and Q respectively. Let T be the intersecting point of the
lines AP and BQ:
a) Prove that T lies on the segment MN 
b) Prove that the sum of the areas of triangles ATQ and BTP
is the smallest possible when t is parallel to AB:
Solution: a) Let the incircle touches AB and PQ at points R and
S respectively. Let MN and SR intersect QB in points T1 and
T2 respectively. Since <) T1MQ =<) T1NC = 180 ; <) T1NB and
<) MT1D =<) BT1N it follows from the Law of Sines for 4MT1D and

16
 DT1 = BT1  so QT1 = MQ : By analogy QT2 = SQ :
4BT1N that MD BN BT BN 1 BT BR 2

It follows now from MQ = SQ and BN = BR that QT1 = QT2 

BT1 BT2
so T1 = T2: In the same manner one can prove that AP passes
through the intersecting point of MN and SR: This implies that
AP BQ MN and SR intersect in T:
b) We have SATQ + SBPT = SABQ + SABP ; 2SABT : Since 4ABC
is isosceles we get MN k AB and therefore SABT is constant. Thus,
SATQ + SBPT is minimal exactly when SABQ + SABP is minimal. The
latter sum equals to AQ:AB sin +2 BP:AB sin =
(AQ+BP )AB sin 
2  where is the angle to the base of the triangle.
Therefore it suces to nd the minimum of AQ + BP: It is easily
seen that AQ + BP = AM + BN + PQ and therefore we have to
nd when PQ is minimal. Let r be the inradius and O be the center
of the incircle of 4ABC: Using that PQ = r(cotg + cotg ) where
 =<) OQP =<) OPQ we obtain from ABPQ that 2 +2 +2 =
360 : Therefore  + = 180 ; : Thus PQ = cos( 2;r sin) + cos :
It is clear now that PQ is minimal exactly when cos( ; ) = 1 so
 = =) PQ k AB:
Problem 10.3 There are n  4 points in the plane such that the
distance between any two of them is an integer. Prove that at least
1 from the distances between them are divisible by 3.
6
Solution: We show rst that the assertion from the problem is true
for n = 4 i.e. for 4 points with integer distances between them at
least one distance is divisible by 3. Denote the points by A B C and
D (it is easy to be seen that WLOG <) BAD =<) BAC + <) CAD).
By the Law of Cosines for 4ABC 4ACD 4ABD we obtain
BC 2 = AB 2 + AC 2 ; 2:AB:AC cos
17
 CD2 = AD2 + AC 2 ; 2:AD:AC cos

BD2 = AB 2 + AD2 ; 2:AB:AD cos 

where =<) BAC
=<) CAD  =<) BAD = +
:
Suppose that all distances are integers not divisible by 3. There-
fore AB 2  AC 2  AD2  BC 2  CD2  BD2  1(mod 3) and
so 2:AB:AC cos  2:AD:AC cos
 2:AB:AD cos   1(mod 3):
Thus, 2:AB:AC cos :2:AD:AC cos
 4:AC 2:AB:AD: cos cos

 AC 2:AB:AD: cos cos

 1(mod 3):
Note that cos  cos
and cos  are rational numbers. Moreover,
if cos = pq  cos
= rs  where p q and r s are relatively prime
then p q r and s are not divisible by 3. Hence p2  q2  r2 
sp2  1(modp 23) and2 cos  = cos cos
; sin sin
= cos cos
;
q ; p s ; r : Therefore 2:AC 2:AB:AD: sin sin
is divisible
2 2

q s
by 3 and after multiplying 2:AB:AD cos   1(mod 3) by AC 2 we
obtain 2:AC 2:AB:AD: cos cos
 1(mod 3) a contradiction to
AC 2:AB:AD: cos cos
 1(mod 3): Therefore at least one of the
distances is divisible by 3.
Let n  4: Since there exist n4 sets with four elements each
 !

there exist at least n distances (counted more than once) divisible

 !

4
by 3. Each such distance is counted exactly n ;2 2 times and we
 !

 
n
= 16 n2 :
 !

get that the desired number is at least n;2

4


Problem 11.1 Let f (x) = xx2 ++74xx++143 :

2

18
 a) Find the greatest value of f (x)
f (x)

x 2
; 5x +
b) Find the greatest value of the function x2 + 5x + 20 10
!

:
Solution: a) We shall prove that the greatest value of f (x) equals
2. Since x2 + 7x + 14 > 0 8x we obtain f (x)  2 =) (x + 5)2  0
and an equality occurs only when x = ;5:
b) Let g(x) = xx2 ; 5x + 10 : Since x2 +5x +20 > 0 8x we obtain
2

+ 5x + 20
g(x)  3 =) (x + 5)2  0 and equality occurs only when x = ;5:
Since x2 ; 5x +10 > 0 8x we have that the function h(x) = g(x)f (x)
is correctly dened. Further, when f (x) > 0 i.e. x 2 ;3 ;1] then
h(x)  32 = 9: From the other hand g(x)  1 =) x  ;1 and so
h(x)  1 when f (x)  0: Therefore the greatest value of h(x) equals
9 and h(x) = 9 when x = ;5:
Problem 11.2 A point A1 is chosen on the side BC of a triangle
ABC such that the inradii of 4ABA1 and 4ACA1 are equal. De-
note the diameters of the incircles of 4ABA1 and 4ACA1 by da:
In the same manner dene db dc: If BC = a CA = b AB = c p =
a + b + c and h  h  h are the altitudes of the triangle ABC and d
a b c
2
is the diameter of the incircle of 4ABC prove that:
q
p(p ; a)
a) da + a d = ha
b) da + db + dc + p  ha + hb + hc :
Solution: a) Let O1 and O2 be the incenters of 4ABA1 and 4ACA1
and p1 and p2 be semiperimeters of the same triangles. We have
SABC = SABA1 + SACA1 = ra:p1 + ra:p2 = rc(p1 + p2) = rc :(p + CC1)
where ra = d2a : Therefore ra(p + AA1) = SABC : If P and Q are
the touching points of the incircles of 4ABA1 and 4ACA1 with

19
BC then the quadrilateral O1PQO2 is rectangle and so O1O2 =
PQ = PA1 + QA1 = p1 ; AB + p2 ; AC = p1 + p2 ; AC ; BC =
p + AA1 ; AC ; BC: It follows from the similarity of 4IO1O2 and
4IBC where I is the incenter of 4ABC that OBC 1 O2
= r ;r ra =)
AA1 + p ; AC ; BC = r ; rc : We obtain the system
BC r
ra(p + AA1) = S q
AA1 + AB ; p = r ; ra Thus, AA1 = p(p ; a) and ra =
c p pr p q

rp =
r p( p ; p ; a) = ha ; r p(p ; a)  which
a 2 a
q
p + p(p ; a)
implies a).
pp ; a pp ; b
b) By a) the inequality is equivalent to a + b +
pp ; c pp pp ; a pp ; b pp ; c
c p  d : Further p a + b + p c =
p;a p;b p;c
(p ;pb) + (p ; c) (p ;p a) + (p ; c) (pp; a) + (p ; b) 
+ +
p p ;pa + p p ;pb + p p ;pc =
2 p ; b p ; c 2 p ; ap p ; c 2 p ; a p ; b
p p
q = d:
(p ; a)(p ; b)(p ; c)
Problem 11.3 See problem 10.3.

20
 XLIX National Mathematical Olympiad
Third Round, 15-16 April 2000
Problem 1. Find all value of the real parameter a such that the
equation
9t ; 4a3t + 4 ; a2 = 0
has an unique root in the interval (0 1):
Solution: After the substitutions x = 3t and f (x) = x2 ; 4ax +
4 ; a2 the problem is equivalent to: nd all values of a such that
the equation f (x) = 0 has an unique root in the interval (1 3): It
is easily seen that if f (1) = 0 or f (3) = 0 then a = 1 a = ;5 or
a = ;13 are not solutions of the problem. Therefore the equation
f (x) = 0 has an unique root in the interval (1 3) when f (1):f (3) < 0
or Df = 0 2a 2 (1 3): It follows now that f (1):f (3) < 0 () (a ;
p < 0 () a 2 (;13 ;5) or 5a2 ; 4 = 0 2a 2 (1 3)
1)2(a +5)(a +13)
and so a = 2 5 5 :

Problem 2. In 4ABC CH (H 2 AB ) is altitude and CM and

CN (M N 2 AB ) are bisectors respectively of <) ACH and <) BCH:
The circumcenter of 4CMN coincides with the incenter of 4ABC:
Prove that S4ABC = AN:BM
2 :
Solution: Let I be the incenter of ABC: Denote by P Q and R the
common points of the incircle of ABC respectively with AB BC
and CA: It follows from IP = IQ = IR and IC = IM = IN that
4IMP 4INP 4ICQ and 4ICR are congruent. Let <) MIP =
<) NIP =<) QIC =<) CIR = : We have that <) MCN = 21 <) MIN
= : Since CM and CN are bisectors we get <) ACB = 2 <) MCN =
2: It follows now from the quadrilateral IQCR that <) QIR+
21
<) QCR = 180 () 4 = 180 ()  = 45 : Therefore 4ABC is
a right angle triangle which implies <) BCM =<) BMC =<) BAC +
<) ABC and <) ANC =<) ACN =<) ABC + <) BAC : Therefore BC =
2 2
BM and AC = AN and so SABC = 2 = AN:BMAC:BC
2 :
Problem 3. Let fang1n=1 be a sequence such that a1 = 43 a2 =
142 an+1 = 3an + an;1 for n  2: Prove that:
a) an and an+1 are relatively prime for all n
b) for every natural number m there exist innitely many natural
numbers n such that an ; 1 and an+1 ; 1 both are divisible by m:
Solution: a) Suppose that there exist natural numbers n and m > 1
such that m divides both an and an+1: It follows from an;1 = an+1 ;
3an that m divides an;1: By induction m divides both a1 and a2
which is impossible since a1 and a2 are relatively prime.
b) Consider the sequence fang dened by an;1 = an+1 ; 3an for
negative indices. Compute a0 = a2 ; 3a1 = 13 a;1 = a1 ; 3a0 =
4 a;2 = a0 ; 3a1 = 1 a;3 = a;1 ; 3a;2 = 1 and so on. We nd a
sequences fang1;1 of integers such that an+1 = 3an + an;1 for every
n: Since the pairs (p(mod m) q(mod m)) are nitely many we get
that there exist integers r and s > r such that ar  as(mod m)
and ar+1  as+1(mod m): It is clear now that ar+i  as+i (mod m)
for every i: Therefore the sequence fang1;1 is periodic. Since a;3 
a;2  1(mod m) there exist innitely many natural numbers n such
that both an ; 1 and an+1 ; 1 are divisible by m:
Problem 4. Given a convex quadrilateral ABCD such that <) BCD
=<) CDA: The bisector of <) ABC intersects the segment CD in
point E: Prove that <) AEB = 90 if and only if AB = AD + BC:

22
Solution: Let <) AEB = 90 : Since <) CEB < 90 there exists a
point F on the side AB such that <) BEF =<) BEC: Thus 4BCE 
=
4BFE which implies BF = BC and <) BFE =<) BCE: From the
other hand
() sin <AE = AF  AE = AD :
) AFE sin <) AEF sin <) ADE sin <) AED
Since <) AED =<) AEF and <) AFE + <) ADE = 180  we obtain
AF = AD and so AB = AD + BC:
Conversely, let AB = AD + BC: There exists a point F on
the segment AB such that AF = AD and BF = BC: Therefore
4BFE = 4BCE so <) BFE =<) BCE and <) BEF =<) BEC: It
follows from () and AF = AD that sin <) AED = sin <) AEF: Since
<) AED+ <) AEF < 180  we have <) AED =<) AEF and therefore
<) AEB = 90 :
Problem 5. Prove that for any two real numbers a and b there
exists a real number c 2 (0 1) such that
ac + b + c +1 1 > 241 :

Solution: Consider f (x) = ax + x +1 1 and let m and M be respec-

tively the minimum and the maximum value of the function f in the
interval 0 1]: Since f (0) = 1 f (1) = a + 12 and f 0(x) = a ; (x +1 1)2
there are four cases for the dierence M ; m:
1. a  41 : Thus f 0(x)  0 for x 2 0 1] and so M ;m = f (0);f (1) =
1 ; a  1:
2 4
23
2. a  1: Thus f 0(x)  0 for x 2 0 1] and so M ; m = f (1) ; f (0) =
a ; 12  12 :

3. 41  a  1: Thus d = p1a ; 1 2 0 1] f 0(x)  0 for x 2 0 d] and

f 0(x)  0 for x 2 d 1]:
3.1 14  a  12 : Since f (0)  f (1) we have M ; m = f (0) ;
p p
f (d) = (1 ; a)  (1 ; 22 )2:
2

3.2 21  a  1: Since f (1)  f (0) we have M ; m = f (1) ;

p p
f (d) = 1 (2 a ; 1)2  1 ( 2 ; 1)2:
2 2
In all four cases M ; m > 121  which implies M + b > 1 or m + b <
24
1 : The assertion of the problem follows now by continuity.
24
Problem 6. Find all sets S of four points in the plane such that: for
any two circles k1 and k2  having diameters with endpoints - points
from S there exists a point A 2 S \ k1 \ k2:
Solution: Let S = fA B C Dg: Consider the circles k1 and k2 with
diameters respectively AB and CD: It follows that at least one of
the angles ACB ADB CAD and CBD is right angle. Without loss
of generality <) ACB = 90 : Let k3 and k4 be circles with diameters
respectively AC and BD: Since <) ABC < 90  we obtain that the
common point of k3 and k4 is one of the points A C or D:
1. Let A 2 k3 \ k4: Now <) BAD = 90 and therefore D 2 l1
l1 ? AB A 2 l1: It is easily seen that <) BDC < 90 and since
<) BAC < 90  <) ABD < 90  it follows that the common point of
24
the circles with diameters AD and BC is the point C and <) ACD =
90 : Therefore D = l1 \ BC:
2. Let C 2 k3 \ k4 : Now <) BCD = 90 and therefore D lies on
the line AC: Since <) BDC < 90  <) BAC < 90 and <) ACD < 90 
we obtain that the common point of the circles with diameters AD
and BC is B and <) ABD = 90 : Therefore D = l2 \ AC where
l2 ? AB B 2 l2:
3. Let D 2 k3 \ k4: Now <) ADC = 90 and therefore D 2 k3:
Since <) BAC < 90 and <) ACD < 90  we obtain that the common
point of the circles with diameters AD and BC is B or D: In the rst
case <) ABD = 90 and thus D 2 l2 from the other hand l2 \ k3 = 
which is a contradiction. In the second case <) BDC = 90  and we
get that D is the orthogonal projection of C on AB:
In conclusion, all sets S satisfying the condition of the problem
are those consisting of the vertices of a right triangle and the foot of
the altitude to the hypothenusis.

25
 XLIX National Mathematical Olympiad
Fourth Round, 16-17 May 2000
Problem 1. In an orthogonal coordinate system xOy a set consisit-
ing of 2000 points Mi (xi yi) is called "good" if 0  xi  83 0 
yi  1 i = 1 2 : : :  2000 and xi 6= xj for i 6= j: Find all natural
numbers n with the following properties: : a) For any "good" set
some n of its points lie in a square of side length 1:
b) There exists a "good" set such that no n + 1 of its points lie
in a square of side length 1:
(A point on a side of a square lies in the square).
Solution: We shall prove that n = 25 is the only solution of the
problem. We show rst that for a "good" set some 25 points lie
in a square of side length 1. All points lie in the rectangle 1 
x  83 0  y  1: Divide this rectangle to 83 squares of side
length 1: If some 25 points lie in one of these rectangles then we
are done. Conversely, in every square there are at least 26 or at
most 24 points. We prove now that there exists a square with at
least 26 points and there exists a square with at most 24 points.
Indeed, if a square with 26 points does not exist then the points are
at most 83:24 = 1992 < 2000: If there is no square with less than
25 points then the points are 83:26 ; 82 > 2000: Further, move the
square with more than 25 points towards the square with less that
25 points. Since the number of points in this square changes at most
by one we get the assertion of the problem.
To prove b) let x1 = 0 xi = xi;1 + 1999 83 and y = 0 for i =
i
0 2 4 : : :  2000 while yi = 1 for i = 1 3 : : :  1999: Let XY ZT be an
unit square. WLOG we assume that it intersects the lines

26
y = 0 and y = 1 in the points P Q and R S respectively. Then
PQ+QR+RS +SP < PY +Y Q+QZ +ZR+RT +TS +SX +XP = 4:
But SP QR  1 and we get PQ + RS < 2: This implies that in the
square XY ZT there are no more than 25 points.
We found that n = 25 is a solution of the problem. It is unique
since 24 < 2000
83 < 25:
Problem 2. Given an acute 4ABC: Prove that there exist unique
points A1 B1 and C1 on BC AC and AB respectively with the fol-
lowing property: Each of the points is the midpoint of the segment
with ends the orthogonal projections of the other two points on the
corresponding side. Prove that 4A1B1C1 is similar to the triangle
formed by the medians of 4ABC:
Solution: It is easy to be seen that the perpendicular through C1 to
AB intersects A1B1 at its midpoint. The same is applied for A1 and
B1: Therefore the tree perpendiculars intersect at the medcenter of
4A1B1C1; the point P:

27
Let T be the midpoint of A1B1 <) PB1A1 = x =<) PCA1 <) PA1B1 =
y =<) PCB1: The Law of Sines for 4B1TP and 4A1TP gives:
sin = B1T = A1T = sin

sin x TP AP sin y
where sin x = sin : Further, if CN is the median in 4ABC similar
sin y sin

arguments imply that sinsin <) ACG = sin : Since x + y =

<) BCG sin

<) ACG+ <) BCG =  and  is acute it follows that x =<) ACG y =
<) BCG: It is clear now that CP is symmetric to the median in
4ABC through C with respect to the bisector of angle : The same
is true for AP and BP: Therefore the point P is unique (and there-
fore A1 B1 and C1 are unique). Further, <) B1C1A1 =<) B1C1P +
<) A1C1P =<) B1AP + <) A1BP =<) BAG+ <) ABG =<) BGK where
K is the symmetric point of G with respect to the midpoint of BC:
By analogy <) C1A1B1 =<) GBK and <) A1B1C1 =<) GKB: There-
fore 4A1B1C1 is similar to the triangle formed by the medians of
4A1B1C1:
28
Problem 3. Let p  3 be a prime number and a1 a2 : : : ap;2 be
a sequence of natural numbers such that p does not divide both ak
and akk ; 1 for all k = 1 2 : : :  p ; 2: Prove that the product of some
elements of the sequence is congruent to 2 modulo p:
Solution: Consider the sequence 1 a1 a2 : : : ap;2: We shall
prove by induction that for any i = 2 3 : : :  p ; 1 there exist in-
tegers b1 b2 : : :  bi each of which is a product of some elements from
the above sequence and bm 6 bn(mod p) for m 6= n: Indeed, for
i = 2 we can choose b1 = 1 b2 = a1(a1 6 1(mod p)): Suppose we
have chosen b1 b2 : : :  bi such that bm 6 bn(mod p) for m 6= n: Con-
sider b1ai b2ai : : : biai: It is easily seen that any two of them are not
congruent modulo p: Further, if bj ai for any j is congruent to bl for
some l we get that b1ai b2ai : : :  biai modulo p is a permutation of
b1 b2 : : : bi modulo p: Thus,
(b1ai):(b2ai): : : : :(biai)  b1:b2: : : ::bi(mod p)
and therefore aii  1(mod p); a contradiction.
It follows now that for any s = 2 3 : : :  p ; 1 there exist few
elements of the sequence from the condition of the problem such
that their product is congruent to s modulo p:

Problem 4. Find all polynomials P (x) with real coecients such

that
P (x):P (x + 1) = P (x2)
for any real x:
Solution: We show rst that for any natural number n there exists
at most one polynomial P (x) of degree n such that P (x):P (x + 1) =
29
P (x2): Indeed, if
P (x) = anxn + an;1xn;1 + : : : + a1x + a0
comparing the coecients in front of the same degrees of x in
P (x):P (x + 1) and P (x2) we get an = 1 and an;1 = ; n2 : It is
easily seen that each of the subsequent coecients is a solution of an
equation of rst degree of the form 2x + b = 0 where b is a function
of already chosen coecients. Therefore, if a polynomial of degree
n with the desired property exists it is unique one.
We prove now that a polynomial P (x) of odd degree such that
P (x):P (x + 1) = P (x2) does not exist. Let P (x) = xm(x ; 1)n :Q(x)
(m or n could be zero), where Q(x) is a polynomial such that Q(0) = 6
0 and Q(1) 6= 0: A substitution gives
(x + 1)m:xn:Q(x):Q(x + 1) = xm:(x + 1)n :Q(x2):
If m 6= n the substitution x = 0 leeds to a contradiction. Therefore
m = n and so Q(x):Q(x + 1) = Q(x2) where Q(x) is of odd degree.
Therefore there exists a real number x0 6= 0 1 such that Q(x0) = 0:
It is clear that x0 = ;1 since otherwise the substitution x = ;1
shows that Q(1) = 0 which is impossible. A substitution x = x0
gives that Q(x20) = 0 thus Q(x40) = 0 and so on. Since x0 6= 1 0 it
follows that therek
are innitely many distinct terms in the sequence
x0 x0 : : : x0  which is impossible.
2 2

Direct verication shows that for any even natural number n =

2k the polynomial P (x) = xk (x ; 1)k is of degree n and is a solution
of the problem.
Therefore all polynomials such that P (x):P (x + 1) = P (x2) are
P (x) = xk (x ; 1)k :
Problem 5. Let D be the midpoint of the base AB of an isosceles
acute 4ABC: A point E is chosen on AB and O is the circumcenter
30
of 4ACE: Prove that the line through D perpendicular to DO the
line through E perpendicular to BC and the line through B parallel
to AC intersect in a point.
Solution: Let <) ABC =<) BAC = and let G be the circumcenter
of 4ABC: Consider points F 0 and F 00 on the line through B parallel
to AC such that OD ? DF 0 and BC ? EF 00: Denote by H 0 and H 00
respectively the projection points of F 0 and F 00 on the line AB: Since
O is inner point for <) ADC and <) ACB < 90  we have that F 0 and
F 00 lie in the interior of <) BAC: It suces to prove that F 0  F 00
i.e. F 0H 0 = F 00H 00: Let O0 and G0 be respectively the projection
points of O on AB and of0 G on OO0 : Since 4DH 0F 0  040OO0D it0
follows that DH = OO0 : Also, <) GOG0 = and so B 0H 0 = OG0 :
0

F H DO
0 0 F H GG
It follows from GG00 =0 DO0  O0G0 = DG and <) DBG = 2 ; 90
that F 0H 0 = BD:O GD
D = ;tg2 :O0D: Denote BC \ EF 00 = I: Since
<) CBF 00 = 180 ; 2 and BE = 2O0D it is clear that F 00H 00 =
BF 00 sin = cos(180 BI sin = ;BE sin cos = ;O0Dtg2 =
; 2 ) cos 2
F 0H 0: This completes the prove.
Problem 6. Let A be the set of all binary sequences of length n
and let 0 2 A be the sequence with zero elements. The sequence c =
c1 c2 : : :  cn is called sum of a = a1 a2 : : : an and b = b1 b2 : : : bn
if ci = 0 when ai = bi and ci = 1 when ai 6= bi: Let f : A ! A be
a function such that f (0) = 0 and if the sequences a and b dier in
exactly k terms then the sequences f (a) and f (b) dier also exactly
in k terms. Prove that if a b and c are sequences from A such that
a + b + c = 0 then f (a) + f (b) + f (c) = 0:
Solution: Consider the sequence e1 = 1 0 : : :  0 0 e2 = 0 1 : : :  0
: : :  en;1 = 0 0 : : :  1 0 en = 0 0 : : :  0 1: It follows from the condi-

31
tion of the problem that (since f (0) = 0) for all p 1  p  n there
exists q 1  q  n such that f (ep) = eq:
It is clear also that f (ep) 6= f (eq ) for 1  p q  n p 6= q:
Therefore
(1) ff (e1) f (e2) : : :  f (en)g = fe1 e2 : : : eng:
Consider an arbitrary sequence a = a1 a2 : : : an with t ones. If
f (ep) = eq  and ap = 1 then the q;th term of the sequence f (a) is
also 1 (otherwise ep and a dier at t ; 1 terms whereas f (ep) = eq
and f (a) dier at t + 1 terms). By analogy if ap = 0 then the q;th
term of f (a) is also 0.
Finaly, consider the sequences a = a1 a2 : : :  an b = b1 b2 : : : bn
and c = c1 c2 : : : cn such that a + b + c = 0: This means that
for every i 1  i  n the sum ai + bi + ci is even number. Fix
i 1  i  n and let f (ei) = ej : It follows now that the j ;th terms of
the sequences f (a) f (b) f (c) coincide with ai bi ci respectively and
using (1) we obtain that f (a) + f (b) + f (c) = 0:

32
 Union of Bulgarian Mathematicians

Sava Grozdev Emil Kolev

BULGARIAN
MATHEMATICAL COMPETITIONS
2001

Soa, 2001
 Winter mathematics competition
Bourgas, 2-4 February 2001
Problem 9.1. a) Draw all points in the plane with coordinates
(x y) such that
(j3x ; yj ; 3)(j3x + yj ; 3) = 0
b) Find all x and y for which
(j3x ; yj ; 3)(j3x + yj ; 3) = 0
y ; f4xg = 0
;1  x  1
(For a real number x we denote the unique number in the interval
0 1) for which x ; fxg is an integer by fxg).
Solution: a) It is easy to see that these are the points on the four
lines l1 : 3x ; y = 3 l2 : 3x ; y = ;3 l3 : 3x + y = 3 and l4 : 3x + y =
;3:
b) There are two types of solutions:
y = ;3x + 3
1. Solutions of y = f4xg
;1  x  1
y = 3x + 3
2. Solutions of y = f4xg
;1  x  1
Denote the integer part of 4x by 4x] i.e. 4x = 4x] + f4xg: In
the rst case we have ;3x + 3 = f4xg = 4x ; 4x] i.e 7x = 3 + 4x]:
3
Since the right hand side is an integer all posiible values of x are
0  71   72  : : :  1: Direct verication shows that only x = 75  y = 67 
x = 76  y = 37  x = 1 y = 0 are solutions.
In the second case 3x + 3 = f4xg = 4x ; 4x] i.e. x = 3 + 4x]:
It follows that x is an integer, i.e. x = ;1 0 or 1: Direct verication
shows that solution is only x = ;1 y = 0.
Thus, there are four solutions:
x = 75  y = 76  x = 76  y = 37  x = 1 y = 0 x = ;1 y = 0:

Problem 9.2. Points A1 B1 and C1 are chosen on the sides BC CA
and AB of a triangle ABC: Point G is the centroid of 4ABC and
Ga Gb and Gc are centroids of 4AB1C1 4BA1C1 and 4CA1B1
respectively. The centroids of 4A1B1C1 and 4GaGbGc are denoted
by G1 and G2 respectively. Prove that:
a) the points G G1 and G2 lie on a straight line
b) lines AGa BGb and CGc intersect in a point if and only if
AA1 BB1 and CC1 intersect in a point.
Solution: a) Let O be an arbitrary point and
OA~ 1 = OB
~ + (1 ; )OC ~  OB
~ 1 =  OA
~ + (1 ;  )OC
~ 
~ 1 =  OA
OC ~ + (1 ;  )OB
~
where    2 (0 1): The existence of   and  follows from
the fact that A1 B1 and C1 lie on the sides BC CA and AB of
4ABC: Then we have
~ = 1 OA ~ + OB~ + OC ~ = 1 3OA ~ + 3OB~ + 3OC ~ 
   
OG 3 9
OG~ 1 = 13 OA
~ 1 + OB

~ 1 + OC
~1 = 

4
1 ( +  )OA
h
~ + ( ;  + 1)OB ~ + (2 ;  ;  )OC
~ :
i

3
By analogy OG ~ a = 1 (1 +  +  )OA
h
~ + (1 ;  )OB~ + (1 ;  )OC
~  i

3
~ b = 1  OA~ + (2 +  ;  )OB
~ + (1 ; )OC ~ 
h i
OG 3
~ c = 1  OA~ + OB ~ + (3 ;  ;  )OC
~ :
h i
OG 3
Thus, OG~ 2 = 1 OA
~ 1 + OB
~ 1 + OC
~1 =


3
1 (1 + 2 + 2 )OA
h
~ + (3 + 2 ; 2 )OB
~ + (5 ; 2 ; 2 )OC~ : i

9
Since GG~ 1 = OG~ 1 ; OG~ we have
~ 1 = 1 ( +  ; 1)OA
GG
h
~ + ( ;  )OB
~ + (1 ;  ;  )OC
~ : i

3
Using the same arguments we obtain GG ~ 2 = OG
~ 2 ; OG
~ =
1 (2 + 2 ; 2)OA
h
~ + (2 ; 2 )OB~ + (2 ; 2 ; 2 )OC~ : It follows
i

9
~ 1 = 3 GG
from the last two equalities that GG ~
2 2 and we are done.
b) Since the lines AA1 BB1 and CC1 intersect in a point Ceva's
theorem gives AC1  BA1  CB1 = 1:
C1B A1C B1A
Denote the intersecting points of AGa BGb and CGc with the sides
BC CA and AB by A2 B2 and C2 respectively. A necessary and
sucient condition for lines AGa BGb and CGc to intersect in a
CA2  BC2  AB2 = 1:
point is A
2 B C2A B2 C
Denote the midpoint of B1C1 by A3: Let h1 and h2 be the alti-

5
tudes of 4AA2C and AA3B1 from A2 and A3: We have
SAA2 C = h1:AC = AA2:AC :
SAA3 B1 h2:AB1 AA3:AB1
By analogy
SAA2 B = AA2:AB :
S AA3 C1AA :AC 3 1
Dividing the above equalities and using that SAA3 C1 = SAA3 B1 and
SAA2 C = CA2 we obtain
SAA2 B BA 2

CA2 = AC  AC1 :
A2B AB AB1
By analogy
BC2 = CB  CB1  AB2 = BA  BA1 :
C2A CA CA1 B2C BC BC1
Multiplying the above equalities gives
CA2 BC2 AB2 = AC AC1  BC CB1  BA BA1 = AC1 BA1 CB1 :
A2B C2A B2C AB AB1 AC CA1 BC BC1 C1B A1C B1A
Therefore the lines AGa BGb and CGc intersect in a point i the
lines AA1 BB1 and CC1 interect in a point.
Problem 9.3. Let An be the number of sequences from 0's and 1's
of length n such that no four consequtive elements equal 0101: Find
the parity of A2001:
Solution: Denote the number of sequences of length n, the last
three terms of which are ijk where i j k 2 f0 1g and no four con-
sequtive elements equal to 0101 by anijk :

6
 Obviously anijk+1 = an0ij + an1ij when ijk 6= 101 and an101+1 = an110:
Adding the above equalities for all values of ijk we obtain
An+1 = 2An ; an010:
Therefore An+1 an010 an001;1 + an101;1 an000;2 + an100;2 + an110;2
a0003 + an100;3 a000
n; n;4 + an;4 + an;4 + an;4 an;5 + an;5 + an;5 +
100 010 110 000 001 010
an100;5 + an111;5 + an110;5 + an101;5 + an011;5+ An;5(mod 2):
Thus, Ak+6m Ak (mod 2) for all k and m:
Therefore A2001 A3 = 8 0(mod 2):
Problem 10a.1. Find all pairs (a b) of integers such that the system
x2 + 2ax ; 3a ; 1 = 0
y2 ; 2by + x = 0
has exactly three real solutions.
Solution: Let (a b) satisfy the condition of the problem. Since there
are exactly three real solutions the rst equation has two distinct
roots x1 < x2 which implies D1 = 4(a2 + 3a + 1) > 0: The discrim-
inant of the second equation in respect to y equals D2 = 4(b2 ; x)
and therefore there are exactly three real solutions i x1 < x2 = b2:
These conditions are satised pi a and b are integers such that
a + 3a + 1 > 0 and b2 = ;a + a2 + 3a + 1: It follows from the
2

latter equality that a2 + 3a + 1 = c2 where c is a positive integer.

Therefore the discriminant of a2 +3a +1 ; c2 = 0 is a perfect square,
i.e. 9 ; 4(1 ; c2) = d2 where d is nonnegative integer. The last
equality can be written in the form (d ; 2c)(d +2c) = 5 which implies
that d ; 2c = 1 d + 2c = 5 i.e. d = 3 c = 1: Hence, a2 + 3a + 1 = 1
with roots a = 0 and a = ;3: Respectively b = 1 or b = 2: Direct
verication shows that all pairs (a b) = (0 1) (0 ;1) (;3 2) and
(;3 ;2) satisfy the condition of the problem.

7
Problem 10a.2. The tangential point of a circle k through the
vertex C of a 4ABC and the line AB is the vertex B . The circle k
intersects for a second time the side AC snd the median of 4ABC
through C at points D and E respectively. Prove that if the inter-
secting point of the tangents to k through C and E lies on the line
BD then <) ABC = 90:
Solution: Let F be the intersecting point of the tangents tC and
tE to k at C and E G = BD \ CE and ACBH is parallelogram.
We have that <) ABC = 90 () tC k AB () FD FB CA= CD :
CD CD GD
From the other hand CA = BH = GB  i.e. it suces to show that
FD = GD ( ): Since 4FBE  4FED and 4FBC  4FCD
FB GB
we obtain FBFE ED= BE and FB = BC : Further, FC 2 = FE 2 =
FC CD
FD CD:ED
FB:FD and so FB = CB:EB : Now the equality from ( ) follows
from CD:ED = SCED = GD  which completes the proof.
CB:EB SCEB GB
Problem 10a.3. Ivan and Peter alternatively write down 0 or 1
until each of them has written 2001 digits. Peter is a winner if the
number, which binary representation has been obtained, cannot be
expressed as a sum of two perfect squares. Prove that Peter has a
winning strategy.
Solution: First we prove that if the binary representation of a pos-
itive integer ends with two ones and even number of zeroes then
this integer cannot be represented as sum of two squares. Indeed,
such a number is of the form 4k (4s + 3) and if we suppose that
x2 + y2 = 4k (4s + 3) then using the fact that x2 + y2 equals 0 mod-
ulo 4 i x and y are even we get that there exist integers p q for
which p2 + q2 = 4s + 3 which is a contradiction.

8
 The winning strategy of Peter could be:
If one of Ivan's digits is 1 then Peter simply repeats all digits
written by Ivan. The nal number is of the form 4k (4s + 3) and
cannot be written as x2 + y2: If all Ivan's digits are zeroes then
the rst three digits of Peter are 1 1 1 after which he writes only
zeroes. The nal number is (0101010 : : : 0)2 = 21:41998 and cannot
be represented as p2 + q2 since 21 cannot be written in that form.
Problem 10b.1. Find all values of the real parameter a such that
the equation
logx (x2 + x + a)2 = 4
has unique solution.
Solution: The equation is equivalent to (x2 + x + a)2 = x4 x >
0 x 6= 1: Further, (x2 + x + a)2 = x4 () p (x + a)(2x2 + x +
a) = 0 with roots x1 = ;a and x2=3 = ;1  4 1 ; 8a  provided
; 1 ; p1 ; 8a
1 ; 8a  0: Since 4 < 0 we obtain that only x1 =
p
;a x2 = ;1 + 4 1 ; 8a can be roots of the equation from the prob-
lem. If a  0 then x1  0 and x2  0 which implies that the
equation has no roots. If a < 0 then x1 > 0 and x2 > 0: Note
that x1 = x2 implies a = 0 which is a contradiction. Therefore if
the equation has unique solution then it is necessary to have a < 0
and one of the two roots equals 1. Thus, x1 = 1 ) a = ;1 and
x2 = 1 ) a = ;3:
Problem 10b.2. On each side of a right isosceles triangle with
legs of length 1 is chosen a point such that the triangle formed from
these three points is a right triangle. What is the least value of the
hypotenuse of this triangle?

9
Solution: Consider a right isosceles triangle ABC with right angle
at C . Let A1 2 BC B1 2 CA and C1 2 AB be such that 4A1B1C1
is right triangle.
1. Suppose <) B1A1C1 = 90 : Assume that the circle k with di-
ameter B1C1 intersects BC at point X 6= A1: It is easy to be seen
that X is an interior point for the line segment BC: Draw a tangent
l to k which is parallel to BC
_
and the tangential point Y belongs
to the smaller of the arcs A1X : Denote the intersecting points of l
with AB and AC by C2 and B2 respectively. Consider a homothety
of center A such that the image of B2(C2) is C (B ): It is clear that
the image of 4B1C1Y is inscribed in 4ABC and its hypotenuse
is less than B1C1: Therefore wlog we may assume that BC is tan-
gent to k: Thus, if <) C1B1A1 =  then <) C1A1B = : From the
Sine Low for 4BC1A1 we obtain BA1 = B1C1 sinsin  sin( + 45) =
45
B1C1 sin (sin  + cos ): Hence
(1) 1 = BA1 + A1C = B1C1 cos  sin  + B1C1 sin (sin  + cos )
Therefore 1 = B1C1(sin2  + sin 2) () B1C1(1 + 2 sin 2 ;
cos 2) = 2: From the other hand 2 sin 2 ; cos 2 =
p 2 p p
5 p sin 2 ; p1 cos 2 = 5 sin(2 ;
)  5 where
is
 !

5 5
such that sin
= p and cos
= p2 : Note that the equality holds
1
5 5
when 2 ;
= 90 and since
< 90 we have that  < 90:
p
Further Bp1C1(1 + 5)  B1C1(1 + 2 sin 2 ; cos 2p) = 2 and
so B1C1  52; 1 : It is easily seen that if B1C1 = 52; 1 and
2 ;
= 90 p then (1) holds true and therefore there exists triangle
with B1C1 = 52; 1 :
2. Suppose <) A1C1B1 = 90: Denote the projection of A1 and
10
B1 on AB by A2 and B2 respectively. Obviously AB2 = B1B2
and BA2 = A1A2: Let P be the midpoint of A1B1 and Q { the
midpoint of A2B2: Thus, A1B1 = 2PC1  2PQ = A1A2 + B1B2 and
A1B1  A2B2 = AB ; AB2 ; BA2 = AB ; B1B2 ; A1A2: Adding p
p
the above equalities gives 2A1B1  AB = 2 and so A1B1  22 :
p p
Since 22 > 52; 1 we obtain that the least possible value equals
p
5 ; 1:
2
Problem 10b.3. An element x is chosen from the set
A = f1 2 : : :  2n g n  3: Questions of the type: Does x belong
to B  A where the sum of the elements of B equals 2n;2 (2n + 1)
are allowed? Prove that one can nd x with exactly n questions
stated in advance.
Solution:
Lemma There exist n sets each with 2n;1 elements, the sum
of the elements of each set equals 2n;2 (2n + 1) with the following
property: the elements from the set f1 2 : : :  2n g get as answers
distinct n;tuples from "yes" and "no".
Proof: Induction by n  3. For n = 3 the sum of the elements
of B is 18 and we use the sets B1 = f1 2 7 8g B2 = f1 3 6 8g and
B3 = f1 4 6 7g: The table shows that the elements from A get as
answers distinct triples.
1 2 3 4 5 6 7 8
B1 = f1 2 7 8g + + - - - - + +
B2 = f1 3 6 8g + - + - - + - +
B3 = f1 4 6 7g + - - + - + + -
+ means "yes", ; means "no".

11
 The number of elements in each set is 22 = 23;1 :
Suppose the assertion is true for some m: Therefore there exist
sets B1 B2 : : : Bm each with 2m;1 elements, the sum of the elements
of each set equals 2m;2 (2m +1) and every element from f1 2 : : :  2mg
gets distinct m;tuple.
Consider the set f1 2 : : :  2m+1g: For any i 1  i  m if Bi =
fa1i a2i : : :  a2m;1i g let
Di = fa1i a2i : : :  a2m;1i a1i + 2m a2i + 2m  : : :  a2m;1i + 2m g:
It is clear that each set Di  1  i  m has exactly 2m elements
and the sum of the elements of Di is equal to: 2(a1i + a2i + : : : +
a2m;1i)+2m;1 :2m = 2:2m;2 (2m +1)+22m;1 = 22m;1 +2m;1 +22m;1 =
2m;1 (2m+1 + 1): It is easily seen that only the elements t and t + 2m
for 1  t  2m get equal m;tuples. Let P and Q be nonintersecting
sets such that jP j = jQj = 2m;1 and P Q = f1 2 : : :  2m g: Consider
a set Q obtained from Q by adding 2m to each of its elements. The
set Dm+1 = P  Q has 2m elements and the sum of its elements
is equal to s2P s + s2Q s + 2m;1 :2m = 2m;1 (2m + 1) + 22m;1 =
P P

2m;1 (2m+1 + 1): It is clear that exactly one element of each pair
(t t + 2m ) for 1  t  2m belongs to Dm+1 : Hence, we have found
the desired sets D1 D2 : : :  Dm+1:
It is obvious now that the sets, given by the Lemma solve the
problem.
Problem 11.1. A sequence a1 a2 : : :  an : : : is dened by
a1 = k a2 = 5k ; 2 and an+2 = 3an+1 ; 2an n  1
where k is a real number.
a) Find all values of k such that the sequence fang1n=1 is conver-
gent.
12
 b) Prove that if k = 1 then
an+2 = 7a1n++1a; +8aan an+1  n  1
" #
2

n n+1
where x] denotes the integer part of x:
Solution: a) Write the given reccurent relation in the form an+2 ;
an+1 = 2(an+1 ; an) and consider the sequence cn = an+1 ; an:
It is obvious that if the sequence fang1n=1 is convergent then the
sequence fcng1 n=1 is also convergent. From the other hand it follows
from c1 = a2 ; a1 = 4k ; 2 and cn+1 = 2cn that the sequence fcn g1n=1
is an arithmetic progression and cn = (4k ; 2):2n;1 for any positive
integer n: Thus, if 4k ;2 6= 0 then the sequence fcn g1n=1 is unbounded
and therefore is not convergent one. Therefore 4k ; 2 = 0 i.e. k = 12 :
For this value of k all terms of the sequence fang1n=1 are equal to 21
and therefore the sequence is convergent.
b) It easily follows by induction that an = 2n ; 1: From the other
hand the equality from the condition of the problem is satised for
n = 1 2: Suppose it is true for n = m and n = m + 1: Then
7a2m+1 ; 8amam+1 = 7(2m+1 ; 1)2 ; 8(2m ; 1)(2m+1 ; 1) =
" # " #

1 + am + am+1 1 + 2m ; 1 + 2m+1 ; 1
3:22m+2 ; 2m+2 + ;1 = 2m+2 ; 1 = a 
" #

m+1
3:2m ; 1 3:2m ; 1
which completes the proof.
Problem 11.2. On each side of a triangle with angles 30  60 and
90 and hypotenuse 1 is chosen a point such that the triangle formed
from these three points is a right triangle. What is the least value
of the hypotenuse of this triangle?
13
Solution: Consider triangle ABC with angles   and  . Let
A1 2 BC B1 2 CA and C1 2 AB be such that 4A1B1C1 is right
with <) A1C1B1 = 90 : Suppose that the circle k of diameter A1B1
intersects AB at point X 6= C1: It is easy to see that if   90
and   90 then X is an interior point for line segment AB: Draw
a tangent l to k which is parallel to AB _
and the tangential point
Y belongs to the smaller of the arcs C1X : Denote the intersecting
points of l and lines AC and BC by B2 and A2 respectively. Con-
sider homothety of center C for which the image of B2(A2) is the
point A(B ): It is obvious that the image of 4B1A1Y is inscribed
in 4ABC and its hypotenuse is less than B1A1: Therefore wlog we
may suppose that AB is tangent to k: In this case if <) B1A1C1 =

then <) B1C1A =

:
The Sine Low for 4AC1B1 and 4BC1A1 gives
AC1 = B1A1 sinsin
sin( +
)  BC = B1A1 cos
sin(90 ;
+  ) :
 1
sin 
Since AB = AC1 + C1B we obtain
(1) AB = B1A1 sinsin
sin( +
) + B1A1 cos
sin(90 ;
+  )
 sin 
This equality is equivalent to
2AB = A1B1(cotg + cotg + 2 sin 2
; (cotg ; cotg ) cos 2
):
From the other hand
2 sin 2
; (cotg ; cotg ) cos 2
=
q q
4 + (cotg ; cotg )2 sin(2
; )  4 + (cotg ; cotg )2
where
cos  = q
2  sin q= cotg ; cotg :
4 + (cotg ; cotg )2 4 + (cotg ; cotg )2
14
It is easily seen that such an angle
always exists.
Finally, the least value is
B1A1 = q
2AB :
cotg + cotg + 4 + (cotg ; cotg )2

It is clear that if B1A1 = q

2AB
cotg + cotg + 4 + (cotg ; cotg )2
and 2
;  = 90 then (1) holds true and therefore there exists a
triangle with B1A1 = q
2AB :
cotg + cotg + 4 + (cotg ; cotg )2
It remains to compute the above expression for the three possible
ways to inscribe a right triangle p p ofpangles 30p 60 and 90
in triangle
and hypotenuse 1. We obtain 3912; 3  214; 3  43 : The least
p p
value is 3912; 3 which is the answer of the problem.
Problem 11.3. The plane is divided into unit squares by lines
parallel to coordinate axes of an orthogonal coordinate system. Find
the number of paths of length n from the point with coordinates
(0 0) to the point with coordinates (a b) moving along the sides of
the unit squares.
Solution: Divide the path into unit paths, i.e. paths between two
neighbouring points of integer coordinates. Denote the number of
moves up by x1 down by y1 right by x2 and left by y2: The condition
of the problem gives x1 + x2 + y1 + y2 = n x1 ; y1 = b x2 ; y2 = a:
Thus, y1 = x1 ; b x2 = n + 2a + b ; x1 y2 = n ; 2a + b ; x1: Since
jaj + jbj = jx2 ; y2j + jx1 ; y1j  x1 + x2 + y1 + y2 = n we obtain
that it is necessary to have a + b  n and a + b n(mod 2): Let

15
 n
 !

us rst x the moves up and rigth. This can be done by n+a+b

2

ways. After that in these already xed n + a + b positions x the

2
moves up and nally, in the remaining n ; 2a ; b positions x the
moves left. We obtain
n+a;b 

n
!
2 n+a+b ! n;a;b !
n;a+b ; i :
X
2 2
n+a+b i
2 i=b 2

n+a;b 
2
X n+a+b ! n;a;b !
The sum 2
i n;a+b ; i equals to the coecient in front
2
i=b 2
of x n;2a+b in the expansion of (1 + x) n+2a+b (1 + x) n;2a;b = (1 + x)n:
n
 !

Therefore the sum equals n;a+b : The number of paths is:

2

n n
 ! !

n+a+b n;a+b :
2 2

Answer: If jaj + jbj > n or a + b 6 n(mod 2) the number of

paths is 0, otherwise the number of paths is n+na+b n;na+b :
 ! !

2 2

16
 Spring mathematics tournament
Kazanlak, 30 March - 1 April 2001
Problem 8.1. Let a be a real parameter such that 0  a  1: Prove
that the solutions of the inequality
jxj + jax + 21 j  1
form an interval of length greater or equal to 1.
Solution: If a = 0 then all numbers from the interval 12  ; 12 of


length 1 are solutions of the problem. If a = 1 then the inequality

becomes jxj + jx + 12 j  1: The solutions of this inequality are all
numbers from the interval ; 3  1  which is of length 1.


4 4
Further, assume that 0 < a < 1 and let x  0: In this case
ax + 21 > 0 and therefore the solutions are x  2(1 1+ a) : Let x < 0:
If ax + 21  0 i.e. x  ; 12  the solutions satisfy x  ; 2(1 1; a) : If
ax + 21 < 0 i.e. x < ; 21a  the solutions satisfy x  ; 3(1 1+ a) :
Let a  12 : Then ; 2(1 1; a)  ; 21a and so all numbers x 
; 2(1 1; a) are solutions of the inequality. Since ; 3(1 1+ a) > ; 21a
it has no other solutions. Therefore all solutions form an interval
; 2(1 1; a)  2(1 1+ a)  of length 2(1 1+ a) + 2(1 1; a) = 1 ;1 a2 > 1:
" #

Let a > 12 : Then ; 2(1 3+ a) < ; 21a and the solutions are x 

17
; 2(1 3+ a) x  ; 21a or x 2 ; 2(1 3+ a)  ; 21a : Therefore all so-

" #!

lutions form an interval ; 2(1 3+ a)  2(1 1+ a)  of length 1 +2 a  1:

" #

Problem 8.2. Given a square ABCD of side length 1. Point

M 2 BC and point N 2 CD are such that the perimeter of 4MCN
is 2.
a) Find <) MAN 
b) If P is the foot of the perpendicular from A to MN nd the
locus of the point P:
Solution: a) Let K be a point on the extension of CB such that
BK = DN: Then MK = MB + BK = MB + DN = 1 ; CM +
1 ; CN = 2 ; (CM + CN ) = MN (using that CM + CN + MN =
2). Since 4ABK  = 4ADN we have that AK = AN: Therefore
4AMN  = 4AMK and it follows that <) MAN =<) MAK i.e.
<) MAN = 12 <) KAN: Furthermore <) KAB =<) NAD which im-
plies that <) KAN =<) BAD = 90: Therefore <) MAN = 45 :
b) It follows from 4AMN  = 4AMK that <) AMN =<) AMK:
Thus, 4APM  = 4ABM and AP = AB = 1: Therefore P lies on a
circle of center A and radius 1: Finally, the locus of P is an arc of a
circle with center A and radius 1 excluding points B and D:
Problem 8.3. a) Prove that
12 + 22 +    + n2 = n(n + 1)(2
6
n + 1)

holds true for any positive integer n:

18
 b) Find the least integer n n > 1 for which
12 + 22 +    + n2
n
is a perfect square.
Solution:
a) Direct verication shows that
n(n + 1)(2n + 1) + (n + 1)2 = (n + 1)(n + 2)(2n + 3)
6 6
for any integer n: Using this equality the proof is easily done by
induction.
b) It follows from a) that n(n + 1)(2n + 1) = 6m2: Since 2n + 1
is odd then n + 1 is even, i.e. n is odd. Let n = 2k ; 1: Then
k(4k ; 1) = 3m2: Therefore 3=k or 3=4k ; 1 i.e. 3=k or 3=k ; 1: Let
k = 3l: Then l(12l ; 1) = m2: Since (l 12l ; 1) = 1 we have l = n2
and 12l;1 = v2: The latter equality is impossible (contradiction both
modulo 4 and modulo 3). Let k = 3l +1: Then (3l +1)(4l +1) = m2:
Since (3l + 1 4l + 1) = 1 we get that 3l + 1 = u2 4l + 1 = v2 u >
1 v > 1: Verifying consequently for v = 3 5 7 9 11 13 : : : we obtain
l = 2 6 12 20 30 42 56 : : : : Thus, 3l + 1 = 7 19 37 61 91 127 169
(rst perfect square). So, the least n is 2k ; 1 where k = 3l + 1 =
3:56 + 1 = 169 and therefore n = 337:
Problem 9.1. Let f (x) = x2 + 6ax ; a where a is a real parameter.
a) Find all values of a for which the equation f (x) = 0 has at
least one real root.
b) If x1 and x2 are the real roots of f (x) = 0 (not necessarily
distinct) nd the least value of the expression
A = (1 +9xa ;)(14a+ x ) ; (1 ; 6a ;70xa)(1+;1 6a ; x ) :
2 3

1 2 1 2

19
Solution: a) The discriminant of f (x) is D = 4(9a2 + a). Therefore
the equation f (x) = 0 has at least one real root i 9a2 + a  0, so
giving a 2 ;1 ; 91  0 +1).

b) Since f (x) = (x ; x1)(x ; x2) we obtain f (;1) = (;1 ;

x1)(;1 ; x2) = (1 + x1)(1 + x2) = (;1)2 + 6a(;1) ; a = 1 ; 7a and
f (1 ; 6a) = (1 ; 6a ; x1)(1 ; 6a ; x2) = (1 ; 6a)2 + 6a(1 ; 6a) ; a =
1 ; 7a: Therefore the denominators of the two fractions of A are
equal to 1 ; 7a and
A = ;70a ;1 4;a 7a+ 9a ; 1 = (1 ; 7a)(10 a2 + 2a ; 1) = 10a2+2a;1:
3 2

1 ; 7a
The quadratic function g(a) = 10a2 + 2a ; 1 attains its minimal
value for a = ; 10 1 62 ;1 ; 1  0 +1): Therefore the minimal

9
value of A equals to the smallest of the numbers g(0) and g ; 1 ,



1 
89
i.e. this value is g ; 9 = ; 81 .

Problem 9.2. Given a convex quadrilateral ABCD such that

OA = OC OB:OD  where O is the intersecting point of its diago-
+ OD
nals. The circumcircle of 4ABC intersects the line BD at point Q:
Prove that CQ is the bisector of <) DCA:
Solution: Let CQ1 Q1 2 BD be the bisector of <) DCO: Therefore
DQ1 = DC
Q1O CO
This equality, combined with the condition of the problem gives
OA(OC + OD) = OB:OD () OA Q1OQ+ODQ1 CD = OB:OD
1

20
 () OA:CO: QDOO = OB:OD () OA:CO = Q1O:OB:
1
Therefore the quadrilateral ABCQ1 is cyclic. Thus, Q1 Q:
Problem 9.3. Prove that there exist eight consecutive positive
integers such that non of them can be written in the form
j7x2 + 9xy ; 5y2j, where x and y are integers.
Solution: Denote f (x y) = 7x2 + 9xy ; 5y2. Since f (0 0) = 0,
f (0 1) = 5, f (1 0) = 7, f (1 1) = 11 and f (0 2) = 20, rst possible
sequence of eight positive integers is 12 13 : : :  19. We shall prove
that non of these integers can be written in the form j7x2 +9xy ;5y2j,
where x and y are integers.
Let f (x y) = k, where x and y are integers. It suces to prove
that f (x y) = k has no solutions for k 2 f12 13 : : :  19g.
Suppose k is even. Then x and y are also even. If x = 2x1 and
y = 2y1 we get the equality 4f (x1 y1) = k which implies that k is
divisible by 4. Thus, k 6= 14 and k 6= 18. Let k = 16 and consider
the equation 4f (x1 y1) = 16 which is equivalent to f (x1 y1) = 4.
As above we conclude that x1 and y1 are both even and let x1 = 2x2
and y1 = 2y2. Therefore f (x y) = 1. By analogy if k = 12 we get
the equation f (x y) = 3.
Multiply the equation f (x y) = k by 28 and write it in the
form
(14x + 9y)2 ; 221y2 = 28k:
Since 221 = 13:17 it is appropriate to consider modules 13 and 17.
Denote t = 14x + 9y and consider all possibilities for k, i.e.
k 2 f1 3 13 15 17 19g.
1) If k = 13 then t2 28:13 (mod 17) and so t2 7 (mod
17). Raising this congruence to 8-th power gives t16 (7)8 ;1
21
(mod 17) which is a contradiction to Fermat's theorem. Therefore
k 6= 13.
2) If k = 15 then t2 28:15 5 (mod 17). Raising in 8-th
power gives t16 (5)8 ;1 (mod 17), a contradiction.
The cases k = 17 19 1 3 are treated similarly.
Problem 10a.1. Let a and b be positive numbers such that both
of the equations (a + b ; x)2 = a ; b and (ab + 1 ; x)2 = ab ; 1 have
two distinct real roots. Prove that if the two bigger roots are equal
then the two smaller roots are also equal.
Solution: Let a and b satisfy the condition of the problem. Both
equations have two distinct roots i a > b and ab > 1: Since a > 0
it follows that a2 > ab > 1 i.e. a > 1: Therefore a > b > a1 and
a > 1: It follows from
p p of the problem thatpa + b +
p the condition
a ; b = ab + 1 + abp ; 1 i.e. pa ; b = (a ; 1)(b ; 1) + ab ; 1:
If a > b  1 then a ; b  ab ; 1 () (a + 1)(b ; 1) 
0 () b  1: Therefore b = 1: Conversely if 1 < b  1 then
p p a
a ; b  ab ; 1 () (a + 1)(b ; 1)  0 () b  1 i.e. b = 1:
Thus, the two bigger roots are equal i b = 1 and a > 1: In this case
the two smaller roots are also equal.
Problem 10a.2. Let A1 and B1 be points respectively on the sides
BC and AC of 4ABC D = AA1 \ BB1 and E = A1B1 \ CD: Prove
that if <) A1EC = 90 and the points A B A1 E lie on a circle, then
AA1 = BA1:
Solution: Let F = AE \ BC: We prove that EA1 is the bisec-
tor of <) BEF which solves the problem. Indeed, then we have
<) BAA1 =<) BEA1 =<) FEA1 =<) ABA1 i.e. AA1 = BA1:

22
 Using Ceva's and Menelaus's theorems for 4AA1C we obtain
AD  A1F  CB1 = 1 AD  A1B  CB1 = 1 and so
A1D CF AB1 A1D CB AB1
(1) A1F = CF
A1B CB
Let B 0 be a point on the ray A1B ! such that <) B 0EA1 =<) A1EF:
Then EA1 is the bisector of <) B 0EF and since <) A1EC = 90 it
follows that EC is the external bisector of the same angle. Therefore
A1F = CF and it follows from (1) that A1B = A1B 0  i.e. B = B 0
A1B 0 CB 0 CB CB 0
which completes the proof.
Problem 10a.3. Find all positive integers x and y such that
x3 + y3 ; x2y2
(x + y)2
is a nonnegative integer.
Solution: Let x and y be positive integers such that
z = x +y ;x y
3 3 2 2

(x + y)2
is a nonnegative integer. Substitute a = x + y and b = xy and write
the expression in the form b2 +3ab ; a2(a ; z) = 0: The discriminant
of this quadratic equation a2(4a + 9 ; 4z) is a perfect square, so
(4a + 9 ; 4z) = (2t + 1)2: Thus, a = t2 + t + z ; 2 and from the
equation for b we obtain that b = a(t ; 1): Since t  2 we have
(x ; y)2 = a2 ; 4a(t ; 1) < (a ; 2(t ; 1))2: From the other hand
a  t2 and therefore a2 ; 4a(t ; 1)  (a ; 2(t ; 1) ; 2)2: Since
a2 ; 4a(t ; 1) 6= (a ; 2(t ; 1) ; 1)2 (the two numbers are of dierent
parity) it follows that (x ; y)2 = a2 ; 4a(t ; 1) = (a ; 2(t ; 1) ; 2)2:
23
Thus, a = t2 i.e. t + z = 2 and so t = 2 z = 0 which implies
a = b = 4: Therefore x = y = 2 and these are the only positive
integers satisfying the condition of the problem.
Problem 10b.1. Solve the equation:
p
3log (cos x+sin x)+ ; 2log (cos x;sin x) = 2:
3
1
2 2

Solution: Note that the admissible values are those x for which
cos x + sin x > 0 cos x ; sin x > 0: After simple calculations the
equation becomes
p p p
( 3 ; 1) cos x + ( 3 + 1) sin x = 2:
This equation is equivalent to
p p p p
6 ; 2 cos x + 6 + 2 sin x = 1 :
4 4 2
p p p p
Since cos 15 = 6+4 2 and sin 15 = 6;4 2 we have
sin(x + 15 ) = 12 :
From the roots of the later equation only x = 15  k360 are ad-
missible. Therefore the roots are x = 15  k360 :
Problem 10b.2. Given a triangle ABC: Let M be such an interior
point of <) BAC that <) MAB =<) MCA and <) MAC =<) MBA:
Analogously, let N be such an interior point of <) ABC that
<) NBA =<) NCB and <) NBC =<) NAB and let P be such an inte-
rior point of <) ACB that <) PCA =<) PBC and <) PCB =<) PAC:
Prove that lines AM BN and CP intersect in a point on circumcircle
of 4MNP:
24
Solution: Suppose <) C < 90 : The case <) C > 90 is treated simi-
larly. Denote the intersecting point of line CP and line segment AB
by C1. Since PC1 is the bisector of <) APB (<) APC1 =<) BPC1 =
 =<) ACB ) we get BCAC1 = AP : Triangles APC and BPC are sim-
1 BP
AP
ilar which gives CP = BC and BP
AC
CP AC= BC =) AP = AC 2 = b2 :
BP BC 2 a2
By analogy, if A1 and B1 are the intersecting points of AM and
BA c2 CB
BN with BC and AC respectively then A C = b2 and B A = c2 :
1 1 a 2

1 1
AC 1 :BA1:CA1
Thus, C B:A C:B A = 1 i.e. according to Ceva's theorem lines
1 1 1
AM BN and CP intersect in a point. Dnote this point by K: Since
<) APB =<) AOB = 2 where O is the circumcenter of 4ABC we _
get that A P O and B lie on a circle and PC1 intersects the arc AB
at its midpoint C1. Therefore OC2 is a diameter of this circle and
<) OPC2 =<) OPC = 90 : It easily follows now that M N and P lie
on a circle of diameter OK:
Problem 10b.3. Consider a set P of six four-letter words over an
alphabet of two letters a and b: Denote by QP the set of all words
over the same alphabet which do not contain as subwords the words
from P . Prove that:
a) if QP is nite then it does not contain words of length  11
b) there exists a set P such that QP is nite and it contains word
of length 10.
Solution: a) Suppose that QP contains a word of length 11. We
will show that QP contains words of any length. Let ! be a word of
length 11. There are 9 subwords of ! of length 3. Since there are 8
distinct words of length 3 it follows that there exists a word  that
appears as subword of ! twise. Let  = 123 where i 2 fa bg:
25
Consider the subword of ! obtained after the second appearance of
 i.e. consider the word
: : : 1231 : : : 123
Write 1 after 3. Obviously the new word does not contain subwords
from P:
: : :2312 : : : 231
By analogy write 2 and so on. Thus, we nd words of any length
without subwords from P: This contradiction shows that there are
no words of length  11 in QP .
b) A direct verication shows that the set
P = f0000 1000 1001 1010 1101 1111 g
is such that QP is nite and 0001011100 2 QP :
Problem 11.1. Prove that there exist unique numbers  and 
such that cos  = 2  tg = 1 and 0 <  <  < 1:
Solution: The function f (x) = cos x ; x2 is continuous in the in-
terval (0 1) f (0) = 1 > 0 and f (1) = cos 1 ; 1 < 0: Therefore there
exists  such that cos  = 2: In the interval (0 1) the function cos x
is decreasing one, whereas the function x2 is increasing. Therefore
there exists unique  2 (0 1) such that cos  = 2:
We show now that in the interval 0 1] there exists a unique
 such that g( ) = 0 where g(x) = xtgx ; 1. The function g(x) is
increasing one because xtgx is increasing as product of two increasing
functions. Moreover tg 4 = 4 ; 1 < 0 g(1) = tg1 ; 1 = tg1 ;



tg 4 > 0 and the uniquennes of  follows. Since sin x < x for positive

26
x it follows that g() = tg ; 1 = sin  ; 1 < 0: Further, g(x) is
increasing in the interval 0 1] and so  < :
Problem 11.2. Let AA1 and BB1 be the altituds of obtuse non-
isosceles 4ABC and O and O1 are circumcenters of 4ABC and
4A1B1C respectively. A line through C intersects the line segments
AB and A1B1 at points D and D1 respectively and E is point on
the line OO1 such that <) ECD = 90 : Prove that EO 1
EO CD= CD1 :

Solution: Let F and F1 be the feet of the perpendiculars from O

and O1 to CD: Since CO1 ? AB we have
(1)
EO1 = CF1 = CO1 cos <) O1CF1 = CO1 sin <) BDC
EO CF CO cos <) OCF CO sin(<) DCB + <) BAC ) :
From the other hand the Low of Sine's for 4A1B1C 4ABC
4A1D1C and 4BDC gives
(2) CO1 = 2 sin A 1C
= A1C
<) A1B1C 2 sin <) ABC

(3) CO = 2 sin BC
<) BAC
(4)
CD1 = CA1 sin <) B1A1C = CA1 sin <) BAC
sin(<) B1A1C + <) D1 CA1) sin(<) BAC + <) DCB )

(5) CD = BCsinsin<) <BDC

) ABC

27
It follows now from (1),(2),(3),(4) and (5) that
EO1 = A1C sin <) BAC sin <) BDC = CD1 
EO BC sin <) ABC sin(<) BAC + <) DCB ) CD
which completes the proof.
Problem 11.3. There are 2001 towns in a country every one of
which is connected with at least 1600 towns by direct bus line. Find
the largest n for which there exist n towns any two of which are
connected by direct bus line.
Solution: Let S1 snd S1 be two towns connected by direct bus line.
If k is the number of towns connected to both S1 and S2 by bus line
then (1599 ; k)+(1599 ; k)+ k  1999 which implies that k  1198:
Therefore there exists town S3 connected to both S1 and S2: Further,
let k be the number of towns connected to all S1 S2 and S3: Therefore
(1197 ; k) + (1598 ; k) + k  1998 so giving k  797: Therefore
there exists town S4 connected to all S1 S2 and S3: By analogy let k
be the number of towns connected to all S1 S2 S3 and S4: We have
(796 ; k) + (1597 ; k ) + k  1997 which implies k  396: Therefore
there exists S5 connected to all S1 S2 S3 and S4: For the number n
from the condition of the problem we obtain n  5: We show that
n = 5: For, number the towns by S1 S2 : : : S2001 and connect Sk
and Sm with direct line for all k and m for which k 6 m(mod5):
Since 2001


5 = 400 we have that each town is connested to 1600 or

1601 other towns, i.e. the condition of the problem is satised.
For arbitrary 6 towns the numbers of at least two are equal mod-
ulo 5 and therefore they are not connected to each other. Therefore
n < 6 and so n = 5:

28
 L National Mathematics Olympiad
3rd round, 28-29 April 2001
Problem 1. For which values of the real parameter a the equation
lg(4x2 ; (8a ; 1)x +5a2 )+ x2 +(1 ; 2a)x +2a2 = lg(x2 ; 2(a +1)x ; a2 )
has exactly one root?
Solution: Write the equation in the form:

lg(4x2 ; (8a ; 1)x + 5a2) + 4x ; (8a ;3 1)x + 5a =

2 2

lg(x ; 2(a + 1)x ; a ) +

2 2 x2
; 2( a + 1) x ; a 2
:
3
Since lg t + t is an increasing function this equality is equivalent
to 3

4x2 ; (8a ; 1)x + 5a2 = x2 ; 2(a + 1)x ; a2

x2 ; 2(a + 1)x ; a2 > 0
The equation in the above system is equivalent to f (x) = 0 where
f (x) = x2 ; (2a ; 1)x + 2a2: Substitution x2 = 2ax ; x ; 2a2 in
the inequality gives x < ;a2: Therefore we have to nd those a for
which the equation f (x) = x2 ; (2a ; 1)x + 2a2 = 0 has exactly one
root such that x < ;a2: There are three cases to be considered:
1. f (x) = 0 has two roots and ;a2 is between the roots. In this
case f (;a2) < 0 ) a2(a + 1)2 < 0 which is impossible.
2. f (x) = 0 has two roots one of which equals ;a2 and the
second one is less than ;a2: in this case f (;a2) = 0 and so a = ;1
29
or a = 0: When a = ;1 we obtain x1 = ;2 x2 = ;1 which implies
that a = ;1 is a solution. If a = 0 we get x1 = ;1 x2 = 0 which
implies that a = 0 is also a solution.
3. f (x) = 0 has one root which is less than ;a2p: Since D = 0 we
get that ;4a2 ; 4a + 1 = 0 and so a1=2 = ; 1  2 : It is easy to
2
check that for both values of a the corresponding root is less than
;a2: p
;
Therefore the solutions are a = ;1 a = 0 a = 2 2 :1 

Problem 2. Diagonals AC and BD of a cyclic quadrilateral ABCD

intersect in a point E: Prove that if <) BAD = 60 and AE = 3CE
then the sum of two of the sides of the quadrilateral equals the sum
of the other two.
Solution: Set <) ABD = x and <) CBD = y. From the Low of
Sine's we obtain
AE = sin x  CE = sin y
AB sin(120 + y ; x) BC sin(60 + x ; y)


and AB = sin(120 ; x) . Therefore 3 = AE = sin x: sin(120 ; x) .

 
BC sin(60 ; y) CE sin y: sin(60 ; y)
Hence 3(cos(2y ; 60 ) ; cos 60) = cos(2x ; 120 ) ; cos 120  i.e.
1 ; cos(2x ; 120 ) = 3(1 ; cos(2y ; 60 )) and so sin2(x ; 60) =
3 sin2(y ; 30). Therefore sin(x ; 60) cos 60 =  cos 30 sin(y ; 30)
i.e. sin x ; sin(120 ; x) = (sin y ; sin(60 ; y)): Again fom the
Low of Sine's we get AD ; AB = (CD ; BC ) i.e.
AD + BC = AB + CD or AD + CD = AB + BC:

30
Problem 3. Find the least positive integer n such that there exists
a group of n people such that:
1. There is no group of four every two of which are friends
2. For any choice of k  1 people among which there are no
friends there exists a group of three among the remaining n ; k
every two of which are friends.
Solution: Consider a group of 7 people A1 A2 : : :  A7 such that
Ai i = 1 2 : : :  7 is not friend only with Ai+1 and Ai;1 (we set that
A8 = A1 and A0 = A7). It is easily seen that there are no four every
two of which are friends. Also, for any choice of k  1 people (in
this case k is 1 or 2) every two of which are not friends there exists
a group of three among remaining 7 ; k people every two of which
are friends. Therefore k  7:
We prove that for any group of 6 which satisfy the condition 1)
it is possible to choose a group of k  1 every two of which are not
friends such that among remaining 6 ; k a group of three every two
of which are friends does not exist.
Denote the people by A1 A2 : : :A6: If some of them is friend with
the other 5 (wlog suppose this is A1) it is clear that there are no three
among A2 A3 : : :A5 every two of which are friends. Therefore the
choice of A1 solves the problem.
Suppose one of then is friend with exactly four others (wlog as-
sume A1 is friend with A2 A3 A4 and A5) Then the choice of A1 and
A6 solves the problem.
Therefore each person has 0 1 2 or 3 friends. It is obviuos that
there exists a group of three any two of which are friends (otherwise
the problem is trivial). Assume that the group A1 A2 and A3 has
this property. Wlog A1 and A4 are not friends. Therefore among
A2 A3 A5 and A6 there is a group of three any two of which are

31
friends. If this is A5 and A6 together with one of A2 or A3 then there
exists a person who is friend with at least four others, a contradiction.
Therefore wlog suppose that this group is A2 A3 and A5. Since A1
and A5 are not friends and since among the others there is a group of
three friends we obtain that either A2 or A3 has at least four friends
which is a contradiction.
Therefore k = 7:
Problem 4. Given a right triangle ABC with hypotenuse AB:
A point D distinct from A and C is chosen on the ray AC ! such
that the line through incenter of 4ABC parallel to the bisector of
<) ADB is tangent to the incircle of 4BCD: Prove that AD = BD:

Solution: First we show that C lies on the line segment AD: For,
suppose the contrary, i.e. <) ADB > 90 : Denote the tangential
point of incircle k(I r) of 4BCD and the side BD by P and the
tangential point of k and the line through the center J parallel to
bisector l of <) ADB by T . Since l ? DI we get T 2 DI: Thus,
<) IJT = <) CBD2 =<) IBP and since IT = r = IP we have that
4IJT and 4IBP are congruent. In particular IJ = IB which
implies that <) BJC < 90 a contradiction. Let A0 be such point on
the ray DA! that DA0 = DB and E is the midpoint of A0B: Denote
the tangential point of incircle of 4A0BC by F and its incenter by
J 0. Then we have
EF = jA0E ; A0F j = 21 jA0B ; (A0B + A0C ; BC )j =
= 12 jBC ; A0C j = 12 jBC ; (A0D ; CD)j =
1
2
jBC + CD ; BDj = r:
Therefore the line J 0F parallel to l is tangent to k: This implies
that J 0 2 JF: From the other hand J 0 2 JC and since JC is not
32
parallel to JF (<) ACJ = 45 > 21 <) ADB =<) ADI ) we get J = J 0:
Thus, <) ABJ =<) CBJ =<) A0BJ i.e. A = A0 which completes the
proof.
Problem 5. Find all triples of positive integers (a b c) such that
a3 + b3 + c3 is divisible by a2b b2c and c2a:
Solution: If d = gcd(a b c) it is easy to see that da  db  dc is
 !

also a solution of the problem. Thus, it suces to nd all triples

(a b c) such that gcd(a b c) = 1: Let gcd(a b) = s and suppose
s > 1: If p is a prime divisor of s then p divides a p divides b and
p divides a3 + b3 + c3: Hence p divides c3 and so p divides c: This
is a contradiction to gcd(a b c) = 1: Therefore gcd(a b) = 1 and
by analogy gcd(a c) = gcd(b c) = 1: It follows from a2=(a3 + b3 +
c3) b2=(a3 + b3 + c3) c2=(a3 + b3 + c3) and gcd(a2 b2) = gcd(a2 c2) =
gcd(b2 c2) = 1 that a3 + b3 + c3 is divisible by a2b2c2: In particular
a3 + b3 + c3  a2b2c2: Wlog we may assume that a  b  c: Then
a 2b2
3c3  a3 + b3 + c3  a2b2c2 ) c  3 : Suppose a > 1: Then
a2b2 > b2  b2 + a(a ; b) = a2 ; ab + b2 and the inequality (1) c >
3
a2 ; ab + b2 holds. From the other hand since b  a  2 we have
that b2  2b  a + b and since c > b2 we obtain (2) c > a + b: Now
(1) and (2) give (a ; ab +c2b )(a + b) < 1 () a c+2 b < 1 which
2 2 3 3

is impossible since a3c+2b3 is an integer. Thus, a = 1: In this case we

have that 1+ b3 + c3 is divisible by b2c2: Consider the following cases:
1) if b = c it is easy to see that b = c = 1: Indeed, (1 1 1) is a
solution of the problem.
2) if b = 1 we obtain the same solution.
3) if b = 2 we obtain c = 3: Triple (1 2 3) is a solution of the

33
problem.
Supose now that c > b  3: Since 1 + b3 + c3  b2c2 it follows
that 2c3 > 1+ b3 + c3 and so 2c > b2 or c > b2 : It follows from 2c > b2
2

that 2c > b2 ; b + 1 i.e. (3) b ;cb + 1 < 2: From the other hand
2

when b  5 the inequalities 2c > b4 > b + 1 ) (4) b +c 1 < 12 hold.

2

Multiplying (3) and (4) gives b c+2 1 < 1 which is impossible since
3

this number is an integer. Direct vericstion shows that when b = 3

or b = 4 we get no new solutions. Therefore all solutions of the
problem are triples (k k k) and (k 2k 3k) (and its permutations)
for arbitrary positive integer k:
Problem 6. Given a pack of 52 cards. The following opperations
are allowed:
1. Swap the rst two cards
2. Put the rst card on the last place.
Prove that using these opperations one can order the cards in
arbitrary manner.
Solution: First we show that it is possible to change any two neigh-
bouring cards. Indeed, using 2) we can move the two cards on the
rst two positions. After that apply 1) and again 2) to put all re-
maining cards on their initial positions.
Consider two arbitrary cards ai and ai+j : We can change these
two cards by changing ai and ai+1 then ai and ai+2 and so on, up to
ai and ai+j . After that we change ai+j and ai+j;1 and so on, up to
ai+j and ai+1:
Since we can change any two cards we can order the cards in
arbitrary manner.
34
 L National Mathematics Olympiad
4th round, 19-20 May 2001

Problem 1. Consider the sequence fang such that a0 = 4, a1 = 22

and an ; 6an;1 + an;2 = 0 for n  2. Prove that there exist sequences
fxng and fyng of positive integers such that an = xyn ;+y7 for any
2

n n
n  0.
Solution: Let xn = an +2an;1 , x0 = 3 and yn = an ;2an;1 , y0 = 1.
Then xn = 3xn;1 +4yn;1 and yn = 2xn;1 +3yn;1 . Since an = xn +yn,
it suces to prove that xn + yn = xyn ;+y7 , i.e. x2n = 2yn2 + 7.
2

n n
We prove this by induction. The assertion is obvious for n = 0.
Suppose that x2n;1 = 2yn2;1 + 7. Writing this equality in the form
(3xn;1 + 4yn;1)2 = 2(3yn;1 + 2xn;1)2 + 7 gives x2n = 2yn2 + 7 which
completes the proof.
Problem 2. Given nonisosceles triangle ABC . Denote the tan-
gential points of the inscribed circle k of center O with the sides
AB BC and CA by C1 A1 and B1 respectively. Let AA1 \ k =
A2 BB1 \ k = B2 and let A1A3, B1B3 be bisectors in triangle
A1B1C1 (A3 2 B1C1 B3 2 A1C1). Prove that:
a) A2A3 is bisector of <) B1A2C1
b) if P and Q are the intersecting points of circumcircles of tri-
angle A1A2A3 and triangle B1B2B3 then the point O lies on the line
PQ.
Solution: a) From the Low of Sine's we obtain
AB1 = sin <) B1A1A2  AC1 = sin <) C1A1A2 
AA1 sin 1 AA1 sin 1
35
where 1 =<) A1B1C1 1 = A1C1B1: Since AB1 = AC1 we get
A2B1 = sin <) B1A1A2 = sin 1 = A1B1 = A3B1 
A2C1 sin <) C1A1A2 sin 1 A1C1 A3C1
which implies that A2A3 is the bisector of <) B1A2C1:
b) Let M = AA1\BB1: It follows from MA1 : MA2 = MB1 : MB2
that M lies on the line PQ. Therefore it suces to prove that
OM ? O1O2, where O1 and O2 are circumcenters of 4A1A2A3 and
4B1B2B3. It follows from a) that the diametrically opposite point
of A3 in k1 { the circumcircle of 4A1A2A3, lies on the line B1C1.
Therefore O1 2 B1C1. Moreover <) B1B3A1 =<) CA1A3 =  + 1 .
2
It easily follows now that O1 coinsides with the intersecting point
of B1C1 and BC . Let OO1 \ A1A2 = N and OO2 \ B1B2 = K . It
follows from 4OA1O1 that ON : OO1 = OA21 = r2 and by analogy
OK : OO2 = r2 where r is the radius of k. Since O, N , M and K lie
on the circle k3 of diameter OM we have that the line O1O2 is the
image of k3 by inversion of center O and degree r2, i.e. OO1 ? OM .
Problem 3. For a permutation a1 a2 : : : an of the numbers 1 2 : : : n
it is allowed to change the places of any two consecutive blocks, i.e.
from
a1 : : :  ai ai+1 ai+2 : : : ai+p ai+p+1 ai+p+2 : : :  ai+q  ai+q+1 : : : an
| {z } | {z }
A B
by replacing A and B one can obtain
a1 : : :  ai ai+p+1 ai+p+2 : : :  ai+q  ai+1 ai+2 : : :  ai+p ai+q+1 : : : an.
| {z } | {z }
B A
Find the least number of such changes after which from n n ;
1 : : : 1 one can obtain 1 2 : : : n.
Solution: Call the change of two blocks a move. We shall prove
that the least number of moves such that from n n ; 1 : : : 1 one can
36
obtain 1 2 : : : n is n +2 1 .
 

Consider the number of pairs ai ai+1 such that ai < ai+1. This
number is 0 in the initial permutation n n ; 1 : : : 2 1 and is n ; 1
in the nal permutation 1 2 : : :  n ; 1 n.
First we show that a move changes the number of pairs ai ai+1
such that ai < ai+1 at most by two. For, consider a move of the
blocks a : : : b and c : : : d of the permutation
: : : p a : : :  b c : : :d q : : :
As a result we obtain
: : : p c : : :  d a : : :b q : : :
It is clear that at most three pairs can change the ordering. Suppose
the elements in all three pairs change the ordering, i.e. from p > a,
b > c and d > q we get p < c, d < a and b < q. Adding the
rst three inequalities gives p + b + d > a + c + q and adding the
last three implies p + b + d < a + c + q, a contradiction. Therefore
a move changes the number of pairs ai ai+1 for which ai < ai+1
at most by two. It is easily seen that the rst and the last moves
change this number by one. Therefore if x is desired number we have
2 + 2(x ; 2)  n ; 1 which implies x  n +2 1 . It remaines to nd
 

a sequence of n +2 1 moves such that from n n ; 1 : : : 2 1 we get

 

1 2 : : :  n ; 1 n. Let n be even number, i.e. n = 2k. Number the

positions from right to left by 1 2 : : : n. First change the places of
the blocks from positions 1 2 : : :  k ; 1 and k k + 1. Next, change
the places of the blocks from positions 2 3 : : :  k and k + 1 k + 2.
Third, change the blocks 3 4 : : :  k +1 and k +2 k +3 and so on. On
the k-th step change the blocks from positions k k +1 : : :  2k ; 2 and
2k ; 1 2k. As a result we obtain k + 1 k + 2 : : : 2k 1 2 : : :  k. The
37
last change is 1 2 : : :  k and k + 1 k + 2 : : :  2k. When n is odd, i.e.
n = 2k + 1 the rst change is of the blocks from positions 1 2 : : :  k
and k + 1 k + 2 after that 2 3 : : :  k + 1 and k + 2 k + 3 and so on.
The last k + 1 change is 1 2 : : : k and k + 1 : : :  2k and we obtain
1 2 : : :  2k + 1.
Problem 4. Let n  2 be xed integer. At any point with integer
coordinates (i j ) we write i + j modulo n. Find all pairs (a b) of
positive integers such that the rectangle with vertices (0 0), (a 0),
(a b), (0 b) has the following properties:
1) the remainders 0 1 : : :  n ; 1 written in its interior points
appear equal number of times
2) the remainders 0 1 : : : n ; 1 written on its boundary appear
equal number of times.
Solution: Let pi and si for i = 0 1 : : : n ; 1 be the number of
residues i on the sides and in the interior of the rectangle respectively.
If a0 = a + kn for an integer k then the corresponding numbers for
(a0 b) are p0i = pi + 2k and s0i = si + k(b ; 1).
Therefore if (a b) is a solution then (a + kn b) and by analogy
(a b + ln) are also solutions. Thus, wlog we may suppose that
1  a b  n.
If n = 2 then all possible values of (a b) are
(a b) = (1 1) (1 2) (2 1) (2 2):
Pair (2 2) is not a solution because there is unique interior point
(s0 = 1, s1 = 0) for the rectangle. The remaining three cases give
solutions.
Therefore for n = 2 solutions are all pairs (a b) where either a
or b is an odd number.
Let n > 2 and suppose (a b) is a solution of the problem for
38
which 1  a b < n. In the vertices (0 0) and (a b) are written 0
and a + b modulo n respectively, whereas all residues 1 2 : : : a +
b ; 1 appear on the boundary even number of times. (once on the
boundary (1 0) : : :  (a 0) (a 1) : : : (a b ; 1) and once on the
boundary (0 1) : : : (0 b) (1 b) : : : (a ; 1 b)). If n does not
divide a + b then 0 and the residue of a + b appear odd number of
times whereas at least one of the remaining residues (n > 2) appears
even number of times. Therefore n divides a + b and so a + b = n
or a + b = 2n. When a + b = 2n we have that a = b = n. This pair
is not a solution since the number of interior points is (n ; 1)2 and
is not divisible by n. It remains to consider the case a + b = n. If
a > 1 and b > 1 then the rectangle has interior points. For any such
point (i j ) we have 0 < i < a, 0 < j < b, 0 < i + j < a + b = n and
therefore the residue modulo n is not 0. Therefore a = 1, b = n ; 1
(or a = n ; 1, b = 1) which is a solution. Therefore for n > 2 all
solutions are a = 1+ kn, b = n ; 1+ ln and a = n ; 1+ kn, b = 1+ ln,
where k l = 0 1 2 : : :.
Problem 5. Find all real numbers t for which there exist real
numbers x y z such that
3x2 + 3xz + z2 = 1 3y2 + 3yz + z2 = 4 x2 ; xy + y2 = t:
p
Solution. We shall prove that the answer is t 2 3 ;2 5  1 :
 !

Let x y t  satisfy the conditions of the problem. Consider four

points A B C O in the plane such that AO = x BO = y CO =
pz and <) AOB = 60 <) BOC =<) COA = 150: By the cosine
3 p
theorem it follows that t > 0 AB = t BC = p2  CA = p1 :
3 3
Since <) ACB <<) AOB = 60 and <) BAC <<) BOC = 150  we

39
have
 2  2 p
p1 + p2 ; ( t)2
(1) 3
2 p
3
1 p2 = cos <) ACB > 21 
3 3

p  2  2
p
( t)2 + p13 ; p23
(2) 2p p2 = cos <) BAC > ; 23 :
3
p
These inequalities
p are equivalent to t < 1 and t + t ; 1 > 0 i.e.
3 ; 5
 !

t2 2 1 :
p
Conversely, let t 2 3 ;2 5  1 : Then we can construct a tri-
 !

p
angle ABC such that AB = t BC = p2  CA = p1 : By (1) and
3 3
(2) it follows that <) ACB < 60 and <) ABC <<) BAC < 150 : Let
kA be the circle through B and C such that the arc BC , lying in
the half-plane (with respect to BC ) containing the point A is equal
to 60 : Denote by kB the analogous circle trough A and C: Then it
is easy to see that the second common point O of kA and kB lies
in the interior of 4ABC: (Indeed, assume the contrary. If O lies in
<) ACB then <) AOB =<) AOC + <) BOC = 150 + 150 = 300
which is impossible. If O lies in the opposite angle of <) ACB
then <) AOB =<) AOC + <) BOC = 30 + 30 = 60 which con-
tardicts to the inequalities <) AOB <<) ACB < 60: The other cases
for the position of O can be rejected in the same manner.) Hence
<) AOB = 360 ; <) AOC ; <)pBOC = 360 ; 150 ; 150 = 60:
Set x = AO y = BO z = CO 3: Applying the cosine theorem for
triangles AOB BOC and COA it follows that x y z t satisfy the
conditions of the problem.
40
Problem 6. Given the equation
(p + 2)x2 ; (p + 1)y2 + px + (p + 2)y = 1
where p is xed prime number of the form 4k + 3: Prove that:
a) If (x0 y0) is a solution of the equation where x0 and y0 are positive
integers, then p divides x0
b) The given equation has innitely many solutions (x0 y0) where
x0 and y0 are positive integers.
Solution: a) Set y ; 1 = z and write the equation in the form
(3) x2 = (z ; x)((p + 1)(z + x) + p):
If z ; x and (p + 1)(z + x) + p are relatively prime then they are
perfect squares which is impossible since the second one is of the
form 4k + 3: Let q be a common divisor of these numbers. It follows
from (3) that q=x and so q=z: Since q=(p + 1)(z + x) + p we have q=p
giving q = p which completes the proof.
b) It suces to prove that (3) has innitely many solutions in
positive integers. Let x = px1 and z = pz1: Then x21 = (z1 ; x1)((p +
1)(z1 + x1) + 1) and therefore there exist positive integers a and b
such that z1 ; x1 = a2 x1 = ab and (p + 1)(z1 + x1) + 1 = b2: It
follows now that
(4) (p + 2)b2 ; (p + 1)(a + b)2 = 1:

q q 2k+1 q q
Let p + 2 + p + 1 = mk p + 2 + nk p + 1 for any k =
0 1 : : :  where mk and nk are positive integers. It is obvious that
q q q q
( p + 2 ; p + 1)2k+1 = mk p + 2 ; nk p + 1
and after multiplying we obtain that (p+2)m2k ;(p+1)n2k = 1 i.e. b =
mk and a + b = nk are solutions of (4): Hence, x = pmk (nk ; mk) and
41
z = pnk (nk ;mk ) are solutions of (3): The assertion of b) follows from
the fact that both sequences m1 m2 : : : and n1 n2 : : : are strictly
increasing.

42
 SPRING MATHEMATICAL COMPETITION

1995

Grade 8.
Problem 1. Find all values of a, for which the system
x + 4jyj = jxj
jyj + jx ; aj = 1
has exactly two solutions.
Solution. Let (x y ) be a solution with x  0. Then y = 0 and jx ; aj = 1, i.e. x = a  1. It
is obvious that when a  1 the system has two solutions with x  0, namely (a ; 1 0), (a +1 0).
When ;1  a < 1 the system has only one solution (a + 1 0) with x  0. Let x < 0. Then
jyj = ; x2 and consequently jx ; aj = 1 + x2 . Since jx ; aj  0, then 1 + x2  0 or x  ;2.
We have x ; a = 1 + x2 and x = 2(a + 1), x ; a = ;1 ; x2 . From here x = 23 (a ; 1). From
;2  2(a + 1) < 0 we get;2  a < ;1 and from ;2  23 (a ; 1) < 0 we get ;2  a < 1.
Obviously if (x y ) is a solution of the system with x < 0, then y 6= 0 and therefore (x ;y )
is a solution too. Thus, when a < ;2 the system has no solution. When a = ;2 we get
x = 2(a + 1) = ;2 and x = 32 (a ; 1) = ;2, i.e. the system has exactly two solutions (;2 1).
If 1 > a > ;2, then 2(a + 1) 6= 32 (a ; 1) and the two values of x give solutions, thus we get four
solutions with x < 0.
Therefore the system has two solutions only when a  1, namely (a ; 1 0) and (a + 1 0) and
when a = ;2, namely (;2 1).
Problem 2. Let M be the midpoint of the side BC of the parallelogram ABCD, N be the
common point of AM and BD, while P be the common point of AD and CN . Prove that
a) AP = AD
b) CP = BD i AB = AC .
Solution. a) Consider 4ABC . The lines AM and BD are medians and thus N is the center
of gravity. If Q is the intersection point of CP and AB , then Q is the midpoint of AB . It's easy
to see that 4APQ  = 4BCQ (Figure. 1), from where AP = BC = AD.
b) Let AB = AC . Then 4ABC is isosceles and AM is a median in it. It is easy to
see that 4NBC is isosceles and BN = CN . Analogously NA is a median and an altitude
in 4PND, thus PN = DN , i.e. PC = PN + CN = DN + BN = BD (Figure 2). Let
CP = BD. Trough B we draw a line BF , parallel to CP . Obviously PFBC is a parallelogram
and therefore 6 CPD = 6 BFP . Since CP = BF , then 4DBF is isosceles. Consequently
1
6 NDA = 6 BFP = 6 CPD and DN = NP . A is the midpoint of PD and NA is perpendicular
to AD and BC . We get that AM is perpendicular to BC and we deduce from here that 4ABC
is isosceles, i.e. AB = AC .
Figure 1. Figure 2.

Problem 3. A convex polygon with n sides, n  4, is given. No four vertices of it lie on

one and the same circle.
a) Prove that there exists a circle through 3 vertices of the polygon which contains the
remaining vertices in its interior.
b) Prove that there exists a circle through 3 consecutive vertices of the polygon which contains
the remaining vertices in its interior.
Solution. a) Let AB be a side of the polygon. All the vertices lie in one of the half-planes
with respect to AB . The segment AB is seen under dierent angles from the vertices which are
dierent from A and B . Let C be the vertex from which AB is seen under the smallest angle.
The circle we are looking for is dened by A, B and C .
b) We shall use the following two lemmas.
Lemma 1. Let the segment AB be seen from the point X under the angle  and from the
point Y under the angle  , where 0 <  <  < 90. Then the radius of the circumcircle of
4ABX is greater than the radius of the circumcircle of 4ABY .
Lemma 2. For each convex polygon there exist a side and a vertex from which this side is
seen under an acute angle.
Proof. Let Ai , Ai+1 and Ai+2 be three consecutive vertices of the polygon (Figure 3). At
least one of the angles Ai+1 Ai Ai+2 and Ai Ai+2 Ai+1 is acute. This is enough for the proof.
Figure 3. Figure 4.

Let now M be the set of all pairs, that are consisted of a side and a vertex from which this
side is seen under an acute angle. Let (AB , C ) be an element of M . Consider the circumcircle of
4ABC and let N be the set of all such circles. We denote by k the circle in N with the biggest
radius. Let k be the circumcircle of 4A1A2 Am , where A1 A2 is a side of the polygon (Figure

2
4). We shall show that k is the circle we are looking for. Assume that there exists a vertex Ap
which is outside k. Then 6 A1 Ap A2 < 6 A1Am A2 < 90 and the circumcircle of 4A1A2 Ap is
from N and its radius is greater than the radius of k (Lemma 1). We get a contradiction.
Consider one of the acute angles in 4A1A2 Am . Let  = 6 A2 A1 Am be acute. We shall prove
that Am = A3 , which means that k passes through 3 consecutive vertices of the polygon. Assume
that Am 6= A3. Then A3 is situated in the way which is shown in the Figure 4. If  = 6 A2 A3 Am ,
then  +  > 180, i.e. 180 ;  <  < 90. Thus 6 A2 Am A3 < 90, which means that (A2A3 ,
Am ) is an element of M . Let c be the circumcircle of 4A2 A3Am and X be the intersection point
of c and the segment bisector of A2 Am . Since 6 A2 XAm = 180 ;  <  = 6 A2 A1Am , then the
radius of c is greater than the radius of k (Lemma 1). This is a contradiction.
Grade 9.
Problem 1. Let M be an arbitrary point on the side AB = 1 of the equilateral triangle
ABC . The points P and Q are orthogonal projections of M on AC and BC , while P1 and Q1
are orthogonal projections of P and Q on AB .
a) Prove that P1 Q1 = 43 .
b) Find the position of M for which the segment PQ is with the smallest length.
Solution. a) We have SABC = SACM + SBCM = Figure 5.
1 (AC:MP + BC:MQ) = 1 (MP + MQ) (Figure 5). On
2 p2 p
the other hand SABC = 4 . Thus MP + MQ = 23 .
3
p triangles P1MPp and MQ1Q
Now from the rectangular
we evaluate P1 M = 23 MP and MQ1 = 23 MQ. From
p
here P1 Q1 = P1 M + MQ1 = 23 (MP + MQ) = 43 .
b) The orthogonal projection of the segment PQ on
AB is the segment P1Q1 and thus PQ  P1 Q1 . There-
fore PQ is with minimal length when it is parallel to
P1Q1. The last is true exactly when AP = BQ. We get
that 4AMP  = 4BMQ and hence AM = BM . So, PQ
is minimal when M is the midpoint of AB .
Problem 2. The quadratic function f (x) = ;x2 + 4px ; p + 1 is given. Let S be the area
of the triangle with vertices at the intersection points of the parabola y = f (x) with the x-axis
and the vertex of the same parabola. Find all rational p, for which S is an integer.
Solution. The discriminant of f (x) is D = 4(4p2;p+1) and D > 0 for all real p. Consequently
f (x) has two real roots x1 and x2, i.e. f (x) intersects the x-axis in two dierent points | A
and B . The vertex C of the parabola has coordinates 2p and h = f (2p) = 4p2 ; p + 1 > 0. We
have q q q
AB = jx1 ; x2j = (x1 ; x2 )2 = (x1 + x2 )2 ; 4x1x2 = 2 4p2 ; p + 1:
Now we nd S = S = AB:h = (4p2 ; p +1) 2 : Denote q = 4p2 ; p +1. Since q is rational
3
ABC 2
and q 3 = S 2 is an integer, then q is an integer too. Then S is rational and  S 2 = q is integer,
q q

3
thus Sq is integer too. Therefore q = n2 , where n is a positive integer, i.e. 4p2 ; p + 1 ; n2 = 0.
The quadratic equation (with respect to p) has a rational root exactly when its discriminant
16n2 ; 15 is a square of a rational number. Consequently 16n2 ; 15 = m2 , and we can consider
m to be a positive integer. From the equality (4n ; m)(4n + m) = 15 we get 4n ; m = 1,
4n + m = 15 or 4n ; m = 3, 4n + m = 5. From here n = 2, m = 7 or n = 1, m = 1. The rational
numbers we are looking for are 0 1 41  ; 34 .
Problem 3. Let n be a positive integer and X be a set with n elements. Prove that
a) The number of all subsets of X (X and included) is equal to 2n .
b) There exist 2n;1 subsets of X each pair of which is with common element.
c) There do not exist 2n;1 + 1 subsets of X , each pair of which is with common element.
Solution. a) We use induction with respect to n. The base of the induction is obvious.
Assume that the assertion is true for a set with n ; 1 elements and let X be with n elements.
We can assume that X = f1 2 : : : ng. Let Y = f1 2 : : : n ; 1g. All subsets of X are divided
into two groups: I group | those which do not contain n and II group | those, which contain
n. Both groups have one and the same number of elements because each set of the II group is
obtained from exactly one set of the I group by annexing n. Thus the number of the elements
of X is twice greater than the number of the subsets of the I group. But the subsets of the I
group are exactly the subsets of Y and according to the inductive assumption their number is
2n;1 . Thus the number of the subsets of X is 2n .
b) According to a) the number of the subsets of X from the II group is 2n;1 and each pair
of them has a common element | the number n.
c) If A
X , let A = X n A. All subsets of X are divided into pairs fA Ag and the number of
these pairs is 2n;1 . Now if we have 2n;1 + 1 arbitrary subsets of X , according to the pigeonhole
principle, it is not possible that they are in dierent pairs of the type fA Ag. Consequently
among the given 2n + 1 subsets there exist two pairs of the type fA Ag, which obviously have
no common element.
Grade 10.
Problem 1. Find all values of the real parameters p and q, for which the roots of the
equations x2 ; px ; 1 = 0 and x2 ; qx ; 1 = 0 form (in a suitable order) an arithmetic
progression with four members.
Solution. Denote by x1 , x2 the roots of the equation x2 ; px ; 1 = 0 and by y1 , y2 the
roots of x2 ; qx ; 1 = 0. It is clear that for all p and q the numbers x1 , x2, y1 , y2 are real
and x1 x2 = y1 y2 = ;1. Assume that x1 < 0 < x2 and y1 < 0 < y2 . If four numbers a, b, c
and d form an arithmetic progression then the numbers d c b a form an arithmetic progression
too. So, we can assume that x1 , x2, y1 , y2 in a suitable order form an increasing arithmetic
progression. If x1 < y1 (the case y1 < x1 is analogous) then there are two possibilities:
I. The arithmetic progression is x1, y1, y2, x2. Then x1 + x2 = y1 + y2, from where p = q,
i.e. x1 = y1 , x2 = y2 which is impossible.
II. Theparithmetic progression is x1, y1, x2, y2. Then x2 ; x1 = y2 ; y1, from where
p2 + 4 = q2 + 4, i.e. p2 = q 2 and since p 6= q, then p = ;q 6= 0.
p

2
In the same way we have y1 = x1 +2 x2 = p2 and therefore p4 ; pq 2 ; 1 = 0. Since p = ;q ,
then p2 = 43 and p =  p2 . Hence (p q ) = p2  ; p2 , ; p2  p2 .
  

3 3 3 3 3
4
 Problem 2. Triangle ABC with AB = 22, BC = 19, CA = 13 is given.
a) If M is the center of gravity of 4ABC , prove that AM 2 + CM 2 = BM 2 .
b) Find the locus of points P from the plane of 4ABC , for which AP 2 + CP 2 = BP 2 .
c) Find the minimal and maximal values of BP , if AP 2 + CP 2 = BP 2 .
Solution. a) We have Figure 6.
p
AM = 23 ma = 23  21 2b2 + 2c2 ; a2 = 105
p

CM = 23 mc = 23  21 2a2 + 2b2 ; c2 = 8
p

BM = 32 mb = 23  21 2a2 + 2c2 ; b2 = 13:

p

Hence AM 2 + CM 2 = 105 + 64 = 169 = BM 2 .

b) Let E be the midpoint of AC and D be symmetric
to B with respect to E (Figure 6). We shall prove that the
locus we are looking for is a circle k with center D and radius 26. For an arbitrary point
P we have 4PE 2 = 2PA2 + 2PC 2 ; AC 2 and 4PE 2 = 2PB 2 + 2PD2 ; BD2, from where
2(PA2 + PC 2 ; PB 2 ) = 2PD2 ; (BD2 ; AC 2). But BD = 2BE = 39, i.e. PA2 + PC 2 ; PB 2 =
PD2 ; (26)2. It follows from here that the equality PA2 + PC 2 = PB 2 is equivalent to PD = 26.
c) Let the circle k intersects the line BD at the points M and N . It follows from b) that
BP is minimal when P coincides with M and then BP = BM = 13. BP is maximal when P
coincides with N , which gives BN = 65.
Problem 3. Find the smallest positive integer n, for which there exist n dierent positive
integers a1  a2 : : : an satisfying the conditions:
a) the smallest common multiple of a1  a2 : : : an is 1995
b) for each i j 2 f1 2 : : : ng the numbers ai and aj have a common divisor =
6 1
c) the product a1a2 : : :an is a perfect square and is divisible by 243.
Find all n-ples (a1  a2 : : : an), satisfying a), b) and c).
Solution. Since 1995 = 3:5:7:19 and ai =1995, then for all i
ai = 3 5 7 19
i i i i
( )
where the numbers i , i , i,
i are equal to 0 or 1. There is no ai which is divisible by 9. We
have a1 a2 : : :an = k2 , where k is a positive integer and since 243 = 35 divides k2 , then 36 divides
k2 and since among the numbers ( ) there is no one which is divisible by 9, then n  6. The
numbers ( ), which are divisible by 3 are
3 3:5 3:7 3:19 3:5:7 3:5:19 3:7:19 3:5:7:19 ( )
They are 8 in number. Let n = 6. Then the numbers a1 a2 : : : an are among ( ). It is
easy to see that the product of any 6 numbers from ( ) is not a perfect square. Thus n  7.
Let n = 7. It is not possible that all the numbers a1  a2 : : : an are divisible by 3, because in
such a case 37=k2 and 38 ;= k2 . Therefore 6 numbers from a1 a2 : : : an are divisible by 3, i.e.
they are from ( ) and at least one of them (for example a1) is not divisible by 3. It follows
from a) that among a1  a2 : : : a7 there is at least one which is divisible by 5, at least one which
is divisible by 7 and at least one which is divisible by 19. Since each pair of these numbers
has a common divisor, and 3 ;= a1 , then a1 must be divisible by 5:7:19, i.e. a1 = 5:7:19. At

5
last note that the product of all numbers ( ) is equal to 38 :54:74:194. The only possibility is
a1a2 : : :a7 = 36 :54:74:194. From here a2 a3 : : :a7 = 36:53:73:193 and the only possibility is
fa2 a3 : : : a7g = f3  5 3  7 3  19 3  5  7 3  5  19 3  7  19g:
Therefore n = 7 and
fa1 a2 a3 a4 a5 a6 a7g = f3  5 3  7 3  19 3  5  7 3  5  19 3  7  19 5  7  19g:

Grade 11.
 !
1 2 n
Problem 1. Let an = n2n++11 21 + 22 +    + 2n , n = 1 2 3 : : :. Prove that
a) an+1  an for all n  3
b) the sequence fan g1n=1 is convergent and nd its limit.
Solution. a) We have
 !
1 2 n+1
an+1 = n2n++22 21 + 22 +    + n2 + 1 = 2(nn++21) (an + 1):
From here
2 ; an) ; (an + 1)  n = 1 2 3 : : ::
an+2 ; an+1 = (n + 2) 2((ann+1
+ 1)(n + 2)
Since an > 0 for all n, then if an+1 ; an  0, we have an+2 ; an+1  0. But a3 = 35 and
a4 = 53 , i.e. a4 ; a3 = 0. Hence a5 ; a4  0, a6 ; a5  0, : : :, an+1 ; an  0.
Figure 7. b) The sequence a1  a2 a3 : : : is decreasing when n 
3 and it is bounded (an > 0 for all n). Therefore this
sequence is convergent. Let nlim !1 n
a = a. From the
n + 2
equality an+1 = 2(n + 1) (an +1) after passing to innity
we get a = 12 (a + 1), i.e. a = 1.
Problem 2. The quadrilateral ABCD is inscribed
in a circle with center O. The diagonals AC and BD
intersect each other in the point M , M 6= O. The line
through M which is perpendicular to OM intersects the
sides AB and CD of the quadrilateral ABCD in the
points X and Y , respectively. Prove that AB = CD i
BX = CY .
Solution. If AB = CD, then ABCD is isosceles
trapezoid. Hence OM ?AD and OM ?BC , from where
XY kBC and BX = CY .
Let BX = CY . Denote by P and Q the midpoints of AB and CD, respectively. The quadri-
laterals OPXM and OQY M are inscribed, thus 6 OPM = 6 OXM and 6 OQM = 6 OY M .
We shall prove that 6 OPM = 6 OQM . The triangles ABM and DCM are congruent and
6
 MP and MQ are medians in them. Therefore 4MPB  4MQC and 6 MPB = 6 MQC .
Then 6 OPM = 90 ; 6 MPB = 90 ; 6 MQC = 6 OQM: (Or 6 OPM = 6 MPB ; 90 =
6 MQC ; 90 = 6 OQM .)
Hence 6 OXM = 6 OY M and OX = OY . Therefore 4OXB  = 4OY C . From here 6 OBA =
6 OCD and the isosceles triangles ABO and DCO are equal, i.e. AB = CD

Remark. If X = P and Y = Q, then PB = QC = 1 AB = 1 CD.

Problem 3. Let n be a positive integer and let 2 2

f (x) = xn + (k + 1)xn;1 + (2k + 1)xn;2 +    + ((n ; 1)k + 1)x + nk + 1:

a) Prove that f (1 ; k) = n + 1.
b) Prove that if n  3 and k is an integer (k =
6 0), then the equation f (x) = 0 has no integer
solution.
Solution. a) Since
xn;1 + 2xn;2 +    + (n ; 1)x + n
= (xn;1 +    + x + 1) + (xn;2 +    + x + 1) +    + (x2 + x + 1) + (x + 1) + 1
n;1 3 2
= x ; 1 + x ; 1 +    + x ; 1 + x ; 1 + x ; 1
n
x;1 x;1 x;1 x+1 x;1
then
 !
1 x n+1 ; 1
n ;1 n ;2
x + 2x +    + (n ; 1)x + n = x ; 1 x ; 1 ; (n + 1)
n+1
= x ;(x(n;+1)1)2 x + n
when x 6= 1. Thus
f (x) = xn + xn;1 +    + x + 1 + k(xn;1 + 2xn;2 +    + n)
n+1 n+1
= x ; 1 + k x ; (n + 1)2 x + n 
x;1 (x ; 1)
i.e. n+2 n+1
f (x) = x + (k ; 1)x (;x ;k(1)n2+ 1) + 1]x + kn + 1
when x 6= 1.
From here
f (1 ; k) = ;k(n + 1) + 1](1
k 2
; k) + kn + 1 = n + 1
when 1 ; k 6= 1. If 1 ; k = 1, i.e. k = 0, then f (1) = n + 1.
b) Let n  3 and k be an integer (k 6= 0). Assume that f (a) = 0. Obviously a 6= 0. We have
an + an;1 +    + a + 1 = ;k(an;1 + 2an;2 +    + (n ; 1)a + n):
If a = ;1, the left hand side of the equation is equal to 0 or 1, while the right hand side is
neither 0 nor 1, because jn ; (n ; 1) + (n ; 2) ;   j > 1 when n  3 (k = 6 0). Thus a =6 ;1.
The equation f (a) = 0 can be written in the following way:
(;a ; k + 1)(an;1 + 2an;2 +    + (n ; 1)a + n) = n + 1:
7
 Hence rn (a) = an;1 + 2an;2 +    + (n ; 1)a + n divides n + 1. If a > 1, then rn (a) > n + 1
when n  3 and this is impossible. From a 6= 0 and a 6= ;1 it follows that a  ;2. We shall
prove that the inequality jrn (t)j  n + 2 is satised for all integers t  ;2 and for all n  3
n+1
except t = ;2, n = 3 and t = ;2, n = 4. Since rn (;2) = (;2) + 3n + 2 , then r3(;2) = 3,
9
r4(;2) = ;2 and r5(;2) = 9, i.e. jr5(;2)j  7. From r3(t) = t2 +2t +3 it follows that jr3(t)j  5
when t  ;3.
Now we shall use induction. If jrn (t)j  n + 2 for t  ;2 and n  3, then rn+1 (t) =
t:rn (t)+ n +1 and jrn+1 (t)j  jtj:jrn(t)j; (n +1)  2(n +2) ; n ; 1 = n +3, i.e. jrn+1(t)j  n +3.
Hence jrn (t)j  n + 2 when n  3 and t  ;2 except the cases n = 3 t = ;2 and n = 4 t = ;2.
Thus rn (a) does not divide n + 1 when n  3 and a  ;2, because r3(;2) = 3, r4(;2) = ;2.
For all others n  3 and a  ;2 we have jrn (a)j  n + 2.
Grade 12.
Problem 1. The function f (x) = p1 ; x (x  1) is given. Let F (x) = f (f (x)).
a) Solve the equations f (x) = x and F (x) = x.
b) Solve the inequality F (x) > x.
c) If a0 2 (0 1), prove that the sequence fan g1 n=0 , determined by an = f (an;1 ) for n =
1 2 : : : is convergent and nd its limit.
Solution. The functions f (x) F (x) and F (x) ; x are dened for all x 2 p 0 1].
p
a) The equation f (x) = x, i.e. 1 ; x = x has only one root  = ;1 +2 5 . It is clear that
 2 (0 1) and the roots of the equation F (x) = x are x1 = 0, x2 = 1, x3 = .
b) In 0 ] the function F (x) ; x has a constant sign. The contrary would imply that there
is  2 (0 ) such that F ( ) =  and this contradicts the result from a). Analogously in  1]
the function F (x) ; x has a constant sign. On the other hand 41 2 (0 ), 43 2 ( 1) and
 p!
F 4 ; 4 > 0, F 4 ; 4 < 0. From here F (x) > x i x 2 0 2 5 .
1
 
1 3
 
3 ; 1 +
c) Let a0 = . It follows from a) that an =  for n = 0 1 2 : : : and hence the sequence is
convergent and its limit is . Let now a0 < . Since f 0 (x) = ; p 1 < 0 for all x 2 0 1),
2 1;x
then f (x) is decreasing. From here frac(a0) > frac() = , i.e. a1 > . By induction
a2n 2 (0 ) and a2n+1 2 ( 1) for all n = 0 1 2 : : :. On the other hand it follows from the
result of b) that for all x 2 (0 ) we have F (x) > x, while for x 2 ( 1) we have F (x) < x,
respectively. Also F 0 (x) = f 0 (f (x)):f 0(x) > 0, i.e. F (x) is increasing and hence F (x) 2 (x ) if
x 2 (0 ) and F (x) 2 ( x) if x 2 ( 1). By induction we get
a0 < a2 <    < a2n <    < 
a1 > a3 >    > a2n+1 >    > :
Both sequences are convergent and let their limits be 1 and 2, respectively. We have
F (a2n ) = a2n+2 and F (a2n+1 ) = a2n+3 for n = 0 1 2 : : :. The function F (x) is continuous and
thus F (1 ) = 1 and F (2 ) = 2 . We get 1 = 2 =  because the only solution of F (x) = x
in (0 1) is . Therefore the sequence fan g1
n=0 is convergent and its limit is . The case  < a0
is analogous.

8
 Problem 2. The sides AC and BC of the triangle ABC are diameters of two circles, each
of which touches internally a circle k, which is concentric to the incircle of 4ABC .
a) Prove that AC = BC .
b) If cos 6 BAC = 34 , nd the ratio of the radii of k and the incircle of 4ABC .
Solution. a) Let I be the center of the incircle of 4ABC , N be the common point of
this circle with AC and M be the midpoint of AC . Let r1 be the radius of k and r be the
radius of the incircle of 4ABC . From the condition it follows that IM = 2b ; r1 and from the
2

rectangular 4INM we get MN = 2 ; r1 ; r2 . On the other hand AN = b + 2c ; a , i.e.

b

2

MN = jAM ; AN j = jc ;2 aj . Therefore
(c ; a)2 = (b ; 2r1)2 ; 4r2 : (1)
Analogously
(c ; b)2 = (a ; 2r1)2 ; 4r2 : (2)
Assume that a 6= b. From (1) and (2) we get (b ; a)(2c ; a ; b) = (b ; a)(a + b ; 4r1), i.e.
2c ; a ; b = a + b ; 4r1. From here b ; 2r1 = c ; a and from (1) it follows that r = 0, which is
impossible. Thus a = b.
b) Let rr1 = t and 6 BAC = . Then c = 2b cos  = 23 b and
s

r = 2c tan 2 = b cos  tan 2 = b cos  11 ; cos  = 3pb :

+ cos  4 7

Since b > 2r1, from (1) it follows that t = b ; (c ;2rb) + 4r and substituting c and r we
p
2 2
p
get t = 23 ( 7 ; 2).
Problem 3. n points (n > 4), no three of which are colinear are given in the plane. More
than n triangles are constructed with vertices among these points. Prove that at least two
triangles have exactly one common vertex.
Solution. Assume the contrary and let k be the smallest number for which the assertion is
not true. This means that there are constructed at least (k + 1) triangles using k points. It
follows from the pigeonhole principle that there exists a point A which is a vertex of at least 4
triangles. Let ABC be the rst triangle. At least one of the points B and C is a vertex of the
second triangle, which we denote by ABD. If ACX is the third triangle then X = D. Thus the
forth triangle must contain B or C , which is impossible. Therefore if A and B are vertices of two
triangles then they are vertices of all the four triangles. Let A be a vertex of t triangles, t  4.
These triangles are of the kind ABA1 , ABA2 , : : :, ABAt , where all the points A1  A2 : : : At
are pairwise dierent. Obviously it is not possible to exist a triangle of the type BXY , where
X and Y are points which are dierent from A1 A2 : : : At. Triangles BAi Aj and Ai Aj Am do
not exist too. Hence the points A, B , A1 , A2 , : : :, At are vertices only of the triangles ABA1 ,
ABA2 , : : :, ABAt . In such a way we use t + 2 points and get t triangles. It is not possible that
t + 2 = k, because all triangles are t < k. The number of remaining points is k0 = k ; t ; 2 and
by them there are constructed at least k + 1 ; t > k0 triangles, such that no two of them have

9
exactly one common vertex. The triangles are more than the points k0, and thus k0 > 4. We
have found a number k0 < k for which the assertion is not true. This contradicts the choice of
k.

10
 SPRING MATHEMATICAL COMPETITION

1996

Grade 8
Problem 1. Prove that for all real a 2 (1 2) the area of the gure encountered by the
graphs of the functions y = 1 ; jx ; 1j and y = j2x ; aj is less than 31 .
Solution. Firstly, we shall nd the common points of the given functions. For this purpose
we solve the equation
j2x ; aj = 1 ; j1 ; xj: (1)
a a a
Since 1 < a < 2, then 2 < 1, and we shall consider the cases: x  2 , 2 < x < 1 and x  1.
We have:
1. When x  a2 the equation (1) takes the form a ; 2x = x. Then x = a3 , which satises
(1), because a3 < a2 .
2. When a2 < x < 1 the equation (1) takes the form 2x ; a = x, i.e. x = a, which does not
satisfy (1), because a > 1.
3. When x  1 the equation (1) takes the form 2x ; a = 2 ; x. Then x = a +3 2 , which
satises (1), because a +3 2 > 1 when a > 1.
Thus, the graphs of the two functions have two common Figure 1.
points (Figure 1), the rst of which (denoted by A) has coor-
dinates xA = a3 , yA = a3 , while the second one (denoted by B )
has coordinates xB = a +3 2 , yB = 2 ; a +3 2 = 4 ;3 a . Denote
by C , D and E the points, with coordinates xC = 1, yC = 1
xD = a2 , yD = 0 and xE = 2, yE = 0, respectively.
Then, the gure encountered by the two graphs is the
quadrilateral ACBD and its area S is obtained by subtracting
the areas of the triangles ODA and BDE from the area of the
triangle OEC . Therefore,
S = SOEC ; SODA ; SBDE
= 1 ; 12  OD  yA ; 12
1  DE  y
B

= 1; 1  a  a ; 1  2; a  4;a
2 2 3 2 2 3
1
 = 1 ; 12 1  a2 ; 1  (4 ; a)2
 2 12 
1
=  ;a + 4a ; 2
6  
= 16  2 ; (a ; 2)2 < 13 :
Figure 2. Problem 2. The altitude CD of the rectangle triangle
ABC (6 ACB = 90) is a diameter of the circle k, which meets
the sides AC and BC in E and F , respectively. The intersec-
tion point of the line BE and the circle k, which is dierent
from E , is denoted by M . Let the intersection point of the lines
AC and MF be K , and the intersection point of the lines EF
and BK be P .
a) Prove that the points B , F , M and P are concyclic
b) If the points D, M and P are colinear, nd the angles
A and B of the triangle ABC .
Solution. a) Since 6 ECF = 90, then EF is a diameter of
the circle k and consequently 6 EMF = 90 = 6 BMK . Then
BC and KM are altitudes in 4BEK (Figure 2), which means
that the point F is the altitude center of this triangle. It follows from here that EP ? BK , i.e.
6 BPF = 90.
Thus, 6 BPF = 6 BMF = 90 and consequently the points B , F , M and P lie on a circle
with diameter BF .
Figure 3. b) We have:
6 BDM = 90 ; 6 MDC = 90 ; 6 MEC

= 6 CBE = 6 FBM = 6 FPM:

If the points D, M and P are colinear (Figure 3), then the
equality between the angles 6 BDM and 6 FPM implies that
the lines AB and EF are parallel, i.e. EF and CD are di-
ameters of the circle k, which are perpendicular to each other.
This means that CD is the bisector of 6 ACB . Consequently,
AC = BC and 6 BAC = 6 ABC = 45.
Problem 3. In a state every town is connected with the
nearest town by a straight way. The distances between the
pairs of towns are pairwise dierent. Prove that
a) no two ways have common points
b) every town is connected by ways with at most 5 other towns
c) there is no closed piecewise line, consisted of ways.
Solution. a) Suppose that the ways AC and BD meet each other (Figure 4) and let C be
the nearest town to A, while D be the nearest town to B . Then AC < AD and BD < BC ,
from where AC + BD < AD + BC .
On the other hand, if O is the common point of AC and BD, then AO + OD > AD and
BO + OC > BC , from where AC + BD > AD + BC . This is a contradiction.
b) Let the town X be connected by ways with the towns A and B (Figure 5). Then AB is
the longest side of 4XAB . Indeed, if we assume that for example AX is the longest side, then
2
A should not be the nearest town to X , and X should not be the nearest town to A as well.
Consequently, the way AX should not exist.

Figure 4. Figure 5. Figure 6.

Therefore 6 AXB is the biggest angle in 4XAB , from where 6 AXB > 60. Now, if we
assume that the town X is connected with at least 6 towns, then the sum of the angles at X
would be greater than 6  60 = 360, which is impossible. Thus, every town is connected with
at most 5 other towns.
c) Assume that there exists a closed piecewise line A1A2 : : :An , consisted of ways (Figure
6). The distances A1 An and A1 A2 are dierent. Let A1An < A1A2 . Then A2 is not the nearest
town to A1 and consequently (because the way A1 A2 exists) A1 is the nearest town to A2 . It
follows from here that A1 A2 < A2 A3 .
Proceeding in this way we obtain the chain: A1 An < A1A2 < A2 A3 < : : : < An A1 , which
leads to contradiction.
Grade 9
Problem 1. Find the values of the real parameter b, for which the dierence between the
maximal and the minimal values of the function f (x) = x2 ; 2bx +1 in the interval 0 1] is equal
to 4.
Figure 7. Figure 8. Figure 9.

- - -
b 0 1 0 1 b 0 b 1

Solution. It is clear that the minimal value of the quadratic function f (x) is obtained when
x = b. We shall consider the following three cases:
1. Let b < 0. In this case the function f (x) is increasing in the interval 0 1] (Figure 7)
and the maximal value is f (1) = 2 ; 2b, while the smallest one is f (0) = 1. From the condition
f (1) ; f (0) = 1 ; 2b = 4 we nd b = ; 23 , which is a solution of the problem.
3
 2. Let b > 1. Now the function f (x) is decreasing in the interval 0 1] (Figure 8) and
the maximal value is f (0) = 1, while the minimal one is f (1) = 2 ; 2b. From the condition
f (0) ; f (1) = 2b ; 1 = 4 we nd b = 25 , which is a solution of the problem.
3. Let 0  b  1 (Figure 9). The minimal value of f (x) in the interval 0 1] is f (b) = 1 ; b2,
while the the maximal one is f (0) or f (1). From f (0) ; f (b) = b2 = 4 we nd b = 2, which are
not solutions (because in this case 0  b  1). From f (1) ; f (b) = (1 ; b)2 = 4 we nd b = 3 or
b = ;1, which are not solutions too.
Finally, the answers are: b = ; 23 and b = 52 .
Figure 10. Problem 2. The quadrilateral ABCD is inscribed in a
circle with radius 1, a circle can be inscribed in it and AB =
AD. Prove that:
a) the area of the quadrilateral ABCD does not exceed 2
p b) the inradius of the quadrilateral ABCD does not exceed
2.
2
Solution. We have (Figure 10) AB + CD = AD + BC and
from AB = AD we get: BC = CD. Let AB = AD = x, BC =
CD = y. The triangles ACB and ACD are equal (by SSS),
from where 6 B = 6 D. But ABCD is inscribed in a circle and
consequently 6 B + 6 D = 180. Therefore 6 B = 6 D = 90.
Then AC is diameter of the circumcircle of the quadrilateral and particularly AC = 2 and
x2 + y2 = AC 2 = 4.
Let S , p and r be the area of the quadrilateral, its semiperimeter and inradius, respectively.
a) We have S = SABC + SACD = 2SACB = xy and
q
xy = x2y2  x +2 y = AC2 = 2
2 2 2

Thus, S  2 (and S = 2 only if x = y and then the quadrilateral ABCD is a square).

b) We shall make use of the formula S = pr, which is true for all polygons that can be
inscribed in a circle. We have:
r2 = Sp2 = (x +S y)2 = fracS 2x2 + y2 + 2xy = 4 +S 2S :
2 2 2

Now
p
r  22 () r2  12 () 4 +S 2S  12
2

() S 2 ; S ; 2  0 () (S + 1)(S ; 2)  0
() S  2

p S > 0). The last inequality is true according to a) and consequently the inequality
(because
r  22 is true too.
Problem 3. This problem is the same as problem 3, grade 8.

4
 Grade 10
Problem 1. Find in the plane the locus of points with coordinates (x y), for which there
exists exactly one real number z , satisfying the equality:
xz4 + yz3 ; 2 (x + jyj) z2 + yz + x = 0:
Solution. Let (x y ) be the coordinates of a point from the locus, we are looking for. This
means that there exists exactly one real number z , for which
xz4 + yz3 ; 2 (x + jyj) z2 + yz + x = 0:
If x = 0, the above equality takes the form
 
z yz2 ; 2jyjz + y = 0:
This equality is satised for at least two dierent values of z (for example z = 0 and z = 1,
if y  0 and z = 0 and z = ;1, if y < 0), which shows that the condition of the problem is not
veried. Consequently, the points from the y -axis do not belong to the locus.
Let x 6= 0. Thus, if z satises the given equality, then z 6= 0. It is easy to see that in this
case the number z1 satises the same equality and consequently z = z1 . From here z 2 = 1, i.e.
z = 1.
Case 1. Let z = 1. Then x + y ; 2 (x + jyj) + y + x = 0, from where; y ; jyj = 0. This
shows that y  0. If y = 0 and then the given equality takes the form x z 2 ; 1 2 = 0, and
consequently it is satised also by z = ;1. Thus, we can assume that y > 0. Then,
xz4 + yz3 ; 2(x + y)z2 + yz + x = 0
from where  
(z ; 1)2 xz 2 + (y + 2x)z + x = 0:
We have one of the following possibilities:
(i) The number z = 1 is the only to satisfy the equality xz 2 + (y + 2x)z + x = 0. This is true
if x + y + 2x + x = 0 and D = (y + 2x)2 ; 4x2 = y (y + 4x) = 0, from where y = ;4x and y > 0.
(ii) There is no real z , which satisfy the equality xz 2 + (y + 2x)z + x = 0. Then, D =
(y + 2x)2 ; 4x2 = y (y + 4x) < 0, from where y < ;4x and y > 0.
Therefore, every time when y  ;4x and y > 0, the point (x y ) belongs to the locus.
Case 2. Let z = ;1. Analogously to the previous case we nd that y < 0 and then the
given equality takes the form
 
(z + 1)2 xz 2 + (y ; 2x)z + x = 0:
We have the following possibilities:
(i) the number z = ;1 is double root of the equation
xz2 + (y ; 2x)z + x = 0: ()
(ii) the equation (*) has no real root.
By computing we deduce that here y  4x and y < 0.
5
 Finally, the locus (Figure 11) consists of the internal as well as of the boundary points of the
angle which is dened by the graphs of the linear functions y = 4x and y = ;4x when x < 0,
without the points from the negative part of the x { axis.
Problem 2. In the triangle ABC , ha and hb are the altitudes from A and B respectively,
`c is the internal bisector of 6 ACB, while O, I and H are the circumcenter, the incenter and
the altitude center, respectively. Prove that if `c + `c = 2, then OI = IH .
ha hb
Solution. Let BC = a, AC = b, 6 ACB =  , CL = `c Figure 12.
1
(Figure 12). Since SABC = SALC + SBLC , then 2  `c a sin 2 + 

1  ` b sin  = 1  ab sin  , from where ` = 2ab cos 2 . But
2 c 2 2 c a+b 
2a cos
ha = b sin  and hb = a sin  . Then h`c + h`c = (a + b) sin2  +
a b
2b cos 2 1 a + b  1 , which
=  =
(a + b) sin  sin 2 a + b a + b sin 2
shows that sin 2 = 12 , i.e. 2 = 30 , because 2 < 90. Conse-
quently,  = 60.
Let CL meet the circumcircle of 4ABC in M . Then
OM ? AB. But CH ? AB, and therefore 6 OMC = 6 MCH . On the other hand OM = OC =
R (R is the circumradius). Consequently, 6 OMC = 6 OCM = 6 MCH . Let H2 be the foot of
the altitude from B . From 4HCH2 we have CH = CH sin 
2
(because 6 H2 HC = 90 ; 6 ACH
= 6 CAB = ). But CH2 = a cos  , i.e. CH = a  cos  = 2R cos  . Since  = 60 , then
sin 
CH = R.
We consider 4COI and 4CHI . Since the point I lies on CM , then 6 OCI = 6 HCI . Also,
CO = R = CH and consequently 4COI = 4CHI . Thus, OI = IH .
Problem 3. Let A1, A2, : : :, An (n  4) be n points in the plane, no 3 of which are colinear.
a) Prove that there is at most one point As , such that all triangles As Ai Aj (i j = 1 2 : : : n)
are acute.
b) Let among A1, A2 , : : :, An be a point, which is a vertex of an acute triangles only. We
consider the angles dened by the given points. Denote by Nk the number of the acute angles
6 Ai Ak Aj (i j = 1 2 : : : n) for which the point Ak is their vertex. Find the minimal value of
Nk .
Solution. a) Assume that there is more than one point with the given property and let A
and B be such two points, while X and Y be any two of the remaining points. There are two
possibilities for the points A, B , X and Y :
(i) A, B , X and Y dene a convex quadrilateral. Because the sum of its internal angles is
360 , then at least one of these angles is not acute. Consequently, at least one of the triangles


with vertex A or B is not acute.

(ii) One of the points A, B , X and Y is inside the triangle dened by the other three. Then,
the sum of the three angles with a vertex inside is 360. Consequently, in this case a triangle
with vertex in A or B is not acute again.
The contradictions show that it is not possible to exist more than one point which is a vertex

6
of acute triangles only.
b) Let A1 be the point which is a vertex of acute triangles only. We consider the angles
6 Ak A1 As (Figure 13). Let A2 A1 An be the biggest one. Because all angles with vertex A1 are
acute, the points A3 , A4 , : : :, An;1 are inside the acute angle 6 A2 A1An . We can assume that
the points are enumerate in such a way that 6 A2A1 Ak < 6 A2A1 Ak+1 for k = 3 4 : : : n.
Consider 6 Ai Ak Aj , where 2  i < k < j  n. Figure 13.
We assume that Ak is internal for 4A1 Ai Aj . Then
6A1Ak Ai + 6 A1Ak Aj > 180
and at least one of the angles 6 A1 Ak Ai and 6 A1Ak Aj will not be
acute, which is a contradiction.
Therefore, the points A1 , Ai , Ak , Aj dene a convex quadrilat-
eral. If we assume that 6 Ai Ak Aj  90, then there exists an angle
of this quadrilateral with vertex A1 , Ai or Aj which is  90. This
is impossible.
Therefore, 6 Ai Ak Aj > 90 , and it is clear that 6 Ai Ak Aj < 90 when i j < k or i j > k.
Particularly, it follows from here that no of the angles Ai Ak Aj is right. The number of all
angles with vertex Ak is equal to (n ; 1)(2 n ; 2) . If Tk is the number of the obtuse angles with
vertex Ak , then Nk = (n ; 1)(2 n ; 2) ; Tk . Thus, Nk is minimal when Tk is maximal.
It is easy to see that T1 = T2 = Tn = 0. Let 3  k  n ; 1. It is clear that the number of
the points Ai for which 2  i < k is k ; 2, and the number of the points Aj for which k < j  n
is n ; k. Then Tk = (k ; 2)(n ; k) for k = 3 4 : : : n ; 1.
But (k ; 2)(n ; k)  (n ;4 2) , and the equality is reached when k ; 2 = n ; k, i.e. when
2

k = n +2 2 . We have two possibilities:

(i) The number n is even. Then n +2 2 is integer and consequently the maximal value of Tk
is T n+2 = (n ; 2)2 . Therefore, the minimal value of N in this case is N n+2 = (n ; 1)(n ; 2 ;
2 2 k 2 2
(n ; 2) = n(n ; 2) .
2

2 4
(ii) The number n is odd. Then, the nearest integers to n +2 2 are n +2 1 and n +2 3 . It is
easy to see that now the maximal value of Tk is T n+1 = T n+3 =
n + 1 ; 2 n ; n + 1  =
2 2 2 2
(n ; 1)(n ; 3) . Consequently, the minimal value of N in this case is
4 k

= (n ; 1)(n ; 2) ; (n ; 1)(n ; 3) = (n ; 1) :
2
N n+1
2
= N n+3
2 2 4 4

Grade 11
Problem 1. Find the values of the real parameter a, for which the inequality x6 ; 6x5 +
12x4 + ax3 + 12x2 ; 6x + 1  0 is satised for all real x.

7
 Solution. When x = 0 the given inequality
 is satised
 for all a. Thus, it is enough to nd
3 1
such a, that x + 3 ; 6 x + 2 + 12 x + + a  0 for all x > 0 and x3 + 13 ;
2 1 1
 x  x x x
6 x +
2 1 1
+ 12 x + +a  0 for all x < 0.
x2 x
Denote t = x + x1 . It is clear that x > 0 () t  2 and x < 0 () t  ;2. But
x2 + x12 = t2 ; 2 and x3 + x13 = t3 ; 3t.
   
We consider the function f (t) = t3 ; 3t ; 6 t2 ; 2 + 12t+ = t3 ; 6t2 + 9t + 12 + a.
The problem is reduced to nd such a, that f (t)  0 for all t  2 and f (t)  0 for all t  ;2
simultaneously.
Figure 14. But f 0 (t) = 3t2 ; 12t + 9 = 3(t ; 1)(t ; 3), from where it is
easy to obtain that (Figure 14) f (t)  0 for all t  2 i f (3)  0
and f (t)  0 for all t  ;2 i f (;2)  0. Since f (3) = a + 12
and f (;2) = a ; 38, we nd for a: a  ;12 and a  38. Finally,
;12  a  38. _
Problem 2. The point D lies on the arc BC of the circumcircle
of 4ABC which does not contain the point A and D 6= B , D 6= C .
On the rays BD! and CD! there are taken points E and F , such that BE = AC and CF = AB .
Let M be the midpoint of the segment EF .
a) Prove that 6 BMC is right. _
b) Find the locus of the points M , when D describes the arc BC .
Figure 15. Figure 16.

Solution. Denote by , ,  the angles corresponding to the vertexes A, B , C of 4ABC ,

b, c the lengths of the sides BC , CA, AB, respectively. Let 6 BCD = '. It follows
and by a, _
from D 2BC that 0 < ' < , i.e. ' 2 (0 ).
;;! 1 ;! ;! 1 ;! ;! ;!
a) We have (Figure 15) BM = 2  BF + BE = 2  CF + BE + BC and ; ;!
CM =
1  ;! ;!  1  ;! ;! ;!  ;;! ;;!
2 CF + CE = 2 CF + BE ; BC . Then we nd for the scalar product BM  CM :
 !
;;! ;;! 1 ;! ;!2 ;!2
BM  CM = 4  CF + BE ; BC

8
 1 ;!2 ;!2 ;! ;! 2 
= 4  CF + BE + 2  CF  BE ; a
 
= 14  c2 + b2 + 2bc cos( ; ) ; a2
 
= 14  c2 + b2 ; 2bc cos  ; a2 = 0:

(We have used that jBE j = jAC j = b, jCF j = jAB j = c and 6 (;! BE ;!
CF ) = 6 BDC =  ; .)
;;! ;;!
Consequently, BM ? CM , i.e. the angle 6 BMC is right.
b) It is clear that the points M and A are in dierent semiplanes with respect to the line BC .
According to a) 6 BMC is a right angle and consequently if the point M is from the locus, then
M lies on a semicircle k with diameter BC , k and A are in dierent semiplanes
 with respect to
the line BC . Let BCM =  . Then CM = a cos  (because BMC = 2 ) and
6 6

;;! ;!
CM  CB = a(a cos )  cos  = (a cos )2:
On the other hand by the sine theorem we nd:

CM  CB = 21  ;!
;;! ;!
CB + ;!BE + ;!CF  ;! CB
 
= 12  a2 + ab cos ( ; ( ; ')) + ac cos '
1 b c 
= 2 a 1 ; a cos( ; ') + a cos '
2

1 sin  cos ' ; sin cos( ; ') 

= 2a 1 +
2
sin 
1
= 2 a2 1 + sin(  ; ' ) ; sin( ;  + ')
2 sin 
1 + '2
= a (1 + cos( + ')) = a cos
2 :
2 2
Consequently, cos2  = cos2 +2 ' . But +2 ' < +2  < 2 and therefore  = +2 ' , i.e.
6 BCM =
+ ' and 6 BCM 2  +  . In addition we have 6 CBM =  ; 6 BCM =
2 2 2 2
 +  ; ' and 6 CBM 2    +  .
2 2 2
Let K and N be points from the semicircle k, for which 6 BCK = 2 and 6 CBN = 2 . Then
angleBCN = 2 ; 2 =  +2 and 6 CBK = 2 ; 2 =  +2  .
_
It follows from the above considerations that when the point D describes the arc BC ,
then CM ! describes the interior of 6 (CK ! CN !), while BM ! describes the interior of
_
6 (BK !  BN ! ). Consequently, the locus, we are looking for is the arc NK from the semi-
circle k (Figure 16).
Problem 3. is the same as problem 3, grade 10.

9
Spring mathematics
tournament|1997
Problem 8'1. Given the equation jx ; aj + 15 = 6jx + 2j, where a
is a real parameter.
(a) Prove that for any value of a the equation has exactly two
distinct roots x1 and x2.
(b) Prove that jx1 ; x2j  6 and nd all values of a for which
jx1 ; x2j = 6.
Solution: (a) When x  ;2, the equation is equivalent to jx ;
aj = ;6x ; 27, which has a solution only if ;6x ; 27  0, i. e., if
x  ; 29 . Considering the cases of both x  a and x  a shows that
if x  ;2, the given equation has a unique root x1, and if x > ;2,
it has a unique root x2:
8 a ; 27 9 8 3 ; a a  1
><  a < ;2 ><
7
x1 = > a + 27 9 x2 = > a +5 3  a > 12
:;  a  ; :
5 2 7 2
1
(b) It follows from (a) that
8 156 ; 12a 9
>>
>< 35  a < ; 2
jx1 ; x2j = > 6 ; 29  a  12
>> 12a + 204 1
:  a >
35 2
It remains to be seen that 156 ; 12a > 6 when a < ; 9 and
35 2
12a + 204 > 6 when a > 1 . Therefore jx ; x j  6 and equal-
35 2 1 2

ity obtains only when a 2 ; 92  12 ].

Problem 8'2. Let O be the intersecting point of the diagonals

of the convex quadrilateral ABCD and let 6 DAC = 6 DBC . The
midpoints of AB and CD are respectively M and N and P and Q
are points on AD and BC respectively such that OP ? AD and
OQ ? BC . Prove that MN ? PQ.

Solution: Denote by E and F the midpoints of AO and BO. Then

4PEM = 4MFQ since MF = 21 AO = PE , QF = 21 OB = ME
and 6 MEP = 6 MEO + 6 OEP = 6 MEO + 26 DAC = 6 MFO +
26 DBC = 6 MFQ (MFOE is a parallelogram). Therefore MP =
MQ. The case when E and F are interior points for 6 PMQ is
treated similarly (prove that there are no other possibilities). By
analogy we conclude that NP = NQ. Hence M and N lie on the
axis of symmetry of PQ and so MN ? PQ.

2
Problem 8'.3. Find all natural numbers n such that there exists
an integer number x for which 499(1997n + 1) = x2+x.

Solution: Let n be a solution of the problem. Then (2x + 1)2 =

1996 1997n + 1997. If n = 1 we get (2x + 1)2 = 19972 and so
2x + 1 =
1997. Therefore x = 998 and x = ;999 satisfy the
conditions of the problem. Let n  2. Now (2x + 1)2 is divisible
by 1997, which is a prime number, and so (2x + 1)2 is divisible by
19972 . But this is impossible, since 1996 1997n +1997 is not divisible
by 19972 when n  2. The only solution is n = 1.

Problem 8.1. Find all values of the real parameter m such that
the equation (x2 ; 2mx ; 4(m2 + 1))(x2 ; 4x ; 2m(m2 + 1)) = 0 has
exactly three distinct roots.

Solution: Suppose m satises the conditions of the problem and

the equations
x2 ; 2mx ; 4(m2 + 1) = 0
x2 ; 4x ; 2m(m2 + 1) = 0
share the root x0. After subtracting we get (2m ; 4)x0 = (2m ;
4)(m2 +1) and so x0 = m2 +1 (note that if m = 2, the two equations
coincide). Substituting x0 in any of the equations gives the equation
(m2 + 1)(m2 ; 2m ; 3) = 0 with roots m = ;1 and m = 3. Direct
verication shows that the condition is satised only for m = 3.
Let now (1) and (2) share no roots. Since D1 = 4 + 5m2 > 0
(1) always has two distinct roots and therefore (2) should have equal

3
roots. Thus D2 = 4 + 2m(m2 + 1) = 0 and so m = ;1. But this
case has already been considered. Thus we determine that m = 3.

Problem 8.2. The area of the equilateral triangle ABC is 7. Points

M and N are chosen respectively on AB and AC so that AN = BM .
Denote by O the intersecting point of the straight lines BN and CM .
The area of BOC is 2.
(a) Prove that MB : AB = 1 : 3 or MB : AB = 2 : 3.
(b) Find 6 AOB .

Solution: (a) Denote MB AB = x. Therefore SABN = 7x = SBMC

and so SBOM = 7x ; 2 and SAMON = SBOC = 2. Further SCON =
; 7x) ,
7 ; 2 ; 2 ; (7x ; 2) = 5 ; 7x, SANO = 1 ;x x SCNO = x(51 ; x
1 ; x 1 ; x
SAMO = x SBOM = x (7x ; 2). It follows from SAMON =
SANO + SAMO that 2 = x(51 ;; 7x) + 1 ; x (7x ; 2), and thus 9x2 ;
x x
9x + 2 = 0: The roots of the above equation are x1 = 31 and x2 = 23 .
(b) Since 4ABN  = 4BMC, we get 6 BOM = 6 BCM +
6 CBO = 6 MBO + 6 CBO = 60 . Further 6 MAN + 6 MON = 180
and therefore the quadrilateral AMON is inscribed in a circle. Let
MB = 1 , i. e., AM = 2BM = 2AN . Denote by Q the mid-
AB 3
point of AM . Triangle AQN is isosceles and has an angle equal to
60 , so it is equilateral. Therefore Q is the circumcentre of AMON
and 6 AOM = 6 ANM = 90 . Thus 6 AOB = 150 . Similarly, if
4
MB = 2 , i. e., 2AM = MB = AN , we get 6 AMN = 6 AON = 90,
AB 3
so 6 AOB = 90.

Problem 8.3. Given n points, n  5, in the plane such that

no three lie on a line. John and Peter play the following game:
On his turn each of them draws a segment between any two points
which are not connected. The winner is the one after whose move
every point is an end of at least one segment. If John is to begin the
game, nd the values of n for which he can always win no matter
how Peter plays.

Solution: Call a point isolated if it is not an end of a segment.

John wins exactly when there are 1 or 2 isolated points before his
last move. Peter is forced to reach the above only if before his move
there are exactly 3 isolated points and any of the remaining n ; 3
points are connected by a segment. Indeed, if there are at least 4
isolated points he could connect one of them with a non-isolated
point. If the isolated points are 3 but not all of the remainig n ; 3
points are connected he could draw a missing segment. Since the
number of segments with ends in n ; 3 points is (n ; 3)(2 n ; 4) ,
we determine that John wins only when (n ; 3)(2 n ; 4) is an odd
integer number. This is true when n is of the form 4k + 1 or 4k + 2.

Problem 9.1. Let f (x) = x2 ; 2ax ; a2 ; 43 where a is real param-

eter. Find all values of a such that the inequality jf (x)j  1 holds
for any x in the interval 0 1].

5
Solution: Let M and m be the maximum and minimum values
of f (x) in the interval 0 1]. Then the condition of the problem is
equivalent to M  1 and m  ;1. There are three cases to consider.
Case 1: a 2 0 1]. Then m = f (a) = ;2a2 ; 43 and M = f (0) =
;a2 ; 43 or M = f (1) = ;a2 ; 2a + 14 . It follows from m  ;1 and
p p
M  1 that a 2 ; 42  42 ] and a 2 (;1 ; 32 ]  ; 12  1). In this
p
case the solution is a 2 0 42 ].
Case 2: a < 0. Now m = f (0) and M = f (1). From m  ;1
and M  1 we get that a 2 ; 12  12 ]: Therefore a 2 ; 21  0).
Case 3: a > 1. Now m = f (1) and M = f (0). It follows from
m  ;1 and M  1 that a 2 ; 25  12 ] which is a contradiction with
a > 1. p
1
Thus the solution is a 2 ; 2  42 ].

Problem 9.2. Let I and G be the incentre and the centre of

4ABC with sides AB = c, BC = a, CA = b.
(a) Prove that (if a > b) the area of CIG equals (a ; b)r where r
6
is the inradius of ABC:
(b) If a = c +1 and b = c ; 1, prove that the segment IG is parallel
to AB and nd its length.

6
Solution: (a) We shall use the usual notation for a triangle. Let
CL and CM be respectively the bisector and the median from C .
It follows from CG = 32 CM that SCIG = 32 SCIM . Thus SCIM =
SCLM ; SILM = LM2 hc ; LM2 r = LM 2 (hc ; r). We nd from
AL + BL = c and BL AL = b that AL = bc and so LM = AM ;
a a+b
c bc c ( a ;
AL = 2 ; a + b = 2(a + b) . Also, hc ; r = 2cS ; r = 2pr
b )
c ;r =
r(2p ; c) = r(a + b) . Therefore S = 2 S = 2 c(a ; b)
CIG
c c 3 CIM 3 4(a + b)
r(a + b) = (a ; b)r .
c 6
(b) The distances from I and G to AB are respectively r and
hc . Hence r = S = chc = chc = hc and so IGkAB . Therefore
3 p 2p 3c 3
the altitude from C of triangle CIG equals 23 hc = 2r. Thus SCIG =
IG r. On the other hand SCIG = (a ;6 b)r = r3 and so IG = 13 .

Problem 9.3. Let n  n be an even number and A be a subset

of f1 2 : : :  ng. Consider the sums of the form "1x1 + "2x2 + "3x3,
where x1 x2 x3 are integer numbers in A (not necessarily distinct),
"1 "2 "3 (at least one of which is not 0) belong to f;1 0 1g and
none of the elements of A appears with coecients 1 and ;1 in any
of the sums. Call A a `free' set if n divides none of the above sums.
(a) Construct a `free' set having  n4 ] elements (x] is the least in-
teger number less than or equal to x).
7
 (b) Prove that no set of  n4 ] + 1 elements is `free'.

Solution: (a) The set A = f1 3 : : :  2 n4 ] ; 1g is `free' and has

 n4 ] elements (prove it!).
(b) Let n = 4m and suppose that A f1 2 : : :  ng is a `free' set
having  n4 ] + 1 = m + 1 elements. Without loss of generality assume
that A f1 2 : : :  2mg (since we can replace x 2 A by n ; x). We
shall show that there exist two elements of A whose sum is equal
to another element of A. Indeed, let a1 < a2 < < am+1 be the
elements of A and consider the set B = fa1 + ai : i = 1 2 : : :  m +
1g. There are 2m + 2 integer numbers a1 a2 : : : am+1 2a1 a1 +
a2 : : :  a1+am+1 from AB and they lie in the interval a1 a1+am+1],
which contains exactly am+1 + 1  2m + 1 integer numbers. This
gives ai = a1 + aj for some i j . But then a1 + aj ; ai = 0, which
is impossible. The case n = 4m + 2 is settled in a similar fashion.
Notice that 2m + 1 cannot be an element of a `free' set.

Problem 10.1. Find the least natural 3number a such that the
equation cos (a ; x) ; 2 cos (a ; x)+cos 2a cos( x
2 x +
2a 3
 )+2 = 0
has a root.

Solution: The roots of the our equations are the common roots of
cos (x ; a) = 1 and cos 32x
a cos( x +  ) + 1 = 0. The roots of
2a 3
the rst one are x = a + 2n n = 0
1
2 : : : and the roots of the
second one are x = 2a(k ; 1 ) k = 0
1 : : : : Therefore a = 6n
3 6k ; 5
8
for some integer numbers n and k. It is easy to see now that the
least natural number with the required property is a = 6.

Problem 10.2. Point F lies on the base AB of a trapezoid ABCD

and is such that DF = CF . Let E be the intersecting point of AC
and BD and O1 and O2 are circumcentres of ADF and FBC respec-
tively. Prove that the straight lines FE and O1O2 are orthogonal.

Solution: Let k1 and k2 be circles with centres O1 and O2 and let

the intersecting points of the two circles be points P and Q. It is well
known that PQ ? O1O2. On the other hand, if L is an arbitrary
point and two lines through L intersect k1 and k2 in points A, B
and C , D respectively. Then L 2 PQ if and only if LA LB =
LC LD. Let k1 and k2 be the circumscribed circles of 4AFD
and 4FBC and let G be the intersecting point of FE with CD.
Denote by C1 and D1 those points on DC for which AD1jjCF and
BC1jjDF , i. e., such that the quadrilaterals AFCD1 and BFDC1
are parallelograms. Using that FD = FC we get 6 CFB = 6 FCD =
6 FDC = 180 ; 6 BC1C . This means that F , B , C and C1 lie on
a circle and so the line DC intersects k2 in C and C1. By analogy
6 AFD = 6 FDC = 6 FCD = 6 AD1D and so line DC meets k1
in points D and D1. In accordance with the initial notes FE is
perpendicular to OO1 if and only if GC GC1 = GD GD1 . It follows
from 4GCE  4FAE , 4GDE  4FBE and 4DCE  4BAE
that GC =
AF EA AB
CE = DC and GD = DE = DC . Thus GC =
BF EB AB
DC AF and GD = BF DC . On the other hand GC = jDC ;
AB AB 1 1

DGj = jBF ;DGj = BF j1; DG BF j = BF j1 ; DC j = BF jAB ;DC j,

AB AB
9
GD1 = jCD1 ; CGj = jAF ; CGj = AF j1 ; CG
AF j = AF j 1 ; DC j =
AB
AF jAB ;DC j. Therefore GC GC = DC AF BF jAB ;DC j =
AB 1
AB 2
GD GD1 .

Problem 10.3. Find all natural numbers n for which a convex n-

gon can be partitioned into triangles through its diagonals in such a
way that there is an even number of diagonals from each vertex. (If
there is a vertex with no digonals through it, assume that there is
an even number (zero) of diagonals from this vertex).

Solution: It is easy to see by induction that if an n-gon is par-

titioned into triangles through d non-intersecting diagonals then
n = d + 3. Let n be a natural number and A1A2 An is a convex
n-gon which can be partitioned into triangles through d diagonals
in a way that there is an even number of diagonals through each
vertex. Since n = 3 is a solution we may assume that n  4.
It is clear that at least one side of each triangle is a diagonal of the
n-gon. We say that a triangle is of type tk , (k = 1 2 3) if exactly k of
its sides are diagonals. Denote by xk the number of triangles of type
tk . It is easy to see that 2x1 + x2 = n = d +3 abd x1 +2x2 +3x3 = 2d.
It follows now that x1 = x3 + 2, so x1 > 0. Therefore there exists
a triangle two of whose sides are sides of the n-gon. Let that be
Aj;1Aj Aj+1. Diagonal Aj;1Aj+1 is a side of another triangle|e. g.,
Aj;1Aj+1As. Assume that Aj;1As or Aj+1As is a side of the n-gon.
If it is Aj;1As then s = j ; 2. It follows now that there are no
diagonals from Aj;1 distinct from Aj;1Aj+1 because such a diagonal
intersects Aj+1As. This contradicts the premise that there is an even

10
number of diagonals from each vertex. Therefore both Aj;1As and
Aj+1As are diagonals so Aj;1Aj+1As is of type t3. Hence there is a
triangle of type t3 adjacent to each triangle of type t1. If distinct
triangles of type t1 are adjacent to distinct triangles of type t3 then
x1  x3 = x1 ; 2 < x1, a contradiction. Therefore there are at
least two triangles of type t1 adjacent to one and the same triangle
of type t3. Without loss of generality assume these are the triangles
A1AnAn;1 and An;1An;2An;3 . Consider the polygon A1A2 An;3.
Obviously the diagonals partition this polygon into triangles and
there is an even number of diagonals through each vertex.
Conversely, if the polygon A1A2 An;3 can be partitioned in the
required way, then adding the vertices An;2, An;1 , An and diagonals
An;3An;1 and A1An;1 shows that the same is true for the polygon
A1A2 An.
Therefore a natural number n  6 is a solution if and only if
n ; 3 is a solution. It is easy to see that n = 3 is a solution, whereas
n = 4 and n = 5 are not. Thus all natural numbers satisfying the
conditions of the problem are n = 3k, k = 1 2 : : :.

Problem 11.1. For any real number b denote by f (b) the maximal
value of j sin x + 3 + 2sin x + bj. Find the minimal value of f (b).

Solution: Substitute t = sin x and g(t) = t + 3 +2 t + b. Since

g(t) is an increasing function in the interval ;1 1], it follows that
f (b) = max(jg(;1)j jg(1)j) = max(jbj jb + 23 j). Now from the graph
of the function f (b) we conclude that min f (b) = f (; 34 ) = 43 .
11
 Problem 11.2. A convex quadrilateral ABCD is such that 6 DAB =
6ABC = 6 BCD. Let H and O be respectively the orthocentre and
the circumcentre of 4ABC . Prove that H , O and D lie on a line.

Solution: Let 6 CAB = , 6 ABC = , 6 BCA =

. Note that
 < and
< . There are three cases to consider: < 90,
= 90 and > 90 . Suppose rst that < 90 . Then O and
H are interior points for 4ABC and 6 ACO = 6 CAO = 6 HCB =
6 HAB = 90 ; . Therefore O is an interior point for 4HAC and
6 HAO = ;
= 6 ACD, 6 HCO = ;  = 6 CAD, 6 HAD =
6 HCD = 2 ; 90 . It follows from the Sine Theorem for 4AHD,

4CHD and 4ACD that sin sin 6 AHD = AD , sin 6 HCD = HD ,

6 HAD HD sin 6 CHD CD
sin 6 CAD = CD . By multiplying the above equalities we get
sin 6 ACD AD
sin 6 AHD sin 6 HCO sin 6 CAO = sin 6 CHD sin 6 HAO sin 6 ACO:
It follows now from Ceva's Theorem that AO, CO and HD intersect
in a point and so H , O and D lie on a line. In the case of = 90
we obtain that H  B , O is the midpoint of AC and AHCD is a
rectangle. Therefore H , O and D lie on a line. Finally, let > 90.
In this case B and O are interior points for 4AHC and 4ADC
respectively. Similarly to the case < 90 we get that the points H ,
O and D lie on a line.

Problem 11.3. For any natural number n  3 denote by m(n)

the maximum number of points which can be placed inside or on
the outline of a regular n-gon with side 1 in a way that the distance
between any two of them is greater than 1. Find all n for which
m(n) = n ; 1.
12
Solution: We prove rst that m(n) = n ; 1 for n = 4 5 6. Let n
be one of the above numbers and let B1 B2  Bn be points satis-
fying the conditions of the problem for the regular n-gon A1A2 An
of side 1 and centre O. It is obvious that OBi  1 and therefore
no three points O, Bi, Bj , 1  i = 6 j  n lie on a line. Further-
more at least one of the angles OBi Bj is less than 2n  90 and
it follows from the Cosine Theorem that BiBj2  OBi2 + OBj2 ;
2OBi OBj cos 2n = Bi0Bj0 2 where Bi0 and Bj0 are points on the seg-
ments OA1 and OA2 such that OBi0 = OBi , Bj0 = OBj . When
n = 4 5 6 the greatest side in 4OA1A2 is A1A2 = 1 and therefore
BiBj  Bi0Bj0  1, which is a contradiction. Thus m(n)  n ; 1
if n = 4 5 6. It is easily seen that for these n there exist n ; 1
points on the outline of a regular n-gon A1A2 An with the re-
quired property. Therefore m(n) = n ; 1 if n = 4 5 6. We shall
prove now that if n  7 then m(n)  n. Let Bi 2 AiAi+1 be points
such that AiBi = 2i;1 ", 1  i  n ; 1 where 0 < " < n1;1 is arbi-
2
trary chosen. From n  7 we obtain cos 6 A1A2A3 = ; cos 2 < ; 1 :
n 2
It follows now that B1B22 > (1 ; ")2 +4"2 +2"(1 ; ")2 = 1+3"2 > 1:
Similarly BiBi+1 > 1 when 1  i  n ; 2. Further, it is clear
that B1Bn;1 > A1An = 1. Since OA1 = OA2 = = OAn > 1,
AiAj > 1 when ji ; j j  2 and Bi ! Ai when " ! 0, it follows
that we can make " so small that OBi > 1 when 1  i  n ; 1 and
BiBj > 1 when ji ; j j  2. Then the points B1 B2  Bn;1  O
satisfy the conditions of the problem and so m(n)  n when n  7.
Since it is obvious that m(3) = 1, we come to the conclusion that
m(n) = n ; 1 only when n = 4 5 6.

13
Spring mathematics
tournament { 1998
Problem 8'.1. Find all values of the real parameter a such that
the inequalities jx + 1j + j2 ; xj < a and 6x 5;a 5;a 8+ 5 < ; 12 are
equivalent.

Solution: We begin with the rst inequality. jx + 1j + j2 ; xj < a.

If x < ;1, it is equivalent to ;x ; 1 ; x + 2 < a or x > 1 ;2 a , i. e.,
1 ; a < x < ;1, which has a solution if 1 ; a < ;1, i. e., if a > 3.
2 2
If x 2 ;1 2], then jx + 1j + j2 ; xj = x + 1 + 2 ; x = 3 and the
above equation has a solution only when a > 3. Finally, if x > 2,
then x + 1 + x ; 2 < a or x < a +2 1 , i. e., 2 < x < a +2 1 , which
has a solution when a > 3. Therefore when a  3, the inequality
has no solution, and when a > 3, the solutions form the interval
1 ; a 1 + a .
2 2

1
 Let us rewrite the second inequality in the form 66xx+;55aa;+115 <
0. Its solutions form either the interval 11 ;6 5a 5a 6; 5 or the


interval 5a 6; 5 11 ;6 5a , depending on which of the two numbers



5a ; 5 and 11 ; 5a is greater.
6 6
Therefore the two inequalities are equivalent if
1 ; a = 11 ; 5a  1 + a = 5a ; 5
2 6 2 6
or
1 ; a = 5a ; 5  1 + a = 11 ; 5a :
2 6 2 6
The rst pair is satis ed by a = 4, the second one by a = 1. But if
a = 1, the rst inequality has no solution, whereas the second one
does. Thus the only solution is a = 4.

Problem 8'.2. Let M and N be the midpoints of the sides BC

and AC and BH (H 2 BC ) the altitude in 4ABC . The straight
line perpendicular to the bisector of HMN intersects the line AC
6

in point P such that HP = 1 (AB + BC ) and HMN = 45 deg.

2
6

a.) Prove that 4ABC is isosceles.

b.) Find the area of 4ABC if HM = 1.

Solution: a) Since HM is a median to the hypothenuse BC of

the right triangle 4BHC and MN is a middle segment in 4ABC ,
it follows that HM = 21 BC (if H  C , again HM = 12 BC ) and
2
MN = 21 AB . Therefore HP = HM + MN . Denote HMN = . 6

There are two cases to consider for the points H N and P : 1. N

lies between H and P 2. H lies between N and P .
1. Let us nd a point K
on the extension of HM such
that MK = MN . Then
HP = HK and HKP =
90 deg ;  . The statement
6

2
of the problem implies that
MP is the bisector of NMK 6

(external to 4HMN ), i. e.,

NMP = KMP . There-
fore 4PNM 
6 6

= 4PKM
and PNM = PKM =
90 deg ; 2 . On the other hand,
6 6

PNM = 45 deg +, i. e.,

6

90 deg ; 2 = 45 deg +,

whence  = 30 deg and HNM = 105 deg. Also MHB = 60 deg
6 6

and since HM = MB , it follows that HMB = 60 deg. Thus

6

6BMN = 105 deg = ANM , i. e., ABMN is a isosceles trapezoid

6

AN = BM . Therefore AC = BC .

3
 2. Let H lie between N
and P and let K be a point
such that MK = MN and
HP = HK . It follows from
the isoscelesness of 4KHP
that HPK = HKP = 2 .
6 6

Since MP is the bisector

of NMK in the isosceles
6

4NMK , it follows that

MP is the axis of symmetry
of NK . Thus NP = KP
and 4NMP  = 4KMP .
Therefore PNM = 2 and
6

since 4NMP , it follows that

 + 2 + 45 deg = 180 deg.
and hence  = 90 deg. Therefore MH ? AC and since BH ? AC ,
we get that H coincides with C and BCA = 90 deg, ABC =
6 6

6 BAC = 45 deg and so AC = BC .

b) 1. When  = 30 deg : ACB = 30 deg, HM = HB = 1,
6

AC = BC = 2. Then SABC = 21 AC BH = 1.
2. When  = 90 deg : AC ? BC , AC = BC = 2BH = 2.
Then SABC = 12 AC BC = 2.

Problem 8'.3. (Problem for the UBM award)

Is it possible to nd 100 straight lines in the plane such that there

4
are exactly 1998 intersecting points?

Solution: Consider 99 lines such that 73 of them are parallel and

the remaining 26 lines pass through a single point and intersect all
73 parallel lines. Then the total number of intersecting points is
73 26 + 1 = 1899. Choose the last line in such a way that it
intersects all lines and does not pass through any of the points. Now
there are 1899 + 99 = 1998 intersecting points.

Problem 8.1. The graph of a linear function is parallel to the

graph of y = 45 x + 95
4 , passing through M (;1 ;25), and intersects
the coordinate axes Ox and Oy in A and B correspondingly.
(a) Find the coordinates of A and B .
(b) Consider the unity grid in the plane. Find the number of
squares containing points of AB (in their interiors).

Solution: (a) The graph of a linear function is parallel to the

graph of y = 45 x + 95 4 when the linear function is of the form y =
5 x+b. From the condition that M belongs to its graph we determine
4
;25 = ; 54 + b, whence b = ; 954 . The coordiantes of A and B are
respectively the solutions of the systems
 
 y = 5 x ; 95  y = 5 x ; 95
 4 4 and  4 4 
y=0 x=0
5
whence we get A(19 0) and B 0 ; 95

4 .
(b) The coordinates of the points on AB satisfy the conditions

 y = 5 x ; 95
(1)  4 4 :
0  x  19
Find the number of points with integer coordinates lying on the
segment AB , i. e., the number of integer solutions of (1). It follows
from y = 45 x ; 954 that y = x ; 23 + x ;4 3 and if x and y are
integer, then x ;4 3 = t is integer. Conversely, if t is integer, then
x = 4t + 3 and y = 5t ; 20 are integer. Since 0  x  19, we nd
0  4t + 3  19, i. e., ; 34  t  4. This observation implies that the
points with integer coordinates lying on AB satisfy the conditions
x = 4t + 3, y = 5t ; 20 for t = 0 1 2 3 4. Therefore there are 5
such points.
Further: the segment AB exits a square and enters another one
only when it intersects a line of the grid. The number of such inter-
sections is 23 + 18 = 41. But when AB passes through a knot (i. e.,
through a point with integer coordinates), the passage from a square
to another one involves crossing two lines. This happens 4 times (not
counting A). Therefore the required number is 41 ; 4 + 1 = 38.

Problem 8.2. Let l1 and l2 be the loci of the centroid G and the
incentre I of the right triangle ABC whose hypotenuse AB is a given
segment of length c.
(a) Find l1 and l2.
6
 (b) Find the area of 4ABC when the length of GI is minimal.

Solution: (a) The vertex C of 4ABC could be placed in either

semiplane cwith respect to AB . The centroid G lies on a circle k1
of radius 6 and centred at the midpoint O of segment AB . The
incentre I lies on one of the two arcs which are the locus of the
points X such that AXB = 135 deg. One of the arcs is part of a
6

circle k2 with centre Q and radius R and the other one is part of a
0 0

circle k2 with centre Q and the same radius and Q and Q lie in
00 00 0 00

dierent semiplanes with respect to AB .

Conversely: Let G be a point on k1 distinct from the intersecting
points M and N of k1 and AB . Then the vertex C of 4ABC is
uniquely determined by CO = 2c . Since AO = BO = CO it follows
that 4ABC is a right triangle. Choose that of the points Q and Q 0 00

which does not lie in the same semiplane as C does with respect to
AB . Without loss of generality this is Q . Denote the intersecting
0

point of the bisector of BAC and k2 by I . We shall prove that

6
0

I is the incentre 4ABC . Since ABI = 45 deg ; BAI , we get

6 6

2 ABI = 90 deg ;2 BAI = ABC .

6 6 6

Therefore BI is the bisector of ABC . 6

Let I be a point of k2 such that AIB = 135 deg. Then the

0 6

vertex C is uniquely determined by: AI is the bisector of BAC , 6

BI is the bisector of ABC and C lies in one and the same semiplane
6

with I with respect to AB . Since BAI + ABI = 45 6

c deg, it follows
6

that 4ABC is a right triangle. Therefore CO = 2 and hence the

intersecting point G of CO and k1 is the centroid of 4ABC .
These observations imply that l1 is the circle k1 without M and

7
N , and l2 consists of two arcs k2 and k2 such that if X 2 k2 or
0 00 0

X 2 k2 , then AXB = 135 deg.

00
6

(b) We shall show that R > 23 c and it will follow (how?) that
k1 lies in the interior of both k2 and k2 . Let P be the intersecting
0 00

point of OQ and k1 which lies between O and Q .

0 0

Since PQ = 3c = 6c + 6c = 0

OP + ON and 4PNO, it fol-

lows that OP + ON > PN
and so PQ > PN . Hence 0

PNQ > PQ N . Therefore

6
0
6
0

PNQ + 45 deg > PQ N +

6 0
6
0

45 deg, so ANQ > AQ N . 6 0 6 0

It follows for 4AQ N that AQ > AN , i. e.,R > 32 c. Let G 2 l1 and

0 0

I 2 l2 be such that Q , O, G and I lie in a line in this order. We

0

shall show that GI is the least possible segment.

Case 1: G1 2 l1, I1 2 l2 and

O lie on a line. Then G1I1 >
GI . Indeed, it follows from
4Q I1O that OQ + OI1 >
0 0

Q I1, i. e., OQ + OG1 + G1I1 >

0 0

R = OQ + OG + GI and so
0

G1 I1 > GI .

8
 Case 2: G2 2 l1 , I2 2 l2
and O are not on a line. Let
OI2 intersect k1 in a point G1.
Then G2I2 > G1 I2. Indeed,
it follows from 4OG2I2 that
OG2 + G2I2 > OI2. But OI2 =
OG1 + G1I2 = OG2 + G1I2 and
therefore G2I2 > G1I2. But
from case 1 we get that G1I2 >
GI and so G2I2 > GI .
Thus 4ABC having the required property is such that Q , O, G
0

and I lie on a straight line. Similarly, Q determines a triangle equal

00

to the rst one. 2Since OQ ? AB , 4ABC is an isosceles triangle

0

and so SABC = c4 .

Problem 8.3. (Problem for the UBM award)

Given n points on a circle such that no three chords with ends
in the given points intersect in a point. Prove that there exists n
2 ; 3n + 4
such that there are n chords with ends in the given points
2
partitioning the interior of the circle into 1998 regions.

Solution: First we shall prove the following Lemma: The number

of
 regions
! into which the interior of a circle is divided by drawing all
n chords with ends in n given points, provided no three chords
2  !  !
intersect in a point, is n4 + n2 + 1:

9
 Use induction by n. It is easy to see that the above formula holds
for n = 2. Suppose it is true for some n. To obtain the result we
count how many new regions are added when a new point appears on
the circle. It is easily seen that if a chord intersects t other chords,
then it `adds' t + 1 new regions. Therefore the new regions are:
nX1
(k(n ; k ; 1) + 1):
;

k=0

Given that 1+2+ + n ; 1 = (n ;2 1)n and 12 +22 + +(n ; 1)2 =

(n ; 1)n(2n ; 1) , we easily obtain that the above sum equals
6
 !  !  !  ! !
n+1 + n+1 +1; n + n +1 
4 2 4 2
which completes the proof.
We show now that if n = 17 it is possible to draw n ; 32n + 4 =
2

121 chords such that there are 1998 regions. Let us draw all chords
with ends in 16 of the given points (there are 120 such chords). !
It follows then that the interior of the circle is divided into 164 +
 !
16 + 1 = 1941 regions. Draw a chord connecting the 17th point
2
with one of the rst 16 in a way that there are 8 and 7 points on the
two sides of the drawn chord. This chord intersects 8 7 = 56 chords
and therefore there are 57 new regions. Therefore the total number
of regions is 1941 + 57 = 1998.

10
Problem 9.1. Find all parameters a such that the inequality jax2;
3x ; 4j  5 ; 3x holds for any x 2 ;1 1].

Solution: Observe that the inequality is equivalent to the system


 ax2 ; 9  0
 ax2 ; 6x + 1
0:

It follows from the second inequality that a > 0, because if a  0

then ax2 ; 6x + 1
0 is not true for x = 1. Further, the second
inequality gives x2  a9 . Since this inequality is true for x 2 ;1 1],
we get 1  a9 ) a  9. Let D = 9 ; a be the discriminant of
ax2 ; 6x + 1. There are two cases to consider:
1) D > 0. The solution of the second inequality is
x 2 (;1 x1] x2 +1), where x1 < x2 are the roots of ax2;6x+1 =
0. Therefore ;1 1]  (;1 x1] or ;1 1]  x2 +1).
1.1) ;1 1]  (;1 x1]. It follows from the above that
 
 a( ; 1) 2
; 6(; 1) + 1
0  a
;7 
 a
;7
 a 12 ; 6 1 + 1
0 ()  3  a
5 ()  a
5 
 x1 + x2  a<3
2 >1
  > 1
a
which is impossible.
1.2) ;1 1]  x2 +1). Therefore
 
 a ( ; 1) 2 ; 6(;1) + 1
0
 a
;7
 a: 1 2
; 6 :1 + 1
0 ()  a
5 
x + x  3
2 < ;1  < ;1
 1 2
a
11
 which is also impossible.
2) D  0 ) the inequality ax2 ; 6x + 1
0 holds for any
real value of x. It then follows that 9 ; a  0 () a
9. Therefore
the solution is a = 9.

Problem 9.2. The quadrangle ABCD is inscribed in a circle. The

tangents to the circle passing through A and C intersect at the
point P . If PA2 = PD PB and P does not lie on DB , prove
that the intersecting point of AC and BD is the midpoint of AC .

Solution:
Let E be the second in-
tersecting point of PB and
the circle. Then PA2 =
PE PB ) PD = PE .
Hence APD = EPC . (If
6 6

O is the centre of the cir-

cle, the above follows from
the similarity of 4ODP
and 4OEP ). Therefore
4ADP  4PCB be-
cause APD = BPC and
AP = DP , so AD =
6 6

BP CP BC
AP : (1)
BP
Also, 4APB  4DCP because APB = DPC = APC ;
6 6 6

AP = BP and so AB = BP : (2)
APD and DP CP DC CP
6

12
 From (1) and (2) we get BC AB AD = AP BP = 1. On the
DC BP CP
other hand,
SABD = 21 AB AD sin DAB = AB AD = 1
6

SCBD 12 BC DC sin DCB BC DC

6

( DAB = 180 deg ; DCB ):

6 6

Therefore SABD = SCBD , i. e., the diagonal BD halves AC .

Problem 9.3. (Problem for the UBM award)

See problem 8.3.

Problem 10.1. Find all values of the real parameter a such that
the inequality x4 + ax3 + (a + 3)x2 + ax + 1 > 0 holds for all real
values of x.

Solution: If x = 0, the inequality holds for any a2. Suppose x 6= 0.

Dividing both sides of the inequality by x2 gives x + ax + a + 3 +
a + 1 > 0 or x + 1 2 + a x + 1 + a + 1 > 0. Substitute
x x2 1 x x
t = x + x . If x > 0, it is true that t
2, and if x < 0, it is true that
t  ;2. Therefore we want to nd all a such that the inequality
t2 + at + a + 1 > 0 holds for any t 2 (;1 ;2] 2 +1). Denote
f (t) = t2 + at + a + 1. The discriminant of f (t) is D = a2 ; 4a ; 4.
p p
Let D < 0, i. e., a 2 (2 ; 2 2 2 + 2 2). Then the inequality
f (t) > 0 holds for any real t, in particular for t 2 (;1 ;2] 2 +1).
p p
Let D
0, i. e., a 2 (;1 2 ; 2] 2 + 2 +1). Then the
inequality f (t) > 0 holds for any t 2 (;1 ;2] 2 +1) if and
13
only if
;2 < ; a2 < 2



f (2) > 0

f (;2) > 0
p p
The solutions of this system when a 2 (;1 2 ; 2] 2 + 2 +1)
are
p p
a 2 ; 53 2 ; 2 2 . So a 2 ; 35 2 + 2 2 .
i 

Problem 10.2. A quadrangle with perpendicular diagonals AC

and BD is inscribed in a circle with centre O and radius 1.
a.) Calculate the sum of the squares of the sides of the quadrangle.
b.) Find the area of ABCD if a circle with centre I is inscribed in
it and OI = p1 .
3

Solution: a.) Denote BAC = , CAD =  . It follows from

6 6

the Sine Theorem for 4ABC and 4ABD that BC = 2 sin , AD =

2 sin ABD = 2 sin(90 deg ;) = 2 cos . Analogously CD = 2 sin  ,
6

AB = 2 cos  . Therefore AB 2 + BC 2 + CD2 + DA2 = 4 sin2  +

4 cos2  + 4 sin2  + 4 cos2  = 8.
b.) The statement of the problem implies that AB + CD =
AD + BC or 2 cos  +2 sin  = 2 cos  +2 sin , i. e. sin( +45 deg) =
sin(+45 deg). Therefore  =  or  = 90 deg ;. We shall consider
only the case  =  (the case  = 90 deg ; is settled in a similar
fashion). Now AD = AB , CD = BC and ABC = ADC = 90 deg. 6 6

The points O and I lie on AC and when 

45 deg, I lies on
14
the segment AO. In 4ACD (DI is a bisector) determine AI =
2 cos  . Consequently OI = AO ; AI = sin  ; cos  = p1 .
sin  + cos  sin  + cos  3
1 ; sin 2  1
Raising both sides to the 2nd power gives 1 + sin 2 = 3 . Hence
sin 2 = 12 . It then follows that the area of ABCD is equal to
S = 2SABC = AB BC = 4 cos  sin  = 2 sin 2 = 1.

Problem 10.3. (Problem for the Atanas Radev award)

Find all natural numbers n such that: If a and b are natural
numbers and n divides a2b + 1, then n divides a2 + b.

Solution: Obviously the condition holds for n = 1 and suppose

n
2. Let a be a natural number coprime to n. It follows from
Bezout's Theorem that there exists a natural number b such that
a2b + 1 is divisible by n. Further, n divides a2 + b and since a4 ; 1 =
a2(a2 + b) ; (a2b + 1), it follows that n divides a4 ; 1.
Let n = 2k, where k is an odd number and 
0. Suppose
k
3. Then n and k ; 2 are coprime and therefore 2 k divides
(k ; 2)4 ; 1, so k divides 15. Hence n is of the form n = 2 3 5 ,
where 
0 0    1. Consequently n divides 114 ; 1 and we
conclude that   4.
It is easy to see now that n divides 24 3 5 = 240.
Conversely, let n be a divisor of 240. Then n satis es the con-
dition of the problem. Indeed, if 3 divides a2b + 1, then 3 divides
both a2 ; 1 and a2 + 1 and therefore 3 divides a2 + b. Similarly, if
5 divides a2b + 1, then 5 divides a2 + b. If we can show that the

15
same property holds for 2 4 8 and 16, we will be done. Assume 2k
divides a2b + 1, where 1  k  3. Then a is an odd integer number
and therefore a2 ; 1 is divisible by 8. Consequently b + 1 is divisible
by 2k and thus a2 + b is divisible by 2k as well. Assume 16 divides
a2b +1. Then a is an odd integer number and it is easy to verify that
a2 is congruent to 1 or 9 modulo 16. Further, b should be congruent
to 15 or 7 modulo 16 correspondingly and again a2 + b is divisible
by 16.
The required numbers are all divisors of 240.

Problem 11.1.
a.) Let p be a positive real parameter. Find the least values of
the functions f (x) = x + xp and g(x) = x + xp2 in the interval
(0 +1).
Let a1 ap2 a3 be positive real numbers. Prove that 3(a1 +
b.) p
a1a2 + 3 a1a2a3)  4(a1 + a2 + a3):

p (x ; pp)(x + pp)
Solution: a.) Since f (x) = 1 ; x2 =
0

xp2 , we get
that the functionpf (x) decreases in the interval (0 p) and increases
in the interval ( p p +1). pTherefore the minimal value of f (x) in
(0 +1) equals f ( p) = 2 p. Analogously it follows from g (px) = 0

1; 2xp3 that the least value of g(x) in (0 +1) equals g( 3 2p) = 3 22p .
q 3

b.) After substituting x = a1 y = pa1a2 z = p3 a1a2a3 our

inequality becomes 0  13 x + 43 yx ; y + 34 zy2 ; z. It follows from a.)
2 3

16
 4 1 4 y 2 1 4 z 3 1
that 3 y  3 x + 3 x and 3 y + 3 y2
z. Therefore 3 x + 3 x ; 4 y 2

y + 43 zy2 ; z
13 y + 43 zy2 ; z
0, which completes the proof. Note
3 3

that equality occurs when x = 2y = 4z, i. e., a1 = 4a2 = 16a3. Note

further that b.) could also be solved by the following inequalities:
p p
a1 + pa1a2 + p3 a1a2a3 = a1 + 12 a14a2 + 41 3 a14a216a3 
 a1 + 41 (a1 + 4a2) + 121 (a1 + 4a2 + 16a3) = 34 (a1 + a2 + a3):

Problem 11.2. Let I and r are the incentre and inradius of 4ABC ,
and N is the midpoint of the median through C . Prove that if
r = CN ; IN , then AC = BC or ACB = 90 deg :
6

Solution: Use the standart notation for the elements of 4ABC

and apply the formula for a median in a triangle. Since IN and
IM are medians in 4CIM and 4AIB , respectively, we get IN 2 =
1 (2CI 2 + 2MI 2 ; CM 2) = = 1 (2CI 2 + AI 2 + BI 2 ; 1 AB 2 ; CM 2).
4 4 2
Therefore
CN 2 ; IN 2 = 14 (2CM 2 + 21 AB 2 ; 2CI 2 ; AI 2 ; BI 2) =
1 (a2 + b2 ; 2(p ; c)2 ; 2r2 ; (p ; a)2 ; r2 ; (p ; b)2 ; r2) = (p ; c)c ; r2:
4 2
It follows from the statement of the problem that IN 2 = (CN ;
r) , so CN 2 ; IN 2 + r2 = 2CN r. It follows from what we have
2

proved above that 2CM r = (p ; c)c. Taking the square of both

17
sides of this equality and using the formul 4CM 2 = 2a2 + 2b2 ; c2
and r2 = (p ; a)(p ; p
b)(p ; c) , we get (2a2 +2b2 ; c2)(p ; a)(p ; b) =
p(p;c)c . After some simple calculations the above equality becomes
2

(a2 + b2 ; c2)(a ; b)2 = 0. Therefore a = b, i. e., AC = BC or

a2 + b2 = c2, i. e., ACB = 90 deg.
6

Problem 11.3. (Problem for the Atanas Radev award)

See problem 10.3.

18
Spring mathematics
tournament|Kazanl^ak,
30 March{1 April 1999
Problem 8'1. Given an inequality jx ; 1j < ax, where a is a real
parameter:
a) Solve the inequality.
b) Find all values of a such that the inequality has exactly two
integer solutions.
Chavdar Lozanov, Kiril Bankov, Teodosi Vitanov

Solution: a) I. Let x  1. Then the inequality is equivalent to

x ; 1 < ax () (1 ; a)x < 1.

1. 1 ; a > 0 a < 1 =) x < 1 ;1 a  1 ;1 a > 1 () a > 0.

Therefore 0 < a < 1 1  x < 1 ;1 a .
1
 2. 1 ; a = 0 =) a = 1 =) 0:x < 1. Therefore a = 1 x  1.
3. 1 ; a < 0 =) 1 < a =) x  1. Therefore 1 < a x  1.
II. Let x < 1. Then the inequality is equivalent to 1 ; x <
ax () 1 < (a + 1)x.

1. a +1 > 0 () a > ;1 a +1 1 < x < 1 a +1 1 < 1 () a > 0.

Therefore a > 0 a +1 1 < x < 1, when ;1 < a  0 no solution
exists.
2. a + 1 = 0 () a = ;1 =) 1 < 0  x, so no solution exists.
3. a + 1 < 0 () a < ;1 =) x < a +1 1 < 0. Therefore
a < ;1 x < a +1 1 .

So when a < ;1, then x < a+1 1

for ;1  a  0 no solution
exists when 0 < a < 1, then a+1 < x < 1;1 a when 1  a, then
1

1  x.
b) It follows from a) that the inequality could have two integer
solutions only if 0 < a < 1. Since in this case 0 < a +1 1 < 1 < 1 ;1 a ,
we nd that there are exactly two integer solutions if and only if
2 < 1 ;1 a  3:

Therefore the answer is 12 < a  32 .

2
Problem 8'2. Let M be the midpoint of the side BC of 4ABC
and 6 CAB = 45 6 ABC = 30.
a) Find 6 AMC .
 BC .
b) Prove that AM = AB2AC
Chavdar Lozanov

Solution: a) Draw CH ? AB . Now 6 ACH = 45 = 6 CAH and

6 HCB = 60 . For 4ACH it is true that AH = HC . Further it
follows from 4CHB that CH = 12 CB = HM . Therefore AH =
180 ; 6 AHM
HM , so MAH = AMH =
6 6
2 . Note that 6 CHM =
60 and 6 AHM = 90 + 60 = 150 . Therefore 6 AMH = 15. We
obtain that 6 AMC = 6 HMC ; 6 AMH = 60 ; 15 = 45 .
b) Let S be the area of 4ABC . We know that S = AB  CH =
2
AB  CB . Since AM is a median, it follows that S S
AMC = . If
4 2
S AM  CP
CP ? AM , then 2 = 2 . But 6 CAM = 6 CAB ; 6 MAB =
45 ; 15 = 30 , and thus CP = AC . Therefore S = AM  AC .
2 2 4
AB  CB AM  AC AB  BC
Now 4 = 2 , and we obtain AM = 2AC .
Problem 8'3. Consider all points in the plane whose coordinates
(x y) in an orthogonal coordinate system are integer numbers and
1  x  19 1  y  4. Each point is painted green, red or blue.

3
Prove that there exists a rectangle with sides parallel to the coordi-
nate axes whose vertices are all of the same colour.
Kiril Bankov

Solution: Since the number of coloured points is 4  19 = 76 and

there are three dierent colours, it follows that there are at least 26
points of the same colour (say blue). Denote by p1  p2 : : : p19 the
lines parallel to the ordinate axis and passing through the points
(1 0) (2 0) : : :  (19 0), respectively. Let n1 n2 : : : n19 be the num-
ber of blue points on the lines p1 p2 : : :  p19. It is clear that 0 
ni  4 for i = 1 2 : : :  19. Without loss of generality assume
n1  n2      n19. Since n1 + n2 +    + n19 = 26, we obtain
that n1  2.
1. Let n1 = 4. Then the remaining 22 blue points lie on 18 lines
and so n2  2. Therefore there are two blue points on each of
the lines p1 and p2 all having the same ordinates. These points
form the required rectangle.
2. Let n1 = 3. Then the remaining 23 blue points lie on 18 lines
and therefore n2  2.
a) If n2 = 3, then there is a blue rectangle with vertices on
p1 and p2.
b) Let n2 = 2. Then n3 = n4 = n5 = n6 = 2. The number
of ways of choosing two blue points on p1 is 3 and the
corresponding number for each of p2 p3 : : :  p6 is 1. The
total number of blue pairs is 3+5  1 = 8, which is greater
that 6|the number of ways of choosing two horizontal
lines out of 4 lines. Therefore there exists a blue rectangle.
4
 3. Let n1 = 2. Then n2 = n3 = : : : = n7 = 2. Since 7 > 6, we
apply the same reasoning as in 2b).

Problem 8.1. Find all rational numbers a such that j4a ; 2j  1

and A = 4a ; 1 is integer. Ivan Tonov
27a4

Solution: It follows from j4a ; 2j  1 that 41  a  34 . Also, it is

clear that when a  14 , then A  0 and A = 0 only if a = 41 . Let
k be a positive integer such that A = k. Then 27a4 ; 4al + l = 0,
where l = 1 . Multiply the above equality by 3 and write it in the
k
following way:
81a4 ; 18a2 + 1 + 18a2 ; 12al + 3l ; 1 = 0 ()
(9a2 ; 1)2 + 2(3a ; 1)2 ; 12a(l ; 1) + 3(l ; 1) = 0
so (9a2 ; 1)2 + 2(3a ; 1)2 + 3(l ; 1)(1 ; 4a) = 0. Therefore 3(l ;
1)(1 ; 4a)  0, which is possible (recall a > 41 ) only if l  1 or k  1.
But since k is a positive integer, it follows that k = 1 and a = 13 .
Therefore the required values are a = 14 and a = 31 .

Problem 8.2. Given a 4ABC . Let M be the midpoint of AB ,

6 CAB = 15 and 6 ABC = 30.
a) Find 6 ACM .
5
  BC .
b) Prove that CM = AB2AC
Chavdar Lozanov

Solution: a) Let AH ? BC . Then 6 HAB = 60 and AH =

AB = HM . It follows from 6 HAC = 6 HAB ; 6 CAB = 45
2  ; 6 MHC
that AH = HC . Thus HM = HC and HCM = 6 180 .
2
But 6 MHC = 6 AHB ; 6 AHM = 90 ; 60 = 30 . Therefore
  
6 HCM = 75 , so 6 ACM = 6 HCM ; 6 HCA = 75 ; 45 = 30 .

b) Let S be the area of 4ABC . We know that S = BC  AH =

2
BC  AB . Since CM is a median, we obtain S = S . If MP ?
ACM
4 2
S AC  PM
AC , then 2 = 2 . It follows from 4PMC that PM =
1 MC (6 PCM = 30). Therefore S = AC  MC , so AC  MC =
2 2 4 2
BC  AB , which implies MC = AB  BC .
4 2AC
Problem 8.3. Given n points on a circle denoted consecutively
by A1 A2 : : :  An (n  3). Initially 1 is written at A1 and 0 at all
remaining points. The following operation is allowed: choose a point
Ai where a 1 is written and replace the numbers a, b and c written
at the points Ai;1 Ai and Ai+1 by 1 ; a, 1 ; b and 1 ; c, respectively.
(Here A0 means An and An+1 means A1.)
a) If n = 1999, is it possible to have a 0 in all points after per-
forming the described operation a nite number of times?
6
 b) Find all values of n such that it is not possible to have a 0 in
all points after nite number of operations.
Kiril Bankov

Solution: a) After performing the transformation from the con-

ditions of the problem consecutively for the points A1 A2 A3 : : : An;2,
we have the following distribution of 0s and 1s:
(1) A1 A2 A3 : : : An;3 An;2 An;1 An
1 1 1 ::: 1 0 1 1
If n = 1999, then arrange the obtained 1998 ones in 666 groups of
three 1s and then perform the operation on each of the groups. We
obtain a zero in every point.
b) If n = 3k + 1 we can repeat the steps from a) and we get
only zeroes in the points.
If n = 3k +2, then starting from (1) and performing the operation
with An;1 one obtains:
A1 A2 A3 : : : An;2 An;1 An
1 1 1 ::: 1 0 0
There are 3k ones which can be arranged in k groups of 3 ones each
and again to obtain anly zeroes.
We shall prove now that if n = 3k, it is not possible to have only
zeroes after a nite number of operations. Assume the opposite, i. e.,
that after a nite number of operations we have only zeroes. Denote
the number of operations performed with the point Ai by ai. Since
each operation changes the number of ones by an odd number (1
or 3), it follows that the sum S = a1 + a2 + : : : + an of all operations
7
is an odd number. On the other hand S = (a1 + a2 + a3) + (a4 +
a5 + a6) + : : : + (a3k;2 + a3k;1 + a3k ). Note that a1 + a2 + a3 is equal
to the number of changes (from 0 to 1 or vice versa ) of the number
written at A2. Since at the beginning there is a 0 written at A2 and
a 0 again at the end, it follows that a1 + a2 + a3 is an even number.
The same applies for a4 + a5 + a6 and so on. Therefore S is a sum
of even numbers, a contradiction to the fact that S is odd. Answer
to b): all numbers divisible by 3.

Problem 9.1. It is known that if the real parameter a equals any

of the numbers p < q < r, then at least one of the remaining two is
a root of the equation
x2 ; (2 ; a)x + a2 ; 2a ; 7 = 0:
Prove that a) p > ; 38 b) p < ;1. Sava Grozdev

Solution: a) It follows from the conditions of the problem that

D = (2 ; a)2 ; 4(a2 ; 2a ; 7)  0, so ; 83  a  4 and therefore
p  ; 38 . When a = ; 83 , then the equation has an unique root
x = 37 , and when a = 37 , the roots are x = ; 83 and x = 73 . Thus in
the case of p = ; 38 three distinct numbers p, q and r satisfying the
condition do not exist. Therefore p > ; 83 .
b) Suppose that when a = p, then x = q is a root. The case
x = r is treated in the same fashion. Since a and x are symmetric in
8
the equality x2 ; (2 ; a)x + a2 ; 2a ; 7 = 0, we get that when a = q,
then x = p is a root. When a = r, at least one of the numbers p and
q is a root. Let that be p. Now when a = p, the equation has x = r
as a root (because of the symmetry). We obtain that when a = p,
the roots are q and r. We conclude now that in all cases p + q + r = 2.
It is clear that the roots of x2 ; (2 ; p)x + p2 ; 2p ; 7 = 0 are greater
than p and therefore p2 ; (2 ; p)p + p2 ; 2p ; 7 > 0, which gives
p < ;1 or p > 73 . But 3p < p + q + r = 2 and p < 32 . Therefore
p < ;1.

Problem 9.2. Through an interior point K of the non-equilateral

4A1A2A3 lines Q2P3 k A2A3, Q3P1 k A3A1 and Q1P2 k A1A2 are
drawn. (Q1, Q2, Q3 lie on A3A1, A1A2 and A2A3, respectively).
Points P1 P2 P3 Q1 Q2 Q3 lie on a circle k. Prove that:
a) 4P1P2 P3
= 4Q1Q2Q3
4A1A2A3
b) point K , the centre of k and the circumcentre 4A1A2A3 lie on
a line.
Rumen Kozarev

Solution: a) It follows from the conditions of the problem that

the arcs P1^Q1 P2^Q2 and P3^Q3 are equal. Therefore 6 P3P1P2 =
P3Q3P2 = P3^Q3 + Q3^P2 = P1^Q1 + Q3^P2 . Since the quadrilat-
2 2 2 2 2
eral A1P1KQ1 is a parallelogram we get 6 A3A1A2 = 6 Q1KP1 =
Q1^P1 + Q3^P2 , so 6 P P P = 6 A A A . Similarly, one can show
2 2 3 1 2 3 1 2

9
 that 6 Q3Q1Q2 = 6 A3A1A2, and the same equalities for the remain-
ing pairs of angles. Thus 4P1P2P3
4Q1Q2Q3
4A1A2A3 and
since the rst two triangles have the same circumcircle, they are
identical.
b) Since A1P1KQ1 is a parallelogram, we obtain 6 A1P1Q1 =
^ ^
6 KQ1 P1 = P2 Q2 + P1 Q2 = 6 P1 P2 P3 = 6 A1A2 A3. Similarly 6 A1Q1P1 =
6 A1 A2A3.
2 2
Let O be the circumcentre of 4A1A2A3. It is easy to see that
OA1 ? P1Q1,OA2 ? P2Q2 and OA3 ? P3Q3. Denote the midpoint
of OK by S . If R1 R2 and R3 are the midpoints of P1Q1 P2Q2 and
P3Q3 then SR1 k OA1 SR2 k OA2 SR3 k OA3 (SR1 is a middle
segment in 4OKA1 =) SR1 k OA1). Therefore S lies on the axes
of symmetry of P1Q1 P2Q2 and P3 Q3, so S is the centre of the circle
through the points P1 P2 P3 Q1 Q2 Q3.

Problem 9.3. Find all polynomials f (x) = xn + an;1xn;1 +    +

a1x + a0 a0 = 6 0 with integer coecients such that f (ai) = 0 i =
0 : : :  n ; 1. Sava Grozdev

Solution: It is clear that n > 1. Since f (x) = (x ; a0)(x ;

a1) : : : (x ; an;1 ), it follows from f (0) = a0 = (;1)n a0 : : : an;1 that
jaij = 1 for i = 1 2 : : :  n ; 1.
First case: ja0j = 1. Now f (x) = (x ; 1)p(x + 1)q  p + q = n > 1.
We know that (x ; 1)p(x + 1)q = (xp ; pxp;1 + : : :)(xq + qxq;1 + : : :)
and comparing the coecients in front of xn;1 and xn;2 we obtain
q ; p = an;1 = 1 :
q(q;1) + p(p;1) ; pq = a
2 2 n;2 = 1

10
 It is easy to nd now that p + q = 3 and so p = 1 q = 2 or
p = 2 q = 1. In the rst case we get that f (x) = x3 + x2 ; x ; 1,
which is a solution and in the second one f (x) = x3 ; x2 ; x + 1,
which is not.
Second case: ja0j  2. Now 0 = f (a0) = jan0 + an;1an0 ;1 n+;1   +
a1a0+a0j  ja0jn ;ja0jn;1;   ja0j2;ja0j;ja0j = ja0j(ja0j;ja2)(0j;ja10j ;1) 
0. Therefore ja0j = 2. Moreover an;1an0 ;1 : : : a1a0 a0 have the same
negativity|the opposite to those of an0 .
We conclude that a0 = ;2 n is an even number and ai = (;1)i+1
for i = 1 2 : : :  n ; 1. If n > 2, then a2 = ;1 and 0 = f (;1) =
(;1)n + (;1)n (;1)n;1 +    + (;1)2(;1) ; 2 = ;n ; 2 6= 0, which
is impossible. Therefore n = 2 and f (x) = x2 + x ; 2, which is a
solution.
Answer:
f (x) = x2 + x ; 2 and f (x) = x3 + x2 ; x ; 1:

Problem 10.1. Prove that the inequality

x(2  3x ; 4xx2 ++xx++12 )  0
2

holds for any real number x. Rumen Kozarev

Solution: 1) Let x  0. Then 2  3x  2. We shall show that

4x2 + x + 2  2. The last inequality is equivalent to 4x2 + x + 2 
x2 + x + 1
2x2 + 2x + 2 () x(2x ; 1)  0 () x 2 (;1 0]  21 +1 and
 

therefore it holds for x  0.

11
 2) Let x > 0. We prove that x2 ++xx++12 ) < 2  3x. Assume
4 x2

the opposite, i. e., 4xx2 ++xx++12  2  3x =) 4xx2 ++xx++12 > 2  30 =

2 2

2 () x(2x ; 1) > 0 () x 2 (;1 0)  ( 21 +1). Since x > 0,

we obtain that x 2 21 +1 . Therefore 4xx2 ++xx++12  2  3x >
  2

2  3 12 > 3 () 4x2 +px + 2 > 3xp2 + 3x + 2 () x2 ;p2x ; 1 >

0 () x 2 (;1 1 ; 2)  (1 + 2 +1) =) x 2 (1 + 2 +1),
because x  12 . Thus 4xx2 ++xx++12  2  3x > 2  31+ 2 > 2  32 = 18.
2 p

Since obviously 4xx2 ++xx++12 < 4xx2 + 4x + 4 = 4 for any x > 0, we

2 2

+x+1
get a contradiction.

Problem 10.2. Let M be an interior point in the square ABCD.

Denote the second points of intersection of the lines AM , BM , CM ,
DM with the circumcircle of ABCD by A1, B1, C1, D1, respectively.
Prove that
A1B1  C1D1 = A1D1  B1C1:
Emil Kolev

Solution: Since 4ABM

4A1B1M 4BCM
4B1C1M 4CDM

4C1D1M 4DAM
4D1A1M , we obtain
AB = BM BC = BM CD = DM DA = DM :
A1B1 A1M B1C1 C1M C1D1 C1M D1 A1 A1M

12
It is easy to see now that
AB  CD = BM  DM = BC  DA :
A1B1 C1D1 A1M  C1M B1C1 D1 A1
Using that AB = BC = CD = DA it follows from the above that
A1B1  C1D1 = A1D1  B1C1, Q. E. D.

Problem 10.3. Consider n points in the plane such that no three

lie on a line. What is the least number of segments having their ends
in the given points such that for any two points A and B there exists
a point C connected to both A and B ? Emil Kolev

Solution: Denote the points by A1 A2 : : :  An. Draw segments

connecting A1 with all remaining points. Also, draw segments A2A3,
A4A5, . . . , An;1 An when n is odd and A2A3, A4A5, . . . , An;2An;1,
A2An when n is even. It is easy to see that the condition of the
problem is met and that there are 3n 2; 3 segments (dM e denotes
 

the least natural number which is greater or equal to M ).

Suppose it is possible to draw less than 3n 2; 3 segments and
 

to meet the condition of the problem. Obviously each point is con-

nected by a segment with another one. If each point is connected to
at least three others, we will have that the number of segments is
at least 32n , which is greater than 3n 2; 3 . Therefore there exists
 

a point (let that be A1) connected with at most two others. If A1 is

connected to exactly one point (let that be A2), a point connected to
both A1 and A2 does not exist. Therefore A1 is connected to exactly
two points (let them be A2 and A3). It is easy to see that A2 and A3
13
are connected by a segment. Consider the pairs A1 and Ai for any
i > 3. It is clear that the point connected to both A1 and Ai could
be either A2 or A3. In both cases Ai is connected to A2 or A3. Since
there are at least two segments from each point Ai i > 3 then the
number of segments from Ai i > 3 is at least 2(n ; 3). Further, since
at least n ; 3 from these points connect some point of Ai i > 3 with
A2 or A3 (and therefore they are counted once) the total number of
drawn segments is at least 3 + n ; 3 + n ; 3 = 3n ; 3 .
   

2 2
This is a contradiction with the number of drawn segments.
Therefore the answer is 3n 2; 3 .
 

Problem 11.1. Given a function f (x) dened for any real x and
f (tgx) = sin 2x for x 2 (; 2  2 ). Find the minimum and the maxi-
mum of the function f (sin3 x)  f (cos3 x).
Oleg Mushkarov, Nikolai Nikolov

Solution: Let t = tgx. Then sin 2x = 1 +2tt2 and it follows from

the conditions of the problem that f (t) = 1 +2tt2 for any t. Therefore
4 sin 3
x  cos3x
f (sin x)  f (cos x) = (1 + sin6 x)(1 + cos6 x) =
3 3

4(sin x  cos x)3

2 ; 3(sin x  cos x)2 + (sin x  cos x)6

14
Let u = sin x  cos x = 21 sin 2x. Then u 2 ; 21  12 ] and we have to nd
the minimum and the maximum of the function g(u) = 2 ; 34uu3 + u6
3

in the interval ; 12  12 ]. We obtain

g0(u) = 12u (1(2;;u3u)(2u+ +u6u)2 + 1) > 0

2 2 4 2

when u 2 ; 12  12 ] and so g(u) is an increasing function in the in-

terval ; 21  12 ]. It follows now that max g ( u) = g ( 1 ) = 32 and
2 81
u2 ; 21  12 ]
min1 1 g(u) = g(; 12 ) = ; 81 32 .
u2 ; 2  2 ]

Problem 11.2.;!A circle

;!is tangent to the circumcircle of 4ABC
and to the rays AB and AC at points M and N , respectively. Prove
that the excentre to side BC of 4ABC lies on the segment MN .
Oleg Mushkarov, Nikolai Nikolov

Solution: Let O be the circumcentre of 4ABC and L be the cen-

tre of the circle tangent to the circumcircle of 4ABC . Firstwe
shall nd the radius  of this circle. For 4OAL we get AL = A ,
sin 2
AO = R, OL = R + , 6 OAL = j B ; C j . From the Cosine Law we
2
2R cos B;2 C
obtain (R + )2 = R2 + 2 A ;
2
. After simplication
sin sin A
2 2

15
the above becomes
 cos2 A2 = 2R sin A2 (sin A2 + cos B ;2 C ) = 2R sin A2 cos B2 cos C2 :
s s s

Since sin A2 = (p ; bbc

)(p ; c) , cos B = p(p ; b) , cos C = p(p ; c)
2 ac 2 ab
and abc = 4RS , it follows from the previous equality that  =
ra , where r is the exradius to side BC of 4ABC . Let I =
a
cos2 A2
AL \ MN and let T be the projection of I on the line AB . Since
AL ? MN , it follows that IT = AIAM  IM = AI  AM  LM =
AM  AL
AI  LM = LM cos2 A = r . Since I lies on the bisector of 6 A, we
AL 2 a
conclude that I is the excentre to side BC of 4ABC .

Problem 11.3. Given an orthogonal coordinate system with ori-

gin O in the plane. Distinct real numbers are written at the points
with integer coordinates. Let A be a nonempty nite set of integer
points which is central-symmetric regarding O and O 2= A. Prove
that there exists an integer point X such that if AX is the image of A
under translation dened by OX ~ , then at least half of the numbers
written at the points of AX are greater than the number written
at X . Avgustin Marinov

Solution: Let us denote the number of points in A, which is obvi-

ously even, by 2s. Connect all integer points X with the points from
AX by arrows so that the arrow points to the smaller number. Sup-
pose no point X with the required property exists. Then there are
at least s + 1 arrows pointing out of any point X . For every natural

16
number n denote the square with vertices (n n) (;n n) (;n ;n)
and (n ;n) by Kn .
Since A is a nite set, there exists a natural number d such that
A  Kd . For every n denote the number of arrows within the square
Kn by Sn . Since there are at least s + 1 arrows pointing out of every
integer point of Kn (and these arrows are within the square Kn+d ),
it follows that (2n + 1)2(s + 1)  Sn+d . On the other hand, since AX
is a central-symmetric set, there are at most s ; 1 arrows pointing to
every integer point of Kn+d . Therefore Sn+d  (2n + 2d + 1)2(s ; 1).
Thus (2n + 1)2(s + 1)  (2n + 2d + 1)2(s ; 1), so s + 1  (1 +
2d )2(s ; 1) for any n. When n ! 1 one obtains s + 1  s ; 1,
2n + 1
a contradiction. Therefore a point X with the required property
does exist.

17
 WINTER MATHEMATICAL COMPETITION

1995

Grade 8 | First Group.

Problem 1. Prove that for every positive integer n the following proposition holds:
\The number 7 is a divisor of 3n + n3 if and only if 7 is a divisor of 3n :n3 + 1."
Solution. If 7 is a divisor of n, then 7 is neither a divisor of 3n + n3 nor a divisor of 3n :n3 +1.
Let 7 be not a divisor of n. In this case 7 divides n6 ; 1 = (n3 ; 1)(n3 + 1) and since 7 is a
prime number, then 7 divides either n3 ; 1 or n3 + 1. Now the above proposition follows from
the equalities:
3n :n3 + 1 = (n3 ; 1)(3n ; 1) + (n3 + 3n )
and
3n :n3 + 1 = (n3 + 1)(3n + 1) ; (n3 + 3n ):
Problem 2. Let ABCDE be a convex pentagon and let M , P , N , Q be the midpoints of
the segments AB , BC , CD, DE respectively. If K and L are the midpoints of the segments
MN and PQ respectively and the segment AE is of length a, nd the length of the segment
KL.
Figure 1. Figure 2.

Solution. Let F be the midpoint of the segment AD (Figure 1). Then the quadrilateral
MPNF is a parallelogram. Hence K is midpoint of the segment FP . It follows from here that
KL = 21 :FQ. On the other hand FQ = 21 :a (because F and Q are midpoints of ED and AD
respectively).
Therefore KL = 14 :a.

1
 Problem 3. Every point in the plane is colored either in black or in white. Prove that
there exists a right angled triangle with hypotenuse of length 2 and an acute angle of 60 , which
vertices are colored in one and the same colour.
Solution. First we shall show that there exist two points which are colored in one and the
same colour and the distance between them is 2. Indeed let ABC be a equilateral triangle of
side 2. Obviously two of its vertices (say A and B ) are colored in one and the same colour (e.g.
white). Let AXY BZT be a regular hexagon with a big diagonal AB (Figure 2).
If one of the vertices X , Y , Z , T (e.g. X ) is white, then the vertices of the triangle ABX
(6 AXB = 90 , 6 XAB = 60 ) are colored in one and the same colour. Otherwise the vertices
of the triangle XY T (6 Y XT = 90 , 6 XY T = 60 ) are colored in one and the same colour.
Grade 8.

Problem 1. Let A = p4x2 +1 4x + 1 and B = px22x;;2x2+ 1 . Find all integer values of x,

for which the number C = 2A + B is an integer.
3
Solution. We have
A = p 1 2 = j2x 1+ 1j  B = p2(x ; 1)2 = 2(jxx;;11)j 
(2x + 1) (x ; 1)
and  1 
2 x ; 1
C = 3  j2x + 1j + jx ; 1j 
1. Let x > 1. Then   4(x + 1)
C = 23  2x 1+ 1 + 1 = 3(2 x + 1) > 0
and
4(x + 1) ; 1 = 1 ; 2x < 0:
C ; 1 = 3(2x + 1) 3(2x + 1)
Hence 0 < C < 1, i.e. C is not an integer for any x > 1.
2. Let ; 21 < x  1. Then x = 0 (because x is an integer) and C = 0. Thus x = 0 is a
solution of the problem.
3. Let x < ; 12 . Then x  ;1 (because x is an integer). It is clear that
2  1  4 (x + 1)  0
C = 3 ; 2x + 1 ; 1 = ; 3(2 x + 1)
and  
C + 1 = 23 ; 2x 1+ 1 ; 1 + 1 = 1 ; 3(24(x + 1) = 2x ; 1 > 0:
x + 1) 2 (2x + 1)
Hence ;1 < C  0, i.e. C = 0 and x = ;1.
Finally only x = 0 and x = ;1 are solutions of the given problem.
Problem 2. Let M and N be the midpoints of the sides BC and AC of the triangle ABC ,
(AB 6= AC , AB 6= BC ) and G be the intersection point of the lines AM and BN . The angle
bisectors of 6 BAC and 6 ABC intersect BC and AC in the points D and E respectively. Prove

2
that the quadrilateral DEMN is inscribed in a circle if and only if there exists a circle, inscribed
in the quadrilateral CMGN .
Figure 3. Figure 4.

Solution. Without loss of generality we can suppose that N is between A and E . There are
two possibilities for the points D and E which are shown in the Figure 3 and Figure 4.
The quadrilateral DEMN is inscribed in a circle if and only if 6 CNM = 6 CDE , i.e.
6 BAN + 6 BDE = 1800 (because MN kAB and 6 CNM = 6 BAN ). Thus the quadrilateral
DEMN is inscribed in a circle i the quadrilateral ABDE is inscribed in a circle. This is
equivalent to 6 DAE = 6 DBE , i.e. to AC = BC .
Therefore we should prove that there exists a circle, inscribed in the quadrilateral CMGN
if and only if AC = BC .
Let AC = BC . Then CM = CN and since G is the cen- Figure 5.
ter of gravity of the triangle ABC , we have GM = 3 AM = 1
1 BN = GN . Hence CM + GN = CN + GM , i.e. there ex-
3
ists a circle, inscribed in the quadrilateral CMGN (Figure
5).
Conversely if there exists a circle, inscribed in the quadri-
lateral CMGN , then CM + GN = CN + GM and 12 BC +
1 BN = 1 AC + 1 AM . Hence AM ; BN = 3 (BC ; AC ).
3 2 3 2
Let K and L be the points of contact of the circle and the
sides CM and CN respectively. Obviously CK = CL.
On the other hand this circle is inscribed in the triangles
ACM and BCN . Hence CK = 21 (AC + CM ; AM ), CL =
1 (BC + CN ; BN ). Thus AM ; BN = 1 (AC ; BC ). It follows from here that 3 (BC ; AC ) =
2 2 2
1 (AC ; BC ), i.e. AC = BC .
2
Problem 3. Thirty points are given in the plane. Some of them are connected with segments
as it is shown in the Figure 6. The points are labeled with dierent positive integers.
If a is a segment and p and q are the numbers, corresponding to its endpoints, we denote
 (a) = jp ; qj.
a) Construct an example of labeling of the points with the integers 1 2 : : : 30, in which there

3
exists exactly one segment a with  (a) = 5
b) Prove that for every labeling there exists at least one segment a with  (a)  5.
Figure 6. Figure 7.

Solution. a) A possible example is shown in the Figure 7.

b) Let the points be labeled with the positive integers m = m1 < m2 < : : : < m30 = M . It
is clear that M  m + 29.
Let A and B be the points labeled with m and M respectively and let a1 = AC1 , a2 = C1C2,
: : :, ak;1 = Ck;2Ck;1, ak = Ck;1B be the shortest path of segments, connecting A and B .
It is not dicult to see that k  7. If we assume that  (ai ) < 5 for i = 1 2 : : :k, then
m + 29  M  m + 4k  m + 28, which is a contradiction.
Grade 9.
Problem 1. Let m be a real number, such that the roots x1 and x2 of the equation
f (x) = x2 + (m ; 4)x + m2 ; 3m + 3 = 0
are real numbers.
a) Find all values of m for which x21 + x22 = 6.
b) Prove that
1 < mx1 + mx2 + 8  121 :
2 2

1 ; x1 1 ; x2 9
Solution. a) Since x1 and x2 are real numbers, then
 
D (f ) = (m ; 4)2 ; 4 m2 ; 3m + 3 = ;3m2 + 4m + 4  0:
Hence ; 23  m  2. On the other hand
6 = x21 + x22 = (x1 + x2 )2 ; 2x1x2 = ;m2 ; 2m + 10:
p
We obtain from here that m = ;1  5. But
p p
;1 ; 5 < ; 32 < ;1 + 5 < 2
4
 p
and therefore only m = 5 ; 1 is a solution of the given problem.
b) We have
mx21 + mx22 = mx21(1 ; x2) + x22(1 ; x1 )]
1 ; x1 1 ; x2 f (1)

= x1 + x2 ; x1x2 (x1 + x2 )
2 2

m;2
= m ; 8m + 13m ; 2 = m2 ; 6m + 1
3 2

m;2
Thus if F = 1mx mx2 + 8, then F = (m ; 3)2 and
2 2
1
+
; x1 1 ; x2
121 = ; 2 ; 32  F > (2 ; 3)2 = 1:
9 3
Problem 2. The point D lies inside the acute triangle ABC . Three of the circumscribed
circles of the triangles ABC , ABD, BCD and CAD have equal radii. Prove that the fourth
circle has the same radius.
Solution. There are two cases:
1. The radii of the circumscribed circles of the Figure 8.
triangles ABD, BCD, CAD are equal.
Let O1 , O2 and O3 be the centers of these cir-
cles (Figure 8). Obviously the quadrilaterals O2 CO3D,
O3 AO1D and O1BO2 D are rhombuses.
Hence O2C kO3DkAO1 and O2 C = AO1 . Thus the
quadrilateral CAO1O2 is a parallelogram. It follows
from here that AC kO1O2 and since O1 O2 ? DB , then
BD ? AC . Analogously CD ? AB and AD ? BC .
Therefore D is the altitude center of 4ABC .
Now it is easy to see that 6 BDC = 1800 ; 6 BAC ,
which implies that the circumscribed circles of 4BCD
and 4ABC are symmetric according to the line BC .
Therefore their radii are equal.
2. The radii of the circumscribed circles of the triangles ABC , ACD and BCD are equal.
In this case the circumscribed circles of 4ABC and 4BCD are symmetric according the
line BC . Hence 6 BDC = 1800 ; 6 BAC . Analogously 6 ADC = 1800 ; 6 ABC . Therefore
6 ADB = 3600 ; 6 BDC ; 6 ADC = 1800 ; 6 ACB and the circumscribed circles of 4ABD and
4ABC are symmetric according to the line AB. Thus they have equal radii.
Problem 3. Let A be a set with 8 elements. Find the maximal number of 3-element subsets
of A, such that the intersection of any two of them is not a 2-element set.
Solution. Let B1  B2  : : : Bn be subsets of A such that jBi j = 3, jBi \ Bj j 6= 2 (i j = 1 : : : n).
Assume that there exists an element a 2 A, which belongs to four of the subsets B1  B2 : : : Bn
(e.g. a 2 B1  B2 B3  B4). Then jBi \ Bj j  1 (i j = 1 : : : 4). But Bi = 6 Bj if i 6= j , i.e.
jBi \ Bj j 6= 3. Thus jBi \ Bj j = 1 (i j = 1 2 3 4). It follows from here that jAj  1 + 4:2 = 9,

5
which is a contradiction. Therefore every element of A belongs to at most three of the subsets
B1 B2  : : : Bn . Then 3n  8:3, i.e. n  8.
If A = fa1 a2 : : : a8g, then the subsets
B1 = fa1 a2 a3g B2 = fa1 a4 a5g B3 = fa1 a6 a7g B4 = fa8 a3 a4g
B5 = fa8 a2 a6g B6 = fa8 a5 a7g B7 = fa3 a5 a6g B8 = fa2 a4 a7g
provide an example of exactly eight 3-element subsets of A, such that jBi \ Bj j 6= 2.
Therefore the searched number is n = 8.
Grade 10.
Problem 1. Find all positive roots of the equation
log(x+a;1) x +4 1 = loga 2
where a > 1 is a real number.
Solution. It is clear that if a > 1 and x > 0, then x + a ; 1 > 0 and 4 > 0. Hence in
x+1
this case log(x+a;1) 4
x + 1 is well dened. Since
log 4
log(x+a;1) 4 =
a x+1
x + 1 loga (x + a ; 1) 
then the given equation is equivalent to
4 = 2loga(x+a;1) : (1)
x+1
The function x +4 1 is strictly decreasing in the interval (0 +1). The function loga (x + a ; 1)
is strictly increasing in the interval (0 +1) and obviously the same is true for the function
2loga (x+a;1) . Therefore the equation (1) has no more than one root in the interval (0 +1).
On the other hand it is easy to check that x = 1 is a root of this equation.
Problem 2. A circle _
k with center O and diameter AB is given. The points C and D are
moving along the arc AB so that C is between B and D and if 6 BOC = 2 and 6 AOD = 2,
then tan  = tan  + 32 . Prove that the lines, which are perpendicular to CD and divide CD in
ratio 1 : 4 measured from C , pass through a xed _
point of the given circle.
Solution. Let E be such a point on the arc AB , not containing C and D, that if 6 BOE = 2 ,
then tan = 2 (Figure 9).
We shall show that the point E satises the problem's conditions.
It is enough to prove that if F is the foot of the perpendicular from E to CD, then the point
F is between C and D and FD CF = 1 .
4
We have
tan(6 ECD) = tan(6 ECA + 6 ACD) = tan(
2 ; + )

6
  
tan
2 ; + tan  1 + tan 
tan 
= 
 = 2 tan  = 12+;2tan 
1 ; tan 2 ; : tan  1 ; 2
 3 
1 + 2 tan  + 2
=  3  = 41+ 2 tan  = 8 + 4 tan  
1 ; 2 tan 
2 ; tan  + 2 2 ; tan 
tan(6 EDC ) = tan(6 EDB + 6 BDC ) = tan( +  ) = 1tan + tan  = 2 + tan  :
; tan : tan  1 ; 2 tan 
Thus Figure 9.
tan(6 ECD) = 4 tan(6 EDC ):
It follows from the last equality that either
6 ECD 
2 and 6 EDC 
2
or
6 ECD <
2 and 6 EDC <
2 :
But these angles belong to the triangle EDC .
Hence they are acute angles. Therefore the point
F is between the points C and D. Since
tan(6 ECD) = EF EF
FC and tan( EDC ) = FD 
6

then EF =
FC FD 4 EF , i.e. FD = 4:FC .
Problem 3. Find all prime numbers p, for which the number p(2p;1 ; 1) is a k-th power
(k > 1) of a positive integer.
Solution. Let p(2p;1 ; 1) = xk (x > 0 is an integer). It is clear that p 6= 2, i.e. p = 2q + 1 is
an odd number. Since p=x, then x = p:y (y is a positive integer) and (2q ; 1)(2q + 1) = pk;1 y k .
At least one of the numbers 2q ; 1 and 2q + 1 is a k-th power of an integer, because they are
relatively prime numbers.
1. Let 2q ; 1 = z k , i.e. 2q = z k + 1. If k is even, then z k + 1 is not divisible by 4 . Hence
q = 1, p = 3 and p(2p;1 ; 1) = 32.
If k = 2l + 1 then 2q = (z + 1)(z 2l ; z 2l;1 +    ; z + 1), i.e. z + 1 = 2, where 0   < q .
On the other hand
2q = (2 ; 1)2l+1 + 1 = 22:A + 2 (2l + 1)
(A is an integer). The last equality contradicts with  < q .
2. Let 2q + 1 = z k , i.e. 2q = z k ; 1. If k is odd, we obtain a contradiction as in the previous
case.
If k = 2l, then (z l ; 1)(z l + 1) = 2q and since GCD(z l ; 1 z l + 1) = 2, we have z l ; 1 = 2,
i.e. q = 3, p = 7, p(2p;1 ; 1) = 7:63 = 212.
The sought numbers are p = 3 and p = 7.

7
 Grade 11.
Problem 1. Find all the values of the real parameter p, for which the range of the function
f (x) = 2(1p;;psin
) + cos x :
2x

contains the interval 1 2].

Solution. Let y = cos x. The problem is to nd all values of p, such that for every k 2 1 2]
the equation
2(1 ; p) + y = k
p ; 1 + y2
has at least one root y0 2 ;1 1], i.e. we should nd all values of q = 1 ; p such that for every
k 2 1 2] the equation
ky 2 ; y ; q (k + 2) = 0
has at least one root y02 6= q , y0 2 ;1 1].
If y02 = q , then ;y0 ; 2q = 0, y0 = ;2q and 4q 2 = q , i.e. q1 = 0 or q2 = 41 .
If q = 0 the roots of the equation ky 2 ; y = 0 are y1 = 0, y2 = 1 . But y2 = 1 2 ;1 1] and
k k
y22 6= q = 0 for every k 2 1 2]. Thus q = 0 satises the problem's conditions.
If q = 1 , then the roots of the equation ky 2 ; y ; 1 (k + 2) = 0 are y1 = ; 1 , y2 = 1 + 1 .
4 4 2 2 k
1
But y2 2 ;1 1] only if k = 2. Thus q = 4 doesn't satisfy the problems conditions.
Let q 6= 0 and q 6= 41 . The equation ky 2 ;y ;q (k +2) = 0 has real roots i D = 1+4k(k +2)q 
0, i.e. i q  ; 4k(k1+ 2) for every k 2 1 2]. Hence q  ; 32 1 . The vertex of the parabola

g(y) = ky2 ; y ; q(k + 2) has as its rst coordinate y 0 = 21k 2 (0 1]  ;1 1]. Therefore the
equation g (y ) = 0 has at least one root y0 2 ;1 1] i at least one of the inequalities g (;1)  0
and g (1)  0 holds. It is easy to obtain from here that q  kk + + 1 for every k 2 1 2]. Hence
2
2  1 2 1
q  3 , i.e. q 2 ; 32  3 and q 6= 4 .
 1 33 
Finally p 2 3  32 , p 6= 43 .
Problem 2. The point O is on the edge AB of the tetrahedron ABCD. The circumscribed
sphere of the tetrahedron AOCD intersects the edges BC and BD in the points M and N
(M 6= C , N 6= D) respectively. The circumscribed sphere of the tetrahedron BOCD intersects
the edges AC and AD in the points P and Q (P 6= C , Q 6= D) respectively. Prove that the
triangles OMN and OQP are similar.
Solution. Since the points A, C , M and O are in one and the same plane and are lying on
a sphere, then the quadrilateral ACMO is inscribed in a circle (Figure 10). Then 4BOM
4BCA, and OM BM OQ AQ
CA = BA (Figure 11). Analogously from 4ABD: BD = AB . Hence
OM = BM  AC : (1)
OQ AQ BD
8
 From 4BCD we have MN =
CD BD
BM , and from 4ACD: PQ = AQ . Thus
CD AC
Figure 10. Figure 11.

MN = BM  AC : (2)
PQ AQ BD
It follows from (1) and (2) that OM MN ON MN
OQ = PQ . Similarly OP = PQ .
Therefore 4OMN 4OQP .
Problem 3. Solve in positive integers the equation:
1 + 5x = 2y + 2z :5t:
Solution. If y  2 and z  2, then the right side of the given equation is divisible by 4. But
1 + 5x  2 (mod 4) and hence min(y z ) = 1. On the other hand 2y  1 (mod 5). Thus y is
divisible by 4 (4 is the index of 2 modulo 5). It follows from here that y  4, z = min(y z ) = 1
and the equation is 1 + 5x = 24y0 + 2:5t, where y = 4y0 (y0 is a positive integer).
If t = 1 then 5x ; 24y0 = 9. Using congruence modulo 3 we obtain that x is even, i.e. x = 2x0.
Hence (5x0 ; 22y0 )(5x0 + 22y0 ) = 9, from where x0 = y0 = 1, i.e. x = 2 and y = 4.
Let t > 1. Then 16y0  1 (mod 25) and y0 is divisible by 5 (5 is the index of 3 modulo 25).
Hence y0 = 5y1 and 1 + 5x = 220y1 + 2:5t. On the other hand 210  1 (mod 11) and 5x  2:5t
(mod 11). Obviously x > t. Thus 5x;t  2 (mod 11), which is not true.
Therefore the given equation has a unique solution (x y z t) = (2 4 1 1).
Grade 12.

p 2Problem 1. For every real number x we denote by f (x) the maximal value of the function
t + 2t + 2 in the interval x ; 2 x].
a) Prove that f (x) is an even function and nd its minimal value
b) Prove that the function f (x) is not dierentiable for x = 0.
c) Prove that the sequence an = ff (n)g, n = 1 2 : : : is convergent and nd its limit.
(For every real number a we denote with fag the unique real number in the interval 0 1)
for which the number a ; fag is an integer.)

9
 p
Solution. It is clear that the maximal value of the function t2 + 2t + 2 in the interval
x ; 2 x] is reached at the endpoints of this interval. Since the inequality (x ; 2)2 +2(x ; 2)+2 
x2 + 2x + 2 is equivalent to x  0, then
8p 2
< x ; 2x + 2
> if x  0
f (x) = > p
: x2 + 2x + 2 if x  0:
a) Obviously f (x) = f (;x), i.e. f (x) is an even function and its minimal value is f (0) = 2.
b) Since p2 p p
lim f ( x) ; f (0) = lim x ; 2 x + 2 ; 2 = ; 2
x!0
x<0 x;0 x !
x<0
0 x 2
and p2 p p
lim f ( x ) ; f (0) = lim x + 2 x + 2 ; 2 = 2
x!0
x>0 x ; 0 x!0
x>0 x 2
then f 0(0) doesn't exist. np o p
c) It is clear that an = n2 + 2n + 2 . On the other hand n + 1 < n2 + 2n + 2 < n + 2,
p
i.e. an = n2 + 2n + 2 ; (n + 1).
Therefore
a = nlim p 1 = 0:
nlim
!1 n !1 n2 + 2n + 2 + n + 1

Figure 12. Problem 2. The points A, B, C and D lie on

a straight line in the given order. A circle k passes
through the points B and C and AM , AN , DK and
DL are tangents to k.
a) Prove that the points P = MN \ BC and Q =
KL \ BC don't depend on the circle k.
b). If AD = a, BC = b (a > b) and the segment
BC is moving along AD, nd the minimal length of
the segment PQ.
Solution. a) From the Steward's formula for the
triangle AMN and the segment AP (Figure 12) we
have:
AP 2 :MN = AM 2:NP + AN 2:MP ; MN:MP:NP
= (AM 2 ; MP:NP ):MN
(here AM = AN because they are tangents to a circle).
Hence
AP 2 = AM 2 ; MP:NP = AB:AC ; BP:CP
= AB:AC ; (AC ; AP )(AP ; AB )
= 2:AB:AC ; AP (AB + AC ) + AP 2 

10
i.e. AP = AB 2AB:AC . Analogously DQ = 2DB:DC . These equalities show that the position
+ AC DC + DB
of the points P and Q doesn't depend on the circle k.
b) Let us denote AB = x, BC = y and CD = z . It follows from a) that
PQ = AD ; AP ; DQ = x + y + z ; 2xy(+x +2xy ) ; 2zy(+
y + z)
2z
   
= (x + y ) 1 ; y +2x2x + z 1 ; 2(yy++2zz)

= (yy+(x2x+)(yy ++ z2)z ) :
2

Having in mind that y = b, x + y + z = a, we obtain

PQ = (a + x ; zb)(aa + z ; x) = a2 ; b(xa; z)2  ba :

2 2 2

Therefore the minimal length of the segment PQ is b and this length is reached i AB =
2

a
CD = x = z = a ; b .
2
Problem 3. Find all prime numbers p and q , such that the number 2p + 2q is divisible by
p:q.
Solution.
Lemma. If k > 1, then k doesn't divide 2k;1 + 1.
Proof. Assume that k divides 2k;1 + 1. Obviously k is odd. Let k = p1 1 p2 2 : : :pr r , where
p1 p2 : : : pr are odd prime numbers (pi 6= pj if i 6= j ), 1  2 : : : r are positive integers and
r  1. Let pi ; 1 = 2mi :ti, where ti are odd integers ( i = 1 2 : : : r). Let m1 be the smallest
number in the sequence m1  m2 : : : mr . It follows from pi  1 (mod(pi ; 1)) that pi  1
(mod 2mi ) and pi i  1 (mod m1 :u
2mi ), i = 1 2 : : : r. Hence k ; 1 = 2m1 :u (u is an integer). If
2  ;1 (mod k), then 2
k ;1 2  ;1 (mod k) and 2(p1;1):u  ;1 (mod p1), because t1 is
odd. But 2 1  1 (mod p1 ) | a contradiction.
p ;1

Let 2p + 2q is divisible by p:q . We have three cases:

1. p and q are odd prime numbers. Then 2p + 2q  0 (mod p) and since 2p  2 (mod p),
then 2q  ;2(modp) and 2pq  (;2)p  ;2 (mod p). Similarly 2pq  ;2 (mod q ). Thus
2pq;1  ;1 (mod pq ), which is a contradiction with the lemma.
2. p = 2, q > 2. Then 4 + 2q  0(modq ) and it follows from 2q  2 (mod q ) that
6  0(mod q ) and q = 3. It is clear that 22 + 23 = 12  0 (mod 2  3).
3. p = q = 2. Then 22 + 22 = 8  0 (mod 2  2).
Therefore the sought numbers are: p = q = 2 p = 2, q = 3 p = 3, q = 2.

11
 WINTER MATHEMATICAL COMPETITION

1996

Grade 8
Problem 1. For which integer values of the parameter a the equation j2x + 1j + jx ; 2j = a
has an integer solutions?
Solution. We shall consider the following cases:
I. x > 2. Then 2x + 1 > 0, x ; 2 > 0 and the equation is equivalent to 2x + 1 + x ; 2 = a.
Thus, x = 1 +3 a , which is a solution when 1 +3 a > 2, i.e. when a > 5.
II. ; 12  x  2. Then 2x+1  0, x;2  0 and the equation has the form 2x+1;(x;2) = a.
Thus, x = a ; 3, which is a solution when ; 1  a ; 3  2, i.e. 5  a  5. In this interval the
2 2
integers are a = 3 4 5 and respectively we get x = 0 1 2, which are integer solutions.
III. x < ; 21 . Then 2x+1 < 0, x;2 < 0 and the equation is equivalent to ;(2x+1);(x;2) =
a or ;3x = a ; 1. Thus x = 1 ;3 a , which is a solution, when 1 ;3 a < ; 12 , i.e. when a > 25 .
According to the considered cases the given equation has solution when a > 25 . When a = 3
the only integer solution is x = 0. When a = 4 there are two integer solutions x = 1 and x = ;1.
When a = 5, x = 2 is the only integer solution. When a > 5 the equation has two solutions
x1 = 1 +3 a and x2 = 1 ;3 a . For these solutions we have: if a = 3k ; 1, then x1 is integer only
if a = 3k + 1, then x2 is integer only if a = 3k, there is no integer solution.
As a result, we get integer solutions when a = 3 or a = 3k  1, where k is a positive integer,
greater than 1.
Problem 2. The bisector AD (D 2 BC ) of the acute isosceles triangle ABC divides it into
two isosceles triangles. Let O and I be the incenter and the circumcenter of 4ABC , respectively.
AO meets BC in point E , while F is the intersection point of the lines BI and DO. Prove that:
a) the quadrilaterals ABEF and ADCF are rhombi with equal side lengths
b) If H is the altitude center of 4ABE , then the points A, D, E , F , H are concyclic.
Solution. Firstly, let us justify the position of AD. If it is a bisector of the angle between the
two equal sides, then it is perpendicular to BC , 4ADB and 4ADC are isosceles. Consequently
AD = BD = CD and 6 BAD = 6 CAD = 45 , i.e. 6 BAC = 90, which contradicts to the
condition that 4ABC is acute. It follows that AD is bisector of the angle belonging to the base
AB, while 4ABD and 4ACD are isosceles (Figure 1).
a) If 6 BAD = 6 CAD = , then 6 ABC = 2. Since 6 ADB > 6 CAD = 6 BAD , the only
possibility is 6 ADB = 6 ABC = 2 and AD = AB . But 6 ACD = 6 ADB ; 6 CAD = 2 ;  =

1
 , i.e. 6 ACD = 6 CAD and CD = AD. From the equality 2 +2 +  = 180 we dene  = 36.
Thus, 6 ACB = 36, 6 BAC = 6 ABC = 72. The points O and I lie on the bisector of 6 ACB ,
which is a segment bisector of AB . We have OA = OC and 6 OAC = 6 OCA = 18 = 6 OAD,
while 6 OAB = 36 + 18 = 54, 6 AEB = 180 ; 72 ; 54 = 54 = 6 BAE . It follows that
BE = AB and the bisector BI of 6 ABC is the segment bisector of AE . In the triangle ACD,
which is isosceles, we have that AO and CO are bisectors and it follows that DO is the bisector
of 6 ADC and the segment bisector of AC . Then F is the intersection point of the segment
bisectors of AE and AC , i.e. F is the circumcenter of 4ACE , the radius of the circumcircle is
AF = EF = CF . We have 6 AFE = 26 ACE = 72 , while 6 AFB = 6 BFE = 36 = 6 ABF =
6 EBF and consequently AF = AB = BE = EF , i.e. ABEF is a rhombus. In the same way
CF = AF = AB = AD = CD and ADCF is a rhombus, which side is AD = AB .
b) It is clear that 6 AEF = 6 AEB = 54 and Figure 1.
1
6 ADF = 6 ADC = frac12(180 ; 72) = 54,
2
the segment AF is seen from the points D and E
under 54. We draw a line through A, perpendicular
to BC , which intersects BI in the point H | the
altitude center of 4ABE , because BI ?AE . But in
4ABD the altitude AH is a bisector of 6 BAD =
36. Thus, 6 BAH = 18, 6 AHF = 18 + 36 =
54 = 6 AEF = 6 AOF . Note that the points D, E ,
H are in one and the same semiplane with respect
to AF . Consequently, D, E , H together with A and
F are concyclic.
Problem 3. Every day a student preparing
himself for the Winter competition in mathemat-
ics has been solving problems during a period of 5
weeks. He has been solving at least one problem daily but no more than 10 problems weekly.
a) Prove that during some consecutive days the student has solved 19 problems exactly.
b) If 1  n  34 is a natural number, prove that during some consecutive days the student
has solved n problems exactly.
Remark. Every week begins on Monday and ends on Sunday.
Solution. Since a) is a particular case of b), we shall solve b) only.
According to the condition the student has been solving problems during 5 7 = 35 days and
has solved at most 5 10 = 50 problems. Let xi be the number of the problems, solved during
the i-th day (i = 1 2 : : : 35).
Let 1  n  34 be a xed natural number. We want to prove that there exist such k < l,
that xk+1 + + xl = n. Denote Xi = x1 + + xi . Obviously,
1  X1 < X2 < : : : < X35  50
and the problem is to prove the existence of such k < l, that Xl ; Xk = n.
Case 1. 1  n  19. We consider the numbers
X1 < X2 < : : : < X35 X1 + n < X2 + n < : : : < X35 + n (1)
which are integers and their number is 70. Obviously they are in the interval 1 50 + n], in
which there are 50 + n  50 + 19 < 70 integers. Consequently, among the numbers (1) there
2
are at least two equal. The rst 35 of them as well as the next 35 are dierent from each other.
Therefore, there exist such k and l, for which Xl = Xk + n, i.e. for which Xl ; Xk = n.
Case 2. 20  n  34. Firstly, we shall prove the following
Lemma. If the integers z1 z2 : : : zm belong to the interval 1 2n] and if m > n, then among
the numbers z1  z2 : : : zm there are two, the dierence of which is equal to n exactly.
Proof. With the numbers from the interval 1 2n] we construct the following pairs:
(1 n + 1) (2 n + 2) : : : (n 2n):
The number of these pairs is n and the dierence of the numbers in each pair is equal to
n. Since m > n, at least two of the numbers z1  z2 : : : zm belong to one and the same pair.
Therefore their dierence is equal to n.
Let us nish now the solution of the problem.
If n  25, then 2n  50, and thus all the numbers X1 X2 : : : X35 are in the interval
1 2n]. On the other hand n < 35 and according to the lemma there are two numbers among
X1  X2 : : : X35 which dierence is n.
If 20  n  24, we represent the interval 1 50] as an union of the intervals 1 2n] 2n +1 50].
In the second one there are 50 ; (2n + 1) + 1 = 50 ; 2n integers. Then, the number of the
integers among X1  X2 : : : X35, which belong to the interval 1 2n], is at least 35 ; (50 ; 2n) =
2n ; 15  40 ; 15 = 25 > n. Consequently, we can apply the lemma again.
Remark. The case 1 can be solved by the lemma proved above.
Grade 9
Problem 1. Let f (x) = x3 ; (p + 5)x2 ; 2(p ; 3)(p ; 1)x + 4p2 ; 24p + 36, where p is a real
parameter.
a) Prove that f (3 ; p) = 0.
b) Find all values of p, for which two of the roots of the equation
p f (x) = 0 are lengths of the
cathetuses of a rectangle triangle which hypotenuse is equal to 4 2.
Solution. a) We have f (x) = (x + p ; 3)(x2 ; 2(p p + 1)x + 4(p ; 3)).
b) The roots of f (x) = 0 are x12 = p + 1  p2 ; 2p + 13 and x = 3 ; p. If p > 3,
 p 2 3
then x1 > 0, x2 > 0 and x3 < 0. The equation x21 + x22 = 32 = 4 2 gives p = 1, which is
impossible. If p = 3, then two of the roots are equal to 0 and this casepgives no solution. If p < 3,
then x1 > 0, x3 > 0 and x2 < 0. Therefore, 32 = x21 + x23 = (p + 1 + p2 ; 2p + 13)2 + (3 ; p)2.
This equation is equivalent to
q
(p + 1)(3(p ; 3) + 2 p2 ; 2p + 13) = 0:
p
If p 6= p;1, we get 2 p2p; 2p + 13 = 3(3 ;pp) > 0, from where 5p2 ; 46p + 29 = p 0 and
23  8 6 23 + 8 6 23 ; 8 6 23 ; 8 6.
p = 5 . Since > 3 and < 3, we nd p 1 = ;1 and p2 =
5 5 5
Problem 2. The incenter of the quadrilateral ABCD is O. The lines lA ? OA, lB ? OB ,
lC ? OC and lD ? OD are drawn through the points A, B, C and D respectively. The lines lA
and lB meet each other in the point K , lB and lC | in L, lC and lD | in M , lD and lA | in
N.
a) Prove that the lines KM and LN meet each other in the point O.

3
 b) If the lengths of the segments OK , OL and OM are p, q and r respectively, nd the
length of the segment ON .
Solution. a) We shall prove that the points N , O and L are colinear. Denote 6 ABC = 6 B ,
6 BCD = 6 C , 6 CDA = 6 D and 6 DAC = 6 A. Since O is incenter, the segments OA,OB ,OC
and OD are the bisectors of the corresponding angles of the quadrilateral ABCD. Note that
each of the quadrilaterals AKBO, BLCO, CMDO and DNAO is inscribed.
Figure 2. Consequently: 6 NOK + 6 KOL =  ;
6 ONA ; 6 OKA +  ; 6 OKB ; 6 OLB =
 ; 6 ADO ; 6 6 ABO6 +  ;6 6 BAO ; 6 BCO =
2 ; ( 2 + 2B + 2A + 2C ) = 2 ;  =  .
6 D

It follows from here that the points N ,O and

L are colinear. Analogously, the points K ,O
and M are colinear. Therefore, O is the inter-
section point of the diagonals of the quadri-
lateral KLMN .
b) Firstly, we shall prove that the quadri-
lateral KLMN is inscribed. Indeed,
6 NKL + 6 NML

= 6 AKO + 6 OKB + 6 DMO + 6 OMC

=
6 B + 6 A + 6 C + 6 D = 2:
2 2 2 2
Thus, OK:OM = OL:ON , from where
ON = OK:OM p:r
OL = q :
Problem 3. A square with side length 5 is divided into unit squares by parallel to its sides
lines. Let A be the set of the vertexes of the unit squares which are not on the sides of the given
square. How many points from A can be chosen at most in a way that no three of them are
vertexes of isosceles rectangle triangle?
Solution. We shall prove that the maximal number is 6. Let us enumerate the points in the
way, shown on the table 1.
It is easy to be seen that no 3 of the points 1,2,3,8,12 and 16 are vertexes of a isosceles
rectangle triangle. Assume that there exists a set of 7 points with the desired property. Note
that if 4 points form a square, then at most 2 of them can be among the already chosen ones.
The points 1,4,16 and 13 2,8,15 and 9 3,12,14 and 5 form squares.
13 14 15 16 Consequently, at most 6 of the chosen points lie on the contour. It
9 10 11 12 follows from here that at least one of the points 6,7,10 and 11 is from the
5 6 7 8 chosen ones. Due to the symmetry we may assume that this is the point 7.
1 2 3 4 Since the points 7,16 and 9 1,7 and 14 form isosceles rectangle triangles,
then at most two of the points 1,9,14 and 16 are from the chosen ones.
Table 1: The points 5,7,13 and 15 form a square and therefore at most one of the
points 5,13 and 15 is from the chosen ones. It follows from here that at
least 3 points are chosen from 2,3,4,6,8,10,11 and 12. By the pigeonhole principle we deduce that
at least two points are chosen in one of the sets 3,6,11,8 and 2,4,10,12. It is easy to see that if
the two points are in the rst set, then we have two possibilities | 3 and 11 or 6 and 8 (in both
4
cases it is not possible to chose more points on the square which encounters 7). Analogously, if
the two points are in the second set, then the possibilities are two again | 2 and 12 or 4 and
10 (in both cases it is not possible to chose more points on the square which encounters 7). The
contradiction shows that the maximal number of points which can be chosen is equal to 6.
Grade 10
Problem 1. Let p and q be such integers that the roots x1 and x2 of the quadratic equation
x + px + q = 0 are real numbers. Prove that if the numbers 1, x1 , x2 (in some order) form a
2
geometric progression, then the number q is a perfect cube.
Solution. There are two possibilities for the order of the numbers in the geometric progres-
sion: x1  1 x2 and 1 x1 x2. In the rst case we get q = x1x2 = 12 , i.e. q = 1 is a perfect
cube. Let now x2 = x21 . We have ;p = x1 + x2 = x1 + x21 , i.e. x1 satises the equation
x21 + x1 + p = 0. On the other hand x1 satises also x21 + px1 + q = 0. From these two equations
we get (p ; 1)x1 + (q ; p) = 0. If p 6= 1, then x1 is rational and q = x1 x2 = x1 x21 = x31 is the
cube of a rational number. Since q is integer, then it is a perfect cube. If p = 1, then q = p and
the quadratic equation becomes x2 + x + 1 = 0. The last equation has no real root.
Problem 2. A triangle ABC with a radius R of Figure 3.
the circumcircle is given. Let R1 and R2 be the radii of
the circles k1 and k2 , respectively, which pass through
C and are tangent to the line AB in A and B , respec-
tively.
a) Prove that the numbers R1, R and R2 form a
geometric progression.
b) Find the angles of 4ABC , if the radius of the
circle which is tangent to k1 , k2 and the line AB , is
equal to R4 .
Solution. a) Let AB = c, BC = a and CA = b.
Denote by O1 and O2 the centers of the circles k1 and k2 (Figure 3). Then O1 is the intersection
point of the perpendicular from A to AB and the segment bisector of AC . Since 6 MAO1 =
j90 ; 6 Aj (M is the midpoint of AC ), then R1 = AO1 = cos j90AM AC
 ; 6 Aj = 2 sin 6 A and by the
sine theorem it follows that
R1 = R ab : (1)
Analogously,
R2 = R ab : (2)
From here R1R2 = R2 and the proposition is proved.
b) Let O be the center of the circle which touches k1 , k2 and the line AB (Figure 4). Denote
by T the tangent point of this circle
p with AB , and by r its radius.
p From the rectangle trapezoid
p
ATOO1 it follows that AT = (R1 + r) ; (R1 ; r) = 2 rR1. Analogously, BT = 2 rR2.
2 2
p p
Then c = AT + TB = 2 rR1 + 2 rR2 and from here we nd r = p c p 2 . Using (1)
2

4( R1 + R2 )
and (2), we get
r = 4R(abc
2

a + b)2 : (3)

5
 Since (a +abb)2  14 and c2  4R2, it follows from (3) that r  R4 . The equation is reached
when a = b and c = 2R. Now, it follows that 6 A = 6 B = 45 and 6 C = 90.
Problem 3. A positive integer n and a Figure 4.
real number ' are given in a way that cos ' =
1
n . Find all positive integers k, for which the
number cos k' is an integer.
Solution. Case 1. n = 1. Then cos ' = 1
and ' = 2m (m = 0 1 : : :). For all k 2 N
the number cos k' is an integer.
Case 2. n = 2. Then cos ' = 21 and
' =  3 + 2m (m = 0 1 : : :). It is clear
that for all k 2 N, which are divisible by 3,
the number cos k' is an integer.
Case 3. n  3. We shall prove that for
all k 2 N the number cos k' is not an integer. Let n be odd. We have cos ' = n1 , cos 2' =
2 cos2 ' ; 1 = 2 ; n and (2 ; n2  n) = 1. We shall prove by induction that cos k' = a , where
2

n2 nk
(a n) = 1. Assume that the assertion is true for all integers from 1 to k. We shall check it for
k + 1. From
cos(k + 1)' + cos(k ; 1)' = 2 cos k' cos '
it follows that cos(k + 1)' = 2 a ; b = 2a ; bn , where cos k' = a and cos(k ; 1)' =
2

n nk nk;1 nk+1 nk
b . We have (a n) = 1 and (b n) = 1 according to the inductive assumption. It is clear that
nk;1
(2a ; bn2 n) = 1 and this ends the proof. The case when n is even is analogous. Now cos k'
is expressed by a fraction which denominator is equal to 2pk with n = 2p, while the nominator
has no common divisor with p.
Answer: if n = 1 8k 2 N
if n = 2 k = 3q , where q 2 N
if n  3 there is no solution.
Grade 11
Problem 1. Find the values of the real parameter a, for which the function
f (x) = x2 ; 2x ; jx ; 1 ; aj ; jx ; 2j + 4
has nonnegative values for all real x.
Solution. Firstly let 1 + a  2, i.e. a  1. Then
8
>
< x2 + 1 ; a x  1+a
f (x) = > x2 ; 2x + 3 + a 1 + a  x  2
: x2 ; 4x + 7 + a x  2:

6
 If 1 + a  2, i.e. a  1, we nd that
8 2
< x + 1 ; a
> x2
f (x) = > x2 ; 2x + 5 ; a 2  x  1 + a
: x2 ; 4x + 7 + a x  1 + a:

Hence the smallest value of f (x) is reached in one of the points 0, 1, 2, 1 + a.

We have f (0) = 2 ; ja + 1j, f (1) = 2 ; jaj, f (2) = 4 ; j1 ; aj, f (1 + a) = a2 ; ja ; 1j + 3.
These four numbers must be nonnegative. We nd that a 2 ;2 1]. Then ja ; 1j = 1 ; a and
f (1 + a) = a2 + a + 2 > 0 for all a.
Let now a 2= ;2 1]. If a < ;2, then x2 ; 2x + 3 + a < 0 when x = 1. Analogously, if a > 1,
then x2 + 1 ; a < 0 when x = 0.
Finally, a 2 ;2 1].
Problem 2. The point O is circumcenter of the acute triangle ABC . The points P and Q
lie on the sides AB and AC respectively. Prove that O lies on the line PQ if and only if
sin 2 = PB
PA sin 2
+ QC sin 2
QA
where  = 6 BAC ,
= 6 ABC ,  = 6 ACB .
Solution. Firstly, let O be on PQ. Denote Figure 5.
x = 6 AOP (Figure 5). By the sine theorem for
the triangles AOP , BOP , AOQ and COQ we nd
that
PA PO 
= sin(90
sin x  ; )

PB PO 
= sin(90
sin(2 ; x)  ; )

QA QO 
= sin(90
sin(180 ; x)  ;
)

QC QO :
= sin(90
sin(2
; 180 + x)  ;
)

Then
PB sin 2
+ QC sin 2 = sin(2 ; x) sin 2
; sin(2
+ x) sin 2 =
PA QA sin x sin x
1 ((sin 2 cos x ; sin x cos 2 ) sin2
; (sin 2
cos x + sin x cos 2
) sin 2 )
sin x
= ; sin 2
cos 2 ; sin 2 cos 2
= ; sin 2(
+  ) = sin 2:
Conversely, let the given equality be satised. Denote by Q0 0the intersection point of OP
and AC . It follows from the above that sin 2 = PA sin 2
+ Q0 A sin 2 . Then Q0A = QC
PB Q C Q 0C

QA .
Since the points Q0 and Q lie on the segment AC , then Q0 = Q.
Problem 3. Find all functions f (x) with integer values and dened in the set of the integers,
such that
3f (x) ; 2f (f (x)) = x
7
for all integers x.
Solution. The function f (x) = x satises the condition of the problem.
Let f (x) be a function which satises the condition. Let g (x) = f (x) ; x. The condition can
be written in the form
2f (f (x)) ; 2f (x) = f (x) ; x
which is equivalent to
g(x) = 2g (f (x)):
From here we obtain
g(x) = 2g (f (x)) = 22 g (f (f (x))) = 23g (f (f (f (x)))) = 24g (f (f (f (f (x))))) = : : ::
Since the numbers g (f (f : : : f (x)) : : :) are integer, then g (x) is divisible by 2n for all integers
x and all natural numbers n. This is possible only if g (x) = 0. Thus, f (x) = x is the only
solution of the problem.

8
Winter mathematics
competition|Burgas, 1997
Problem 8.1. Let F be the set of points with coordinates (x y)
such that jjxj ; jyjj + jxj + jyj = 2.
(a) Draw F .
(b) Find the number of points in F such that 2y = j2x ; 1j ; 3.

Solution: (a) If jxj  jyj, then jjxj ; jyjj + jxj + jyj = jxj ; jyj +
jxj+jyj = 2jxj = 2, thus jxj = 1 and therefore 1  jyj, so ;1  y  1.
We conclude that the segments ;1  y  1 on the lines x = 1 and
x = ;1 belong to F .
If jxj  jyj, then jjxj ; jyjj + jxj + jyj = ;jxj + jyj + jxj + jyj =
2jyj = 2, thus jyj = 1 and therefore 1  jxj, so ;1  x  1. We
conclude that the segments ;1  x  1 on the lines y = 1 and
y = ;1 also belong to F .
Thus we have determined that F consists of the sides of a square
with vertices A(;1 ;1) B (1 ;1) C (1 1) D(;1 1).

1
 (b) We nd the number of solutions of 2y = j2x ; 1j; 3 on each
of the segments AB BC CD DA.
The segment CD consists of all points (x y) such that ;1  x 
1, y = 1. The equation 2 = j2x ; 1j ; 3 has no solution x when
;1  x  1. Therefore 2y = j2x ; 1j ; 3 has no solution on CD.
The segment AB consists of all points (x y) such that ;1  x 
1, y = ;1. The equation ;2 = j2x ; 1j ; 3 has two solutions: x = 0
and x = 1. Therefore 2y = j2x ; 1j ; 3 has two solutions on AB .
As above we get that 2y = j2x ; 1j ; 3 has a unique solution
on AD: (x y) = (;1 0) and a unique solution on BC : (x y) =
(1 ;1). Note that the last one has already been obtained as a point
on AB . Thus there are three solutions of 2y = j2x ; 1j ; 3 in F :
(x y) = (;1 0) (0 ;1) (1 ;1).

Problem 8.2. Let H be the orthocentre of an acute triangle ABC .

Prove that the midpoints of AB and CH and the intersecting point
of the internal bisectors of 6 CAH and 6 CBH lie on a line.

2
Solution: Denote by M , P and N the midpoints of AB and CH
and the intersecting point of the internal bisectors of 6 CAH and
6 CBH (g. 1).

Let AA1 (A1 2 BC ) and BB1 (B1 2 AC ) be altitudes in 4ABC .

We show rst that M , N and P lie on the axis of symmetry l of A1B1.
From 6 CA1H = 6 CB1H = 90 we get PA1 = PB1 = CH 2 . Similarly,
from 6 AA1B = 6 AB1B = 90 we get MA1 = MB1 = AB 2 . Therefore
M 2 l and P 2 l. We prove now that 4NMA1  = 4NMB1.
Now 6 BAA1 = 90 ;  and 6 ABB1 = 90 ; , thus 6 NBB1 =
6 NBA1 = 1 6 A1 BB1 = 45 ;  . By analogy 6 NAA1 = 45 ;  ,
2 2 2
so 6 ANB = 180 ; 6 NAB ; 6 NBA = 90 . Therefore 4ABN is a
right triangle and MN = MA1 = MB1 = AB 2 .

Further 6 NMA1 = 6 NMB ; 6 A1MB = 90 ;  . By analogy

6 NMB1 = 90 ;  and therefore the considered triangles are iden-
tical. It follows now that NA1 = NB1, so N 2 l.

Problem 8.3. The n points A0 A1 : : : An;1 lie a circle in this
order and divide it into equal arcs. Find an ordering B0 B1 : : :  Bn;1
3
of the same points such that the length of B0B1 : : :Bn;1 is maximal.

Solution: Let rst n = 2k +1. Clearly a chord AiAj is of maximal

length if ji ; j j = k or k + 1. Cosider the following points:
A0 Ak  A2k Ak;1 A2k;1 Ak;2 A2k;2 : : : A1 Ak+1:
Since each segment is of maximal length, it follows that the length
of A0Ak A2kAk;1A2k;1Ak;2A2k;2 : : :A1Ak+1 is maximal.
Let now n = 2k. A chord AiAj is of maximal length if ji ; j j =
k. There are k such segments: A0Ak  A1Ak+1 : : : Ak;1A2k;1. The
second longest chord AiAj is obtained when ji ; j j = k ; 1 or k + 1.
Cosider the following points:
A0 Ak A2k;1 Ak;1 A2k;2 Ak;2 : : : Ak+1 A1:
It is easy to see that there are k segments of maximal length and
k ; 1 segments of the second greatest length. Trivially, this is the
required ordering.

Problem 9.1. Let  6=  be the roots of the equation x2 + px + q =

0. For any natural number n denote:
n ; n
an =  ;  :
(a) Find p and q such that for any natural number n the following
equality holds:
an+1 an+2 ; anan+3 = (;1)n :
4
 (b) Prove that for these p and q it is true that
an + an+1 = an+2
for any natural number n.
(c) Prove that for any natural number n, an is integer and if 3
divides n, then an is even.

Solution: (a) Since  and  are the roots of x2 + px + q = 0, we

know that  +  = ;p,  = q and therefore:
n+1 ;  n+1 n+2 ;  n+2 n ;  n n+3 ;  n+3
(;1)n =   ;  ; ; ; ;
= ( ;1  )2 ;( +  )( )n+1 + (3 +  3)( )n]
= p2 ;1 4q (pqn+1 ; p(p2 ; 3q)qn)
= p2 q; 4q (;p3 + 4pq) = ;pqn:
n

Thus
(1) pqn = (;1)n+1 :
It follows from (1) for n = 1 and n = 2 that pq = 1 and pq2 = ;1
and so p = ;1, q = ;1. Direct verication shows that p = ;1 and
q = ;1 satisfy (1) for any n. Also  6=  .
(b) Since  and  are the roots of x2 ; x ; 1 = 0, we know that
 =  + 1 and  2 =  + 1. Therefore
2

n ;  n n+1 ;  n+1 n (1 + ) ;  n(1 +  )

an + an+1 =  ;  + ; = ;
5
  n :2 ;  n: 2 n+2 ;  n+2
= ; =  ;  = an+2
which completes the proof of (b).
(c) Since a1 =  ; ;
 = 1 and a = 2 ;  2 =  +  = 1, it
2
;
follows by induction from
(2) an + an+1 = an+2
that an is integer for any n. From (2) we obtain:
an+3 = an+2 + an+1 = an+1 + an + an+1 = 2an+1 + an:
Observe that a3 = a1 + a2 = 2 is an even number. It is easy to see
now (again by induction) that an is even for n = 3k.

Problem 9.2. A pentagon ABCDE is inscribed in a circle. Let P

be the intersecting point of AC and BD, and let Q be the intersecting
point of AD and CE . Prove that if the triangles ABP , AEQ, CDP ,
CDQ and APQ have the same area, then ABCDE is a regular
pentagon.

Solution: It suces to prove that the sides of ABCDE are equal.

Since 4ABP and 4CDP have equal areas, so do the triangles ACD
and ADB . Therefore the quadrilateral ABCD (g. 1) is a trapezoid
inscribed in a circle, so AB = CD. By analogy from SAEQ = _SCDQ
we get AC kDE and AE = CD. Now, 6 PCQ = 6 ACE = AE 2 =
_
AB = 6 ADB = 6 PDQ. Therefore CDQP is inscribed in a circle.
2
6
On the other hand it follows from SCDP
_
= _SCDQ that CDQP is a
trapezoid. Thus 6 CDP = 6 DCQ )BC =DE and therefore BC =
DE . It remains to show that AB = BC . Consider the triangles
ABP and APQ, whose areas are equal. They have AP as a common
side and 6 AQP = 6 APQ = 6 ACD = 6 ABD = 6 ABP . It is easy to
see now that 4ABP  = 4APQ. Since 6 APB = 6 ADP + 6 PAQ >
6 PAQ,

it follows that 6 APB = 6 APQ = 6 AQP = 6 ABP . Therefore

_ _ _ _ _ _ _
AB + CD=AE + DE , so AB=DE =BC or AB = BC , which
completes the proof.

Problem 9.3. Given a rectangular table of 100 rows and 1997

columns. The table is lled with zeroes and ones in such a way
that there are at least 75 ones in any column. Prove that it is
possible to remove 95 rows in such a way that there is at most one
7
column consisting of zeroes in the remaining table (5 rows and 1997
columns).

Solution: We show rst that there is a row with at least 1498

ones. Assume the contrary. Denote by ai the number of ones in the
i-th row (i = 1 2 : : :  100) and by bi the number of ones in the i-th
X
1997 X
100
column (i = 1 2 : : :  1997). Now bi = ai. Note that the sum
i=1 i=1
on the left-hand side is at least 1997 75, whereas the sum on the
right-hand side is at most 1497 100, a contradiction.
Without loss of generality assume that the rst row begins by 1498
ones. Consider the table formed by the last 499 columns. As above
we prove that there is a row of this table (not necessarily distinct
from the rst one) with at least 375 ones. Let that be the second
row (if it is not the rst one) and let it begin with 375 ones in the
new table. Now consider the table formed by the last 124 columns.
Analogously, there exists a row having at least 93 ones. Let that be
the third row (if it is not the rst or the second one) and let it begin
with 94 ones in the new table. Consider next the table formed by
the last 31 columns and note that there exists a row that originally
has 24 ones. Finally consider the table formed by the last 7 rows
and note that there exists a row that originally has 6 ones.
We now have 5 rows (if there are fewer of them, we add arbitrary
rows). Remove the remaining 95 rows of the original table. Since
1498+375+93+24+6 = 1996, there is at most one column consisting
of zeroes.

8
  
Problem 10.1. Find all real numbers x such that tan 12 ; x ,
  
tan 12 and tan 12 + x form (in some order) a geometric progres-
sion.

Solution: Denote a = tan 12 and y = tan x. There are three cases
to consider:
   
1. tan 12 ; x tan 12 + x = tan2 12 . Now 1a+;ayy 1a;+ayy =
a2. Therefore a2 ; y2 = a2(1 ; a2y2), and so (a4 ; 1)y2 = 0.
Since a 6= 1 we get that y = 0, so tan x = 0. Obviously
all numbers x of the kind x = k, k 2 Zare solutions to the
problem.
    
2. tan 12 tan 12 + x = tan 12 ; x . We obtain
2

 !2
a + y a ; y
a 1 ; ay = 1 + ay =) (a2 + 1)yay2 + (a2 ; 1)y + 3a] = 0:
The case of y = 0 is settled in 1. Let y1 and y2 be the roots of
the equation ay2 + (a2 ; 1)py + 3a = 0. Since a = p tan 15 deg =
tan(45 deg ;30 deg) = 2 ; 3 we get y1 = y2 = 3, so tan x =
p
3. Obviously all x of the kind x = 3 + k, k 2 Zare solutions
of the problem.
    
3. tan 12 tan 12 ; x = tan 12 + x . The substitution z =
2

;x transforms this case to the previous one. Therefore x =

; 3 + k, k 2 Z.
9
The required numbers are x = k and x =  3 + k, k 2 Z.

Problem 10.2. Two points C and M are given in the plane. Let
H be the orthocentre of 4ABC such that M is a midpoint of AB .
(a) Prove that CH CD = jAM 2 ; CM 2j where D 2 AB and
CD ? AB:
(b) Find the locus of points H when AB is of given length c.

Solution: (a) It is easy to see that the equality holds if 4ABC

is a right triangle. If 4ABC is not a right triangle, then 4BDH
and 4ADC exist and 4BDH  4ADC . Therefore DH =
AD CD
BD
and so
(1) CD DH = AD BD
There are three cases:
Case 1: 4ABC is an acute triangle (g. 1).
Case 2: 4ABC is obtuse triangle and 6 C > 90 (g. 2).
Case 3: 4ABC is obtuse triangle and 6 A > 90 or 6 B > 90
(g. 3).
In cases 1 and 2 it follows from (1) that
CD DH = (AM DM )(AM DM ) = AM 2;DM 2 = AM 2;(CM 2;CD2)

10
 so
(2) AM 2 ; CM 2 = CD(DH ; CD):
In case 1 we have CD = CH + DH and from (2) we get
AM 2 ; CM 2 = CD(DH ; CH ; DH ) = ;CH CD:
In case 2 we have CD = DH ; CH and from (2) we get
AM 2 ; CM 2 = CD(DH ; DH + CH ) = CH CD:
In case 3 it follows from (1) that CD DH = (DM AM )(DM 
AM ) = DM 2 ; AM 2 = CM 2 ; CD2 ; AM 2 and therefore CM 2 ;
AM 2 = CD(DH + CD). Now DH = CH ; CD and so CM 2 ;
AM 2 = CD(CH ; CD + CD) = CH CD. Summarising all three
cases considered (and the case of a right triangle 4ABC ) we get
CH CD = jAM 2 ; CM 2j:

11
 (b) It follows from the conditions of the problem that A and B
are diametrically opposite in the circle (k) of centre M and radius
c . When the diameter AB varies in (k) we obtain all triangles ABC
2
with the xed vertex C . There are three cases:

Case 1: 2c = jCM j (g. 4) Case 2: 2c > jCM j (g. 5) Case 3:

c < jCM j (g. 6).
2
In case 1 C lies on (k) and triangles ABC are right triangles and
their orthocentres coincide with C . In this case the locus consists of
point C .

12
 Consider now case 2. Let AB be a diameter in (k) perpendicular
to CM . The orthocentre H of 4ABC lies on line CM . Moreover
H is an external point for (k) because 4ABH is acute. The last
follows from 6 AHB = 180 deg ;6 ACB and 6 ACB > 90 deg.
Let A1B1 be a arbitrary diameter of (k) such that C 2= A1B1:
Denote by H1 the orthocentre of A1B1C and by D1 the foot of the
altitude from the vertex C . Since 4A1B1C is obtuse, we get that
H1 6= C . Consider 4MD1C and 4CHH1 and use a). We are in
case 2 of (a). For 4ABC we obtain that CH CM = c4 ; CM 2 and
2

for 4A1B1C that CH1 CD1 = c4 ; CM 2. Therefore CH CH = CD1 ,

2

CM
so 4MD1C  4CHH1. Since 4MD1C is a right triangle, we
1

know that H1H ?CM and so H1 lies on a line l through H which is

perpendicular to CM . Let H1 6= H be an arbitrary chosen point of
l and let A1B1 be a diameter of (k) such that AB ? CH1.
Since H is an external point for (k) and l?CM through H , it
follows that all points of l are external for (k). In particular H1 6= C
and therefore A1B1 exists. The orthocentre of 4A1B1C lies on CH1

13
and it follows from the above that it lies on l. Since H1 is the
intersecting point of these two lines it is obvious that H1 is the
orthocentre of 4A1B1C .
Therefore the locus is a line l perpendicular to CM through a
point H on the ray opposite to CM~ and of distance 4jCM c2 ;jCM j
j
from point C .
Consider case 3. Let AB be a diameter of (k) perpendicular to
CM . The orthocentre H of 4ABC lies on CM but now H is an
internal point for (k). Consider an arbitrary diameter A1B1 of (k)
such that H 2 A1B1. Let H1 be the orthocentre of 4A1B1C and
CD1{ the altitude. Consider 4MD1C and 4CHH1. It is essential
to show that 4CHH1 is uniquely determined (for 4MD1C it is
obvious). It suces to prove that H1 6= C . If the contrary is true,
then 4A1B1C is a right triangle with its right angle at C and since
A1B1 is a diameter, C must lie on (k), a contradiction. Since 4ABC
is acute, we apply case 1 of (a), so CH CM = CM 2 ; c4 . There
2

are three cases for 4A1B1C :

1. if it is an acute triangle, we apply case 1 of (a) and therefore
CH1 CD1 = CM 2 ; c4 
2

2. if it is an obtuse triangle, we apply case 3 of (a) and we get the

same equality (note that 4A1B1C is not obtuse at C because
C is an external point for (k))
3. if it is a right triangle (say at B1) then D1  B1  H1 and
therefore CH1 CD1 = CB12 = CM 2 ; c4 .
2

14
The further considerations follow those from case 2. We conclude
that the locus is a line perpendicular to CM and passing through
H 2 CM~ of distance jCM j ; 4jCM c2 from C .
j
In particular it follows that PQ?CM , a well-known property.

Problem 10.3. How many natural numbers a1a2 : : : a2n exist such
that:
(a) none of the digits ai is zero
(b) the sum a1a2 + a3a4 + : : : + a2n;1a2n is an even number?

Solution: Denote the required number by An. The product a2i;1a2i

is even if at least one of the digits a2i;1 and a2i is even. Therefore
there are 5 4 + 4 5 + 4 4 = 56 choices for a2i;1 and a2i such that
a2i;1a2i is even. Similarly, a2i;1a2i is odd when both a2i;1 and a2i
are odd. There are 5 5 = 25 choices for a2i;1 and a2i such that
a2i;1a2i is odd. Thenumber
 i n;ofi a1a2 : : :a2n such that
P inof the items
a2i;1a2i are odd n
is i 25 56 : Therefore An = i 2i 25 56n;2i . 2 i

Let Bn = Pi 2in+1 252i+1 56n;2i;1 . Obviously
n n!
X
An + Bn = 25i 56n;i = (56 + 25)n = 81n 
i
i=0
n  !
An ; Bn =
X n 25i 56n;i = (56 ; 25)n = 31n :
(;1)i
i=0 i
Thus 2An = 81n + 31n and An = (81n + 31n )=2.

15
Problem 11.1. The sequence fxng1n=1 is dened as:
x1 = 3 xn+1 = x2n ; 3xn + 4 n = 1 2 3 : : :
(a) Prove that fxng1
n=1 is monotone increasing and unbounded.

(b) Prove that the sequence fyn g1n=1 dened as yn = x 1; 1 +

1
1 + + 1 , n = 1 2 3 : : :, is convergent and nd
x ;1
2 x ;1
n
its limit.

Solution: (a) xn+1 ; xn = (xn ; 2)2  0 and so fxng1n=1 is a

monotone increasing function. We now prove by induction that xn 
n + 2. Obviously this equality holds for n = 1. Suppose it is true
for n = k  1. Then
xk+1 = xk (xk ; 3) + 4  (k + 2)(k ; 1) + 4  k + 3:
Therefore xn  n + 2 when n = 1 2 3 : : : and so the sequence is
unbounded.
(b) It follows from the recursive denition of our sequence that
xk+1 ; 2 = (xk ; 1)(xk ; 2). Hence
1 = 1 = 1 ; 1 
xk+1 ; 2 (xk ; 1)(xk ; 2) xk ; 2 xk ; 1
so 1 = 1 ; 1 :
xk ; 1 xk ; 2 xk+1 ; 2
By adding the above equalities for k = 1 2 : : :  n we get
yn = x 1; 2 ; x 1 ; 2 = 1 ; x 1 ; 2 :
1 n+1 n+1

16
Since 0  x 1 ; 2  n1 it follows that nlim 1
!1 xn+1 ; 2
= 0, so
n+1
nlim y = 1.
!1 n

Problem 11.2. Given 4ABC such that 6 ABC  60 deg and

6 BAC  60 deg. Let BL (L 2 AC ) be the internal bisector of
6 ABC and AH H 2 BC be the altitude from A. Find 6 AHL if
6 BLC = 36 AHL.

Solution: Denote  = 6 AHL,  = 6 BAC ,  = 6 ABC and  =

6 ACB . From the Sine Law for 4AHL and 4CHL we get
sin  =
AL
sin 6 ALH and sin(90 deg ;) = sin 6 CLH . Since CL = BC =
AH CL CH AL BA
sin  and AH = tan  (from 6 ACB < 90 deg it follows that H lies
sin  CH
on a ray CB~), we get
(1) cos  = sin  :
sin  cos 
It follows from the conditions of the problem that 3 =  + 2 , so
6 ;  =  +  . Thus cos  = ; cos(6 ; ) and (1) is equivalent to
cos  cos(6 ; ) + sin  sin  = 0 () cos(7 ; ) + cos(5 ; ) +
cos( ; ) ; cos( + ) = 0 () ; sin 4 sin(3 ; ) + cos(3 ;
) cos 2 = 0 () cos 2(2 sin 2 sin(3 ; ) ; cos(3 ; )) = 0.
There are two cases to consider:
1. cos 2 = 0. From 0 <  < 90 deg it follows that  = 45 deg.
(Working backwards we get that 135 deg = 6 BLC = 36 AHL
17
 for any 4ABC such that  + 2 = 135 deg).

2. cos 2 6= 0. It follows from 3 ;  = 2 that

(2) 2 sin 2 = cotg 2 :

But 60 deg   < 180 deg and so
p
(3) cotg 2 < cotg30 deg = 3:
On the other hand, also from the conditions of the problem we
get 180 deg > 3 =  + 2  90 deg, so 30 deg   < 60 deg.
Therefore
p
(4) 2 sin 2  2 sin 60 deg = 3:
The inequalities (3) and (4) show that (2) holds only if  =
30 deg. In this case  =  =  = 60 deg. (It is obviously true
that 90 deg = 6 BLC = 36 AHL for any equilateral 4ABC ).

Problem 11.3. Find all integer numbers m n  2 such that

1 + m3n + m2:3n
n
is integer.

18
Solution: Let m and n satisfy the conditions of the problem. Since
n is an odd integer number, we get (m n) = 1 and n > 2. When
n = 3, all m  4 such that m  1(modn 3) are solutions, because
if m  ;1(mod 3), then 1 + m +n m23  1 ; 1 + 1  1(mod 3).
3n

Let nown n > 3. nIt follows that m3 6 1(mod n), because otherwise
1 + m3 + m23 n+13(mod n), i. e., n=3. On the other hand 1 +
m3n + m23n = mm3n ;;11 and therefore m3n+1  1(mod n). Let k
3

be the least natural number such that mk  1(mod n). Further,

k=3n+1 and k ;= 3n , so k = 3n+1 . Let '(n) be Euler's function. From
(m n) = 1 it follows that m'(n)  1(mod n), so k  '(n). Therefore
3n+1  '(n)  n ; 1, which is impossible.
The required numbers are: n = 3 and all m  4 such that
m  1(mod 3).

19
Winter mathematics
competition|Pleven, 6{8
February 1998
Dedicated to the One Hundredth Anniversary of the UBM

Problem 8.1. Let three numbers a, b and c be chosen so that

a = b = c.
b c a
1
 a.) Prove that a = b = c.
b.) Find the sum x + y if 3xy = 2x ;y 5y = 6x ;x 15y and the
expression ;4x2 + 36y ; 8 has its maximum value.

Solution:
a.) It is obvious that a 6= 0, b 6= 0, c 6= 0. The rst equality
gives b2 = ac, whence by multiplying both sides by b we get
b3 = abc. Similarly a3 = abc and c3 = abc. Hence a3 = b3 = c3
and therefore a = b = c.
b.) By multiplying both the numerator and the denominator of
the second fraction by 3 and using the result of a.) we obtain
x = 3y. Thus ;4x2 + 36y ; 8 = ;9(4y2 ; 4y + 1) + 1 =
;9(2y ; 1)2 + 1, and its maximum value is 1 when 2y ; 1 = 0.
Therefore y = 21 and x = 23 , i. e., x + y = 2.

Problem 8.2. In the acute triangle 4ABC with 6 BAC = 45 deg,

BE (E 2 AC ) and CF (F 2 AB ) are altitudes. Let H , M and
K be the orthocentre of ABC and the midpoints of BC and AH ,
respectively.
a.) Prove that the quadrangle MEKF is a square.
b.) Prove that the diagonals of the quadrangle MEKF intersect
at the midpoint of OH , where O is the circumcentre of 4ABC .
c.) Find the length of EF when the circumradius of 4ABC is 1.

2
Solution:
a.) The segments EM and FM are medians to the hypotenuses
of 4BCE and 4BCF and therefore EM = FM = 21 BC .
Similarly, for 4AHE and 4AHF we get EK = FK = 12 AH .
Since 6 BAC = 45 deg, we nd that 4AEB and 4CEH are
isosceles. Hence AE = BE and EC = EH , i. e., 4AHE  =
4BCE . Therefore EK = EM . Thus MEKF is a rhombus.
Furthermore,
6 MEK = 6 MEB + 6 HEK = 6 CBE + 6 HEK

= 6 EAH + 6 HEK = 6 EAH + 6 AHE = 90 deg

i. e., the quadrangle is a square.

b.) It follows from a.) that the intersecting point S of the diagonals
of the quadrangle MEKF is the midpoint of both diagonals.
Since 4AEB is isosceles, E lies on the axis of symmetry of the
segment AB and therefore EO?AB , i. e., EOkHF . Similarly
FOkEH . Thus the quadrangle EOFH is a parallelogram.
From the above we conclude that S is the midpoint of OH .
3
 c.) a.) implies that in the acute triangle 4ABC with orthocentre
H and 6 BAC = 45 deg it is true that AH = BC . 4AFE
is of the same type and therefore EF = AO = 1. (It follows
from b.) that O is orthocentre of this triangle.)

Problem 8.3. Let 1998 points be chosen on the plane so that out of
any 17 it is possible to choose 11 that lie inside a circle of diameter 1.
Find the smallest number of circles of diameter 2 sucient to cover
all 1998 points.
(We say that a circle covers a certain number of points if all
points lie inside the circle or on its outline.)

Solution: Consider a regular hexagon with a side of length 3.

Choose 1998 points as follows: the 6 vertices of the hexagon and
1992 points inside a circle of diameter 1 centred at the centre of the
hexagon. It is clear that the above 1998 points satisfy the condition
of the problem. Moreover any circle of radius 1 covers at most one of
the vertices of the hexagon. Therefore the required number is no less
than 7 (in our case: 6 circles for each vertex and a single circle for
the remaining points).
Now we shall prove that the required number is no greater than 7.
Arbitrarily choose 8 points and add other 9, for a total of 17. It is
clear that there is a circle of diameter 1 covering at least 11 of these
17 points. At most 6 points lie outside the circle and therefore
at least 2 of the initially chosen 8 points lie inside the circle. The
distance between these two points is no greater than 1. We have
proved that among any 8 points there always exist 2 such that the
distance between them is no greater than 1.

4
 Now choose a circle of radius 1 centred in one of the points. If
the remaining points lie inside the circle, the required number is 1
and thus no greater than 7. If this is not the case, take another
point outside the rst circle. If all points lie in the two circles, then
the required number is 2 and thus no greater than 7. Continuing in
this way we either obtain no more than 7 circles covering all points
or have 7 circles and a point that lies outside all circles. Consider
this point and the centres of the chosen circles. There exist 2 points
among these 8 such that the distance between them is no greater
than 1. But this is impossible because of the way we chose our
points.
Together the two parts of the proof demonstrate that the required
number is 7.

Problem 9.1. Find all quadratic functions f (x) = x2 ; ax + b with

integer coecients such that there exist distinct integer numbers m,
n, p in the interval 1 9] for which jf (m)j = jf (n)j = jf (p)j = 7.

Solution: Let f (x) be a function satisfying the conditions of the

problem. Such a function cannot take one and the same value
for three dierent arguments (otherwise we would have a quadratic
equation having three distinct roots). Therefore two of the numbers
f (m), f (n) and f (p) equal 7 (or ;7) and the third one equals ;7
(or 7).
Case 1. f (m) = f (n) = 7, f (p) = ;7. Without loss of
generality we may assume that m > n. Since m, n are roots of
x2 ; ax + b ; 7 = 0, we obtain that a = m + n, b = mn + 7.

5
Subtracting the two equalities
m2 ; am + b = 7
p2 ; ap + b = ;7
we nd
14 = m2 ; p2 ; a(m ; p) = (m ; p)(m + p ; a) = (m ; p)(p ; n):
Thus the numbers m ; p and p ; n are either both positive or both
negative and since m > n, they are positive. Moreover they are
integer and therefore are equal to 1 and 14 or to 2 and 7. But since
m n p 2 1 9] it follows that neither m ; p nor p ; n is 14. There
are two cases to consider: m ; p = 2, p ; n = 7 and m ; p = 7,
p ; n = 2, i. e., either m = p + 2, n = p ; 7 or m = p + 7, n = p ; 2.
It is obvious that in both cases at least one of m n p lies outside the
interval 1 9].
Case 2. f (m) = f (n) = ;7, f (p) = 7. As in Case 1 we get
a = m + n, b = mn ; 7 and (m ; p)(p ; n) = ;14. Using similar
arguments we obtain that either m ; p = 2, p ; n = ;7 or m ;
p = ;2, p ; n = 7. (Without loss of generality we suppose that
jm ; pj < jp ; nj.) Therefore the two options are m = p + 2,
n = p + 7 and m = p ; 2, n = p ; 7. Simple calculations show that
all triples (m n p) satisfying the conditions are (3 8 1), (4 9 2),
(6 1 8) and (7 2 9). So the functions are f (x) = x2 ; 11x + 17,
f (x) = x2 ; 13x + 29, f (x) = x2 ; 7x ; 1 and f (x) = x2 ; 9x + 7.

Problem 9.2. Three points A1, B1 and C1 lie on the sides BC ,

CA and AB of 4ABC so that AB1 = C1B1 and BA1 = C1A1. Let
D be the reexion of C1 in A1B1 (D 6= C ). Prove that the line CD
is perpendicular to the straight line through the circumcentres of
4ABC and 4A1B1C .
6
Solution: It suces to prove that D is the second intersecting
point of the two circumcircles. We know that
6 A1DB1 = 6 A1C1 B1 = 180 deg ;6 BC1A1 ; 6 AC1B1

= 180 deg ;6 C1BA1 ; 6 C1AB1 = 6 A1CB1:

On the other hand, A1D = A1C1 = A1B and B1D = B1C1 =

B1A, which shows that A1 and B1 are the circumcentres of 4BC1D
and 4AC1D. Therefore 6 ADB = 6 ADC1 + 6 BDC1 = 12 6 AB1C1 +
2 BA1C1 = 90 deg ; C1AB1 + 90 deg ; C1BA1 = ACB . Since C
16 6 6 6
and D lie in one and the same semiplane in regard to both A1B1 and
AB , it follows from 6 A1DB1 = 6 A1CB1 and 6 ADB = 6 ACB that
D is the second intersecting point of the circumcircles of 4A1B1C
and 4ABC . This completes the proof.

Problem 9.3. All natural numbers from 1 to 1998 inclusive are

written 9 times (so that there are 9 ones, 9 twos and so on) in the
7
cells of a rectangular table with 9 rows and 1998 collumns, so that
the dierence between any two elements lying in one and the same
column is no greater than 3. Find the maximum possible value of
the smallest sum amongst all 1998 sums of the elements lying in one
and the same column.

Solution: Consider the placement of the ones in the columns. If

they are all in a single column, then the minimum sum of elements
lying in one column is 9. Let all ones lie in exactly 2 columns.
Therefore there are at least 5 ones in a single column and thus the
minimal sum is no greater that 5  1 + 4  4 = 21. If all ones are
placed in exactly 3 columns, then the sum of all numbers in these
three columns is at most 9  1 + 9  4 + 9  3 = 72. Hence the minimal
sum is at most 72 : 3 = 24. If all ones are placed in exactly 4
columns, then the sum of all numbers in these columns is at most
9  1 + 9  4 + 9  3 + 9  2 = 90 and therefore the minimal sum is
at most 90 : 4, i. e., 22. It is impossible to have ones in more than 4
columns, because in that case the total number of 2s, 3s and 4s does
not suce to ll the remaining cells. Therefore the required sum is
at most 24.
1 1 1 2 2 2 7 8 . . . 1998
1 1 1 2 2 2 7 8 . . . 1998
1 1 1 2 2 2 7 8 . . . 1998
The following example 3 3 3 5 5 5 7 8 . . . 1998
shows that this sum can 3 3 3 5 5 5 7 8 . . . 1998
be 24, consequently the 3 3 3 5 5 5 7 8 . . . 1998
answer is 24: 4 4 4 6 6 6 7 8 . . . 1998
4 4 4 6 6 6 7 8 . . . 1998
4 4 4 6 6 6 7 8 . . . 1998

8
Problem 10.1. Find all values of the real parameter a for which
the equation x3 ; 3x2 +(a2 +2)x; a2 = 0 hasthree distinct roots
 x1,
x2 and x3 such that sin 23 x1 , sin 23 x2 and sin 23 x3 form
(in some order) an aritmetic progression.

Solution: Since x3 ; 3x2 + (a2 + 2)x ; a2 = (x ; 1)(x2 ; 2x + a2),

in order for there to be three distinct real roots itpis necessary that
D = 1 ; a2 > 0. Therefore a2 < 1 and thus p 1 2 1 ; a2 >p0. The2
roots of our equation are x1 = 1, x2 = 1+ 1 ; a , x3 = 1 ; 1 ; a .
It follows now that x2 + x3 = 2 and 2  x2 > 1 and 1 > x3  0.
There are two cases to consider: 
2
1. The second term of the progression is sin 3 x1 . Then
2   2   2 
sin 3 x2 + sin 3 x3 = 2 sin 3
  
2 sin 23 x2 +2 x3 cos 23 x2 ;2 x3 = 2 sin 23

cos 3 (x2 ; x3) = 1:
p
But 3 jx2 ; x3j = 23 1 ; a2 23 , and hence 3 (x2 ; x3) 2
2  2   
; 3  3 ]. Therefore cos 3 (x2 ; x3) = 1 when x2 = x3, which is
impossible, since the roots are distinct.
2  
2. The rst or the third term of the progression is sin 3 x1 .
Then   
sin 23 + sin 23 xi = 2 sin 23 (2 ; xi)
9
for i = 2 or 3. Hence
  
sin 23 + sin 23 xi = 2 sin 43 cos 23 xi ; 2 cos 43 sin 23 xi :
 
After simple calculations we get cos 3 xi = ; 21 . From the re-
2
strictions for x2 and x3 we obtain xi = 1 or xi = 2. In the rst case
a2 = 1, which is impossible, and in the second case x2 = 2 x3 = 0
and a2 = 0.
Thus a has a unique value and it is a = 0.

Problem 10.2. A point C lies on the periphery of a circle. Two points

A and B are chosen anticlockwise away from C such that if 6 CAB =
 and 6 CBA =  , the following equality holds:
   
2 cos 2 +  = sin 2 ;  :
Prove that the bisectors of 6 CAB pass through a xed point.

Solution: It is easy to see that  = 90 deg,  = 45 deg and tan 2 =

1 ,  = 90 deg satisfy the condition for  and  . Therefore the
2
required point is the midpoint of OE {F (g. 1).
From the premises of the problem we obtain
2 cos 2 cos  ; 2 sin 2 sin  = sin 2 cos  ; cos 2 sin 

10
and after dividing by cos 2 cos  (cos 2 6= 0 (why?), and if cos  = 0,
i. e.,  = 90 , we have one of the two cases already considered)
we obtain
tan 2 (1 + 2 tan  ) = 2 + tan 
It follows in particular that if  is xed, then  is uniquely deter-
mined.
Suppose that tan 2 > 2. Thus 2 > 45 and therefore  > 90.
If tan  < 0, we get  > 90 , which is impossible, since  +  < 180.
If tan  > 0, we get 2 + tan  > (1 + 2 tan  ) 2 i. e., tan  < 0, which
is a contradiction.
Therefore tan 2 2 and B lies on CED where 6 GCF = 6 FCD
and tan 6 GCF = 2 (g. 1).
Fix the point B such that  < 90. Let T be the midpoint of the
arc CB and let A1 be the intersecting point of TF and the circle.
We shall show that A1
A. We obtain 6 OA1F = 2 ; (90 ;  ) =
 + 2 ; 90 and 6 A1FO = 45 ; 2 +  ; 45 =  ; 2 . It follows from

11
  
sin  ; 2
the Sine Theorem for 4A1FO that    = 2, which is
   sin 
 2+ ; 90 

equivalent to 2 cos 2 +  = sin 2 ;  . Therefore A1

A.
The case of  > 90 can be dealt with by analogy. The condition
for B to lie on CED shows that A1 lies between C and B (g. 3).

Problem 10.3. Let n be a natural number. Find the number of

sequences a1a2 : : : a2n, where ai = +1 or ai = ;1 for i = 1 2 : : :  2n,
such that  
 X 2l 
 ai 2
i=2k;1 
for all k and l for which 1 k l n.

Solution: It is clear that a sequence having a2k;1 + a2k = 0 for

1 k n satises
P the condition of the problem, because any sum
l
k;1 ai equals zero. There are 2 such sequences.
of the form 2i=2 n
Let us determine the number of sequences such that there exists
a k for which a2k;1 + a2k 6= 0. Let k1 k2 : : : ks be all k with the
above property. It is easily seen that if a2k ;1 + a2k = 2 (;2), then
a2k +1;1 + a2k +1 = ;2 (2). Therefore all sums a2k ;1 + a2k (and so
i i

i i i i

also a2k ;1a2k ) are uniquely determined by a2k1;1 + a2k1 (there are
i i

two possibilities for a2k1;1 a2k1 ). There are two possibilities for any
of the remaining n ; s pairs (for which a2t;1 + a2t = 0). Therefore
there are
 !  !  !  !  !
;1 n ;2 n ;k n n n :
2 +22 1 +22 2 +  +22
n n n n
k + : : : +2  2 n;1 +2  n

12
sequences with the required property. By adding and subtracting 2n
to and from the above expression we get:
  !  !  !  !  !!
2: 2 0 +2 1 +2 2 +    +2 n ; 1 + nn ;2n =23n ;2n
n n n;1 n n ;2 n n

Thus there are 2  3n ; 2n sequences.

Problem 11.1. Consider the function f ( x) = px + px ; 4 ;

p p
x ; 1 ; x ; 3, x  4.
a.) Find xlim
!1
f (x).
b.) Prove that f (x) is an increasing function.
q x;3
c.) Find the number of real roots of the equation f (x) = a x ,
where a is a real parameter.

Solution: a.) By grouping the rst and third radicals and the
second and fourth radicals and rationalising we get that when x > 4,
f (x) = px+p1 x;1 ; px;4+1 px;3 . Therefore xlim
!1
f (x) = 0.
b.) When x > 4,
f 0(x) = 2p1 x ; p 1 ; p 1 + p 1
2 x;1 2 x;3 2 x;4
= p p 1
p p +
2 x;3 x;4( x;3 + x;4)
; 2pxpx;1(1px+px;1) > 0:
13
Therefore f (x) is an increasing function if x  4.
c.) It follows from a.) and b.) that f (x) < 0 when x  4, i. e.,
the equation couldq have a q solution only if a < 0. Let a < 0. The
function g(x) = a x = a 1 ; x3 is decreasing and continuous and
x ;3

xlim
!1
g(x) = a < 0. Since f (x) is an increasing and continuous func-
tion and xlim
!1
f (x) = 0, in accordance with the Bolzano{Weierstra
Theorem the equation f (x) = g(x) has a solution
p (and it is a unique
one) exactly when f (4) g(4), i. e., if 2(1 ; 3) a < 0.

Problem 11.2. The convex quadrangle ABCD is inscribed in a

circle with centre O. Let E be the intersecting point of AC and BD.
Prove that if the midpoints of AD, BC and OE lie on a straight line,
then AB = CD or 6 AEB = 90 deg.

Solution: It suces to prove that if 6 AEB 6= 90 deg, then AB =

CD. Let 6 AEB = 6 90 deg. If O
E , then ABCD is a rectangle
and therefore AB = CD. Suppose O 6
E . Let M , N , P , Q be the
midpoints of AD, BC , AC , BD, respectively, and R be the inter-
secting point of the straight lines through P and Q perpendicular
to BD and AC , respectively. It is clear that MPNQ and OPRQ
are parallelograms. Therefore the midpoints of MN and OR coin-
cide with the midpoint of PQ, and since the midpoint of OE lie
on MN , we get that RE kMN . On the other hand R is the ortho-
centre of 4PQE and therefore RE ?PQ. Hence MN ?PQ, i. e.,
the parallelogram MPNQ is a rhombus. It is easy to see now that
AB = 2PN = 2NQ = CD, which solves the problem.

14
 Note: The above solution shows that if O is the intersecting
point of the axes of symmetry of AC and BD, then the assertion of
the problem and its opposite are true for a quadrangle that is not
inscribed in a circle. This could be demonstrated by using complex
numbers or trigonometry.

Problem 11.3. Let famg1m=1 be a sequence of integer numbers

such that their decimal representations consist of even digits (a1 = 2,
a2 = 4, a3 = 6,. . . ). Find all integer numbers m such that am = 12m.

15
Solution: Let m be an integer number such that m = b0 + b1  5 +
   + bn  5n . Denote f (m) = 2b0 + 2b1  10 +    + 2bn  10n . It is clear
that ff (m) j m 2 Ng is the set of integer numbers with only even
digits in their decimal representation. Since f (m1) < f (m2) ()
m1 < m2, it follows that am = f (m) for any m. Therefore it suces
to nd all m such that
12(b0 + b1  5 +    + bn  5n ) = 2b0 + 2b1  10 +    + 2bn  10n 
i. e.,
(1) 6(b0 + b1  5 +    + bn  5n ) = b0 + b1  10 +    + bn  10n :
Since b0+b15+  +bn5n 5n+1 ;1 and b0+b110+  +bn 10n  10n ,
it follows from (1) that 6(5n+1 ; 1)  10n , i. e., 6  5n+1 > 10n . Thus
2n < 30 and therefore n 4. If n = 4, we get from (1) that
b0 + 4b1 + 10b2 = 50b3 + 1250b4  1250, which is imposible. In the
same way it is easy to show that n  3, i. e., n = 3. In this case
b0 +4b1 +10b2 = 50b3. Obviously b3 = 1 and b0 = b1, because b0 ;b1 is
divisible by 5. As a result we have the equation b0 +2b2 = 10, and its
solutions are b0 = 2 b2 = 4 and b0 = 4 b2 = 3. Therefore all integer
numbers m with the required property are m = 2+2  5+4  52 +53 =
237 and m = 4 + 4  5 + 3  52 + 53 = 224.

16
Winter mathematics
competition|Varna, 1999
Problem 8.1. Find all natural numbers x and y such that:
a) x1 ; y1 = 13 

b) x1 + y1 = 13 + xy
1.

Solution: a) The equation is equivalent to 3y ; 3x = xy, so

x = y 3+y 3 = 3y y++9 3; 9 = 3 ; y +9 3 . Therefore y + 3 = 9 and thus
y = 6. Hence there is a unique solution x = 2 y = 6.
b) Let x  y be a solution of the problem. Now 31 = x1 + y1 ;
1  2 ; 1 = 2y ; 1 < 2y = 2 , giving x < 6. When x = 1, x = 2
xy y xy xy xy x
or x = 3, no solution exists. When x = 4, it follows that y = 9, and
x = 5 implies y = 6. If y  x, we apply the same reasoning. The

1
problem has four solutions:
x = 4 y = 9 x = 5 y = 6 x = 6 y = 5 x = 9 y = 4:

Problem 8.2. Given an acute 4ABC with centroid G and bisec-

tors AM (M 2 BC ) BN (N 2 AC ) CK (K 2 AB ). Prove that one
of the altitudes of 4ABC equals the sum of the remaining two if
and only if G lies on one of the sides of 4MNK .
Solution. We shall repeatedly use
the following property: A segment con-
necting a vertex of a triangle with a
point on the opposite side divides the
triangle into two triangles such that the
ratio of their areas equals the ratio of
the parts into which the point divides
the side.
Let G 2 MN and G1 , G2, G3 be the projections of G on BC , AC
and AB , respectively. Further, denote the projections of N on BC
and AB by N1 and N3 and those of M on AC and AB by M2 and M3.
We shall prove that GG3 = GG1 + GG2 and from the above property
it will follow straightforwardly that the altitude from C is equal to
the sum of the remaining two altitudes. We obtain
GG1 = CM  GG1 = SGMC = GM :
NN1 CM  NN1 SNMC NM
By analogy MM GG2 = GN , implying that GG1 + GG2 = 1, so
2 NM NN1 MM2
(1) GG1  MM2 + GG2  NN1 = MM2  NN1:

2
Let M4 and G4 be the projections of M and G on NN3 . It easily
follows now that
NN3 ; GG3 = GN = GG2 :
NN3 ; MM3 NM MM2
Further, using that NN3 = NN1 and MM3 = MM2, we obtain
GG3  MM2 = GG1  MM2 + GG2  MM2 and therefore GG3 =
GG1 + GG2.
Conversely, let the altitude through C be the sum of the remain-
ing two. Now GG3 = GG1 + GG2. If G? = GG3 \ MN , then it
follows straightforwardly that the sum of the distances from G? to
AC and BC equals to G?G3. It is easy to check now that G?  = G.

Problem 8.3. Let n be a natural number. Find all integer values

of m such that k = 2m;2 is integer and A = 1999k + 6 is a sum
of the squares of n integers (not necessarily distinct and dierent
from zero).

Solution: It is sucient to consider only nonnegative values of m.

1) n = 1 and k = m is integer only if m = 4p and m = 4p + 2.
2
If m = 4p, then A = (2  1000 ; 1)2p + 1 and it follows by induction
that A is of the form A = 4a + 7, so A is congruent to 3 modulo 4.
We conclude that A is not a perfect square. If m = 4p + 2, then
A = (25  80 ; 1)2p+1 + 6 and it follows by induction that A is of
the form A = 25a + 5. Therefore A is not a perfect square, because
5 divides A, but 25 does not.
2) n = 2 and k = m is integer for any m. Now A = 1999m + 6.
When m = 0, we get A = 7, which is not a sum of two squares.
3
When m = 1, we obtain A = 2005 = 412 + 182 . If m  2, then
A = (2  999 + 1)m + 1 and as above A is congruent to 3 modulo 4.
On the other hand the sum of two perfect squares is congruent to
0, 1 or 2 modulo 4 and so no solution exists in this case.
3) n = 2 and k = m is integer for any m. Now A = (8  250 ;
1) + 6, which can be written in the form A = 8a + 7. Therefore A
2 m
is congruent to 7 modulo 8, whereas a sum of three perfect squares
is congruent to 0, 1, 2, 3, 4, 5 or 6 modulo 8. Thus no solution exists
in this case.
4) n  4 and k is integer for ;
any m. Now A = (1999m2n;3 )2 +
22 + 12 + 12 and if a1 = 1999m2  a2 = 2  a3 = a4 = 1 a5 = a6 =
n 3

   = an = 0, then A = a21 + a22 +    + a2n.

Answer: if n = 1, no solution exists
if n = 2, there is an unique solution m = 1
if n = 3, no solution exists
if n  4, any m  0 is a solution.

Problem 9.1. Let p be a real parameter such that the equation

x2 ; 3px ; p = 0 has real and distinct roots x1 and x2.
a) Prove that 3px1 + x22 ; p > 0.
b) Find the least possible value of
A = 3px +px2 + 3p + 3px2 +px2 1 + 3p :
2 2

1 2

When does equality obtain?

4
 Solution: a) It follows from the equation that x22 = 3px2 + p and
so 3px1 + x22 ; p = 3p(x1 + x2) = 9p2 > 0. The inequality is strict
because otherwise x1 = x2 = 0.
b) As in a), we obtain 3px1 + x22 + 3p = 3px2 + x21 + 3p =
9p2 + 4p > 0 (the last inequality follows from the conditions of the
problem x1 and x2 to be distinct and real, giving p 6= 0). Therefore
A = 9p2p+ 4p + 9p p+2 4p  2
2 2

(from the Arithmetic{Geometric Mean Inequality) and equality ob-

tains when 9p2 + 4p = p2, i. e., when p = ;1=2.

Problem 9.2. Given an acute 4ABC such that AC > BC , let M

be the midpoint of AB and let CD, AP and BQ be the altitudes.
Denote the circumcircle of 4PQC by k1 and the circumcircle of
4DRP by k2, where R is the point of intersection of AB and PQ.
Prove that:
a) MP is tangent to both k1 and k2.
b) RH ? CM , where H is the orthocentre of 4ABC .

Solution: a) Note that H 2 k1. Since 6 APM = 6 PAM =

90 ; 6 ABC = 6 BCD, we obtain that MP is a tangent to k1.
On the other hand 6 MPD = 6 MPB ; 6 DPB = 6 MPB ; 6 DPB =
6 MBP ;6 BPR (4ABC  4DBP ), so 6 ARP = 6 MBP ;6 BPR =
6 MBP ; 6 QPC = 6 MBP ; 6 BAC (4ABC  4PQC ). Therefore
6 MPD = 6 MBP and thus MP is a tangent to k2.

5
 b) Let L = CM \ k1. It follows from
a) that ML  MC = MP 2 = MD  MR.
We conclude that L lies on the circum-
cicrle of 4DRC and therefore RL ?
CM . Further HL ? CM , since HC is
a diameter of k1. Hence RH ? CM .

Problem 9.3. A square table lled with nonnegative (not neces-

sarily distinct) integer numbers is said to be a magic square with sum
m if the sum of the numbers in each row and each column equals m.
Prove that the number of magic squares 3  3 of sum m such that
the minimal element among ! the elements on the main diagonal lies
in the centre is 4 .m + 4

Solution: It is evident that knowing the elements of main diagonal

and the element in the cell (1 2) (see g. 1) one can determine all
elements in the table. Indeed, there is a unique choice for all remain-
ing cells (see g. 2). Therefore it suces to see when all elements
are nonnegative and b is the minimal element among the elements
on the main diagonal.
a d a d m;a;d
b m+c;a;b;d b a+d;c
c b+d;c m;b;d c
Fig. 1 Fig. 2
It is clear from g. 2 that the following inequalities hold:
(1) a + d  m
(2) b + d  m
6
(3) c  a + d
(4) c  b + d
(5) a + b + d ; c  m.
The conditions of the problem imply b  a and b  c. It is
clear now that (3) follows from (4) and (2) and (5) follow from (1).
Therefore we can consider only (1) and (4).
Consider the following chain of inequalities
b  2b + d ; c  a + b + d ; c  a + d  m
(the rst follows from (4), the second from b  a, the third from
b  c, and the fourth is equivalent to (1)). It is easy to see that
knowing the quadruple (b 2b+d;c a+b+d;c a+d) we can uniquely
determine a b c and d and so nd a magic square. Therefore the
required number equals the number of quadruples, which is m4+4 .

Problem 10.1. Find all values of the real positive parameter a

such that the inequality acos 2x + a2 sin x  2 holds for any real x.
2

Solution: We know that acos 2x + a2 sin x = a1;2 sin x + a2 sin x =

a + a2 sin
2 2

x . Substitute t = a2 sin x . Since 0  sin2 x  1,

2 2
2

a2 sin x
2

we obtain that t is between 1 and a2. Our inequality now becomes

a + t  2 () t2 ; 2t + a  0. Since it holds true for any x
t
(i. e., for any t between 1 and a2), it follows that the roots of f (t) =
t2 ; 2t + a = 0 lie outside the open interval determined by 1 and a2.
Therefore f (1)  0 and f (a2)  0. The rst inequality gives a  1
and the second one implies a4 ; 2a2 + a  0 () a3 ; 2a + 1 
0 () (a ; 1)(a2 + a ; 1)  0. Since a  1, we obtain a2 + a ; 1  0.

7
 " p p#
The solution of this inequality is a 2 ; 1 ; 5  ;1 + 5 . So we
" p# 2 2
obtain a 2 1 ;1 + 5 .
2

Problem 10.2. Let CH and CM be an altitude and a median in

a non-obtuse 4ABC . Let the bisector of angle BAC meet CH and
CM at points P and Q, respectively. If 6 ABP = 6 PBQ = 6 QBC ,
prove that:
a) 4ABC is a right triangle
b) BP = 2CH .
Solution: a) Let R = BC \ AP ,
T = AC \ BP and S = AC \
BQ. Denote AB = c, BC = a,
CA = b. It is easy to see that
P lies between A and Q (otherwise
6 ABP > 6 PBQ). It follows from
Ceva's Theorem for point P that:

(1) AH  BR  CT = 1 () b cos   c  SBTC = 1

HB RC TA a cos  b SABT
b cos  c BT  a  sin 2
cos  sin
() a cos   b  BT  c sin  = 1 () cos   sin 3 = 1
3
2

3 3
 beta
  2 sin 3 cos 3 = 1 () cos  = 1 :
() coscos  sin 3 cos  2 cos 3
8
 It follows from Ceva's Theorem for point Q that:

(2)
AM  BR  CS = 1 () c  SBSC = 1 () c  BS  a  sin 3 = 1
MB RC SA b SABS b BS  c sin 23
c a sin  a sin  a 
() b  c sin 2 = 1 () 2b sin  cos
3 3
 = 1 () b = 2 cos 3 :
3 3 3

Now (1) and (2) imply cos  = b . From the Sine Law we obtain
cos  a
sin  = b , so cos  = sin  () sin 2 = sin 2 . If  =  , then
sin  a cos  sin 
the triangle is isosceles and therefore P  Q, implying that 6 ABP =
6 PBQ = 6 QBC = 0 , which is impossible. Thus  +  = 90 and
therefore 6 ACB = 90 .
b) It follows from 4BCS that cos  = a . Combining the
3 BS
above with (2) gives 2b = BS . Note that 4ABC  4BHC , which
BP = BS = 2. Therefore BP = 2CH .
implies CH AC
Problem 10.3. Let A be a set of natural numbers with no zeroes in
their decimal representation. It is known that if a = a1a2 : : :ak 2 A,
then b = b1b2 : : :bk , where bj  1  j  k is the remainder of 3aj
modulo 10, belongs to A and the sum of the digits of b equals the
sum of the digits of a.
a) Prove that the sum of the digits of a k-digit number in A equals
5k.
9
 b) Find the smallest k-digit number which could be an element
of A.

Solution: a) Let a = a1a2 : : : ak be a k-digit number from A, the

sum of whose digits is S .
Consider the following numbers: b = b1b2 : : : bk  c = c1c2 : : : ck
and d = d1d2 : : :dk , where bj  1  j  k is the remainder of 3aj
modulo 10, cj  1  j  k is the remainder of 3bj modulo 10 and
dj  1  j  k is the remainder of 3cj modulo 10.
By the conditions of the problem all b, c and d belong to A.
Further
X
k X
k X
k X
k
(1) S= aj = bj = cj = dj :
j =1 j =1 j =1 j =1
Direct verication shows that for xed j the sum aj + bj + cj + dj
is equal to 20 (e. g., if aj = 3, then bj = 9 cj = 7 dj = 1 and
therefore
Pk aPj + bj + cPj + dj = P Pk now from (1) that 4S =
20). It follows
k k k
j =1 aj + j =1 bj + j =1 cj + j =1 dj = j =1 (aj + bj + cj + dj ) = 20k .
Therefore S = 5k, Q. E. D.
b) We shall prove that the required number is a = a1a2 : : :a2t,
where a1 = 1 a2 = 1 : : :  at = 1 at+1 = 9 at+2 = 9 : : : a2t = 9 if
k = 2t and b = b1b2 : : : b2t+1, where b1 = 1 b2 = 1 : : : bt = 1 bt+1 =
5 bt+2 = 9 : : : b2t+1 = 9 if k = 2t + 1. It is easy to see that a and b
could be elements of a set having the required property.
Let k = 2t and suppose there exists c = c1c2 : : :c2t 2 A such
that c < a. Since there are no zeroes among the digits of c, we
obtain c1 = c2 = : : : = ct = 1. But it follows from a) that the
sum of the digits of c is 5k = 10t. The last is possible only if
10
ct+1 = ct+2 = : : : = c2t = 9. Hence c = a, a contradiction with the
choice of c.
Similarly, suppose k = 2t + 1 and there exists c = c1c2 : : :c2t+1 2
A such that c < b. Since there are no zeroes among the digits of c
we obtain c1 = c2 = : : : = ct = 1. But it follows from a) that the
sum of the digits of c is 5k = 10t + 5. The latter is possible only if
ct+1  5 and since c < b, it follows that ct+1 = 5. It is easy to see
now that ct+1 = ct+2 = : : : = c2t = 9. Hence c = b, a contradiction
with the choice of c.
p
Problem 11.1. Given the sequence an = n+a n2 + 1 n = 1 2 : : :,
where a is a real number:
a) Find the values of a such that the sequence fang1n=1 is conver-
gent.
b) Find the values of a such that the sequence fang1n=1 is mono-
tone increasing.

Solution: a) Ifp a = ;1, the sequence fang1n=1 is convergent be-

cause an = n ; n2 + 1 = p;12 = ; n1 1 +1 1 ! 0 when
n+ n +1 n 2
n ! 1. Conversely,
p let the sequence
p f a n g 1 be convergent.
n=1 Since
an = pn ; n2 + 1 + (a + 1) n2 + 1,pwe get that the sequence
(a+1) n2 + 1 is also convergent. Since n2 + 1 ! 1 when n ! 1,
6 0, so a = ;1.
it follows that a + 1 =
b) Let fang1 n=1 be a monotone increasing sequence, i. e., an+1 

11
an for any n. This inequality is equivalent to
(?) a(2n + 1)  ;1:
q p
(n + 1)2 + 1 + n2 + 1
Since
2n + 1 2 + 1

nlim
!1
q p = lim q n q =1
(n + 1)2 + 1 + n2 + 1 n!1 (1 + n1 )2 + n12 + 1 + n12
it follows from (?) that a  ;1.
Conversely, let a  ;1. It follows from q 2n + 1
p <
(n + 1)2 + 1 + n2 + 1
2n + 1 = 1 that (?) holds true so the sequence fa g1 is increas-
n n=1
n+1+n
ing. The required values of a are a 2 ;1 +1).

Problem 11.2. Given a 4ABC with circumcentre O and circum-

radius R. The incircle of 4ABC is of radius r and touches the sides
AB BC and CA in the points C1 A1 and B1. Let the lines deter-
mined by the midpoints of the segments AB1 and AC1, BA1 and
BC1, CA1 and CB1 meet at points C2 A2 and B2. Prove r that the
circumcircle of 4A2B2C2 is of centre O and radius R + 2 .

Solution: We show rst that the projection B3 of B2 on AC is the

midpoint of AC . Let Ba and Bc be the midpoints of AB1 and CB1.
We shall use the standard notation for the elements of 4ABC .

12
 We obtain B a B3
B2B3 = tg  = r
2 p;a
and Bc B3 = tg 2 = p ;r c , so
B B

2 3
BaB3 = p ; c . Since B B = b ,
a c
BcB3 p ; a 2
it follows that BaB3 = 2 = p ; c
CBc and B B = p ; a = ABa ,
c 3
2 2 2
which gives AB3 = CB3. There-
fore B2O = B2B3 + B3O =
(p ; c)(p ; a) + R cos  .
2r
We shall show that the above expression equals R + 2r and by analogy
A2O = C2O = R + 2r , which will complete the proof.
We obtain that (p ; c2)(rp ; a) +R cos  = 2r +R () 2(pS; b) ;
S = R(1 ; cos  ) () Sb = 2R sin2  () r =
2p 2p(p ; b) 2 p;b
2 
4R sin2  () tg  = 2 sin 2 () sin  = 2 sin  cos  , which is
b 2 2 sin  2 2
a true equality.

Problem 11.3. Find the smallest natural number n such that the
sum of the squares of its divisors (including 1 and n) equals (n +3)2.

Solution: It is clear that n has at least three divisors and let 1 <
d1 < d2 <    < dk < n be those dierent from 1 and n. The
13
conditions of the problem imply
(?) d21 + d22 +    + d2k = 6n + 8:
Let n = p, where p is a prime number. It follows now from (?) that
p2 + p4 +    + p2;2 = 6p + 8, so pn8 and therefore p = 2. The
above equality implies 1 + p2 + p4 +    + p2;4 = 6p;2 + 2, which
is impossible.
Therefore k 6= 1 3 5, because otherwise the number of divisors
of n equals 3, 5, 7, i. e., n = p2 , n = p4 or n = p6, where p is a
prime number. Suppose that k  6. Since didk;i = n, it follows
from (?) that (dk;1 ; d1)2 + (dk;2 ; d2)2 + (dk;3 ; d3)2  8. The
last inequality is impossible, since the numbers dk;1 ; d1 dk;2 ; d2
and dk;3 ; d3 are distinct (if for example dk;1 ; d1 = dk;2 ; d2 = A,
then d1(A + d1) = d2(A + d2), so d1 = d2). We conclude now that
k = 2 or k = 4.
Assume k = 4. Then n has 6 divisors and thus n is of the form
n = p  q2, where p and q are distinct prime numbers. (n is not of
the form n = p5). It follows from (?) that
(??) p2 + q2 + q4 + p2 q2 = 6pq2 + 8:
If q  5, then q4 + p2q2  2pq3  10pq2 > 6pq + 8 and therefore
q = 2 or q = 3. Direct verication shows that inequality (??) is
impossible. Thus k = 2 and hence n = pq, where p and q are
distinct prime numbers such that
p2 + q2 = 6pq + 8:
Since qnp2 ; 8, it is easy to see that if p  17 then p = 7, q = 41
and n = 287. Since 172 = 289 > 287, we conclude that the smallest
n with the required property is n = 287.
14
 Annual Vojtĕch Jarník
International Mathematical
Competition

1991-2014
 The 1st Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 10th April 1991
Category I

Problem 1 Show that there is no arithmetic progression with three elements in the infinite geometric sequence
{2k }∞
k=0 .
Solution Suppose 2k , 2l , 2m form an arithmetic progression, where 0 < k < l < m. Then 2l − 2k = 2m − 2l
and therefore 2l+1 = 2k (2m−k + 1). Because of k < l < m, m ≥ k + 2, 2m−k + 1 is an odd number greater or
equal to 5. This odd number must divide 2l+1 . This is a contradiction. 
 The 1st Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 10th April 1991
Category I

Problem 2 Compute the determinant

 
0 a12 a13 ... a1n
 −a12 0 a23 ... a2n 
 
 −a13
det  −a23 0 ... a3n 
,
 ... 
−a1n −a2n −a3n ... 0

where n is an odd number.

Solution Let  
0 a12 a13 ... a1n
 −a12 0 a23 ... a2n 
 
 −a13
A= −a23 0 ... a3n 

 ... 
−a1n −a2n −a3n ... 0
and d = det A. If we multiply all rows in A by −1, we get new determinant d0 . The following holds:

d0 = (−1)n d = −d.

On the other hand,

d0 = det AT ,
where AT is the transposed matrix of A. Thus

−d = d0 = det AT = det A = d.

This implies that d = 0. 

 The 1st Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 10th April 1991
Category I

Problem 3 Let [x] be the integer part of x. Find the limit

√ √
lim ( 3 + 1)n − [( 3 + 1)n ] .


n→∞

Solution For n even we have

√ √ √ √ √ √
( 3 + 1)2n − [( 3 + 1)2n ] = ( 3 + 1)2n + ( 3 − 1)2n − ( 3 − 1)2n − [( 3 + 1)2n ]
√
= 1 − ( 3 − 1)2n ,

but for n odd we get

√ √ √ √
( 3 + 1)2n+1 − [( 3 + 1)2n+1 ] = ( 3 + 1)2n+1 − ( 3 − 1)2n+1
√ √
+ ( 3 − 1)2n+1 − [( 3 + 1)2n+1 ]
√
= ( 3 − 1)2n+1

So the limit does not exists.


 The 1st Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 10th April 1991
Category I

Problem 4 Let f (x) be an even, twice continuously differentiable function and f 00 (0) 6= 0. Prove that there is
an extremum of f (x) at the point x = 0.
Solution
f (h) − f (0) f (h) − f (0)
lim = lim− = f 0 (0) ,
h→0+ h h→0 h
and because f (h) = f (−h) we have

f (h) − f (0) f (h) − f (0)  f (h) − f (0) f (−h) − f (0) 

lim + lim = lim + = 0,
h→0+ h h→0− h h→0+ h −h
which implies that 2f 0 (0) = 0 . Hence f 0 (0) = 0 and because f 00 (0) 6= 0 we obtain that f (x) has an extremum
at the point x = 0. 
 The 2nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 28th April 1992
Category I

Problem 1 Find the nth derivation of the function

x
f (x) = .
x2 − 1
Solution
x (n) 1 1 1 (n)
f (x)(n) = = +
−1 x2 2 x+1 x−1
(−1)n  1 1 
= n! + .
2 (x + 1)n+1 (x − 1)n+1


 The 2nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 28th April 1992
Category I

Problem 2 Prove that there exist two real convex functions f, g such that

f (x) − g(x) = sin x

for all x ∈ R.
Solution Let f (x) = x2 + sin x and g(x) = x2 , then f (x) − g(x) = sin x and f (x), g(x) are convex function
because

f 00 (x) = 2 − sin x > 0 ,

g 00 (x) = 2 > 0 .


 The 2nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 28th April 1992
Category I

Problem 3 Prove that for all integers n > 1,

(n − 1)|(nn − n2 + n − 1) .

Solution For n = 2 we get (n − 1) = 1 and (nn − n2 + n − 1) = 1 so (n − 1)|(nn − n2 + n − 1).

Now we prove the result for n ≥ 3. First we use mathematical induction to prove that (n − 1)|(nk − n2 ) for
k ≥ 3. For k = 3 we have
(n3 − n2 ) : (n − 1) = n2 ⇒ (n − 1)|(n3 − n2 ) .
Suppose that (n − 1)|(nk − n2 ). We have to prove that (n − 1)|(nk+1 − n2 ). Hence (nk+1 − n2 )/(n − 1) = nk
with remainder (nk − n2 ), and because n − 1 divides (nk − n2 ) we have shown that (n − 1)|(nk − n2 ) for k ≥ 3.
Hence from the facts that (n − 1)|(nn − n2 ) and (n − 1)|(n − 1) we obtain that (n − 1)|(nn − n2 + n − 1) for all
n > 1. 
Second solution Let n > 1 be a fixed integer. Let f (x) be the polynomial xn − x2 + x − 1. Note that f (1) = 0.
It follows from the factor theorem that (x − 1)|f (x). Substituting in n for x and noting that n − 1 6= 0, we see
that
n − 1|f (n) = nn − n2 + n − 1 .

 The 2nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 28th April 1992
Category I

Problem 4 Let X be a finite set and f : X → X be map. Prove that f is an injective map if and only if f is
a surjective map.
Solution Let X have n elements.
1. Let f be injective, i.e. ∀xi , xj ∈ X; f (xi ) 6= f (xj ), i 6= j. Then we have ∀xi ∈ X, f (xi ) = yi ∈ X and
because f is injective then yi 6= yj whenever i 6= j. Because f : X → X we get f is surjective.

2. Suppose f is surjective, but it is not injective. Then ∃xi , xj such that f (xi ) = f (xj ). But because
f : X → X we obtain that f is not surjective. This is contradiction.

 The 2nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 28th April 1992
Category II

Problem 1 Prove that for a continuously differentiable function f (x), where f (a) = f (b) = 0,
Z b
0 1
max |f (x)| ≥ |f (x)| dx .
x∈[a,b] (b − a)2 a

Solution
b Z b
maxx∈[a,b] |f (x)|
Z
1 1
|f (x)| dx ≤ max |f (x)| 1=
(b − a)2 a
2
(b − a) x∈[a,b] a b−a
maxx∈[a,b] |f (x) − f (a)| maxx∈[a,b] |(x − a)f 0 (ξ)|
= =
b−a b−a
≤ max |f 0 (ξ)| .
ξ∈[a,b]


 The 2nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 28th April 1992
Category II

Problem 2 Find all functions f : R → R which satisfy the equality

xf (y) + yf (x) = (x + y)f (x)f (y) .

Solution For x = y = 1 we get f (1) = f 2 (1). Hence we have two possibilities. For f (1) = 0, y = 1 we obtain
that f (x) = 0 and for f (1) = 1, y = 1 we have
(
1 x 6= 0
f (x) = .
a∈R x=0


 The 2nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 28th April 1992
Category II

Problem 3 Let Zk be the additive group of residual classes modulo k. Decide if Z6 is isomorphic to Z2 × Z3 .
Solution Let Z6 be
⊕ 0 1 2 3 4 5
0 0 1 2 3 4 5
1 1 2 3 4 5 0
2 2 3 4 5 0 1
3 3 4 5 0 1 2
4 4 5 0 1 2 3
5 5 0 1 2 3 4
and Z2 × Z3

(0,0) (0,1) (0,2) (1,0) (1,1) (1,2)
(0,0) (0,0) (0,1) (0,2) (1,0) (1,1) (1,2)
(0,1) (0,1) (0,2) (0,0) (1,1) (1,2) (1,0)
(0,2) (0,2) (0,0) (0,1) (1,2) (1,0) (1,1)
(1,0) (1,0) (1,1) (1,2) (0,0) (0,1) (0,2)
(1,1) (1,1) (1,2) (1,0) (0,1) (0,2) (0,0)
(1,2) (1,2) (1,0) (1,1) (0,2) (0,0) (0,1)
Let f : Z6 → Z2 × Z3 be the function defined by

f (0) = (0, 0), f (1) = (1, 1), f (2) = (0, 2),

f (3) = (1, 0), f (4) = (0, 1), f (5) = (1, 2)

It is easy to check that this function is an injective function and the condition

f (x ⊕ y) = f (x)
f (y)

is fulfilled. Thus Z6 is isomorphic to Z2 × Z3 . 

 The 2nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 28th April 1992
Category II

p
Problem 4 Prove that each rational number q 6= 0 can be written in the form

p b2 bn
= b1 + + ··· + ,
q 2! n!

where n is a sufficiently large positive integer and bk ∈ Z (k > 1) such that 0 ≤ bk < k, bn 6= 0.
Solution Let
p p(q − 1)!
= .
q q!
Further we have

p(q − 1)! = sq q + bq ,
sq = sq−1 (q − 1) + bq−1 ,
..
.
s1 = b1 .

where bi ∈ {0, . . . , i − 1}. So we obtain

p b2 bq
= b1 + + ··· + .
q 2! q!

 The 3rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 13th -14th April 1993
Category I

Problem 1 Decide whether there is a nontrivial homomorphism from the additive group of rational numbers
to the additive group of integers.
Solution (by Stijn Cambie) Suppose it is possible to have a non-trivial homomorphism φ. In that case there
q q
is some rational q such that φ(q) = n for some non-zero integer n. By additivity we have φ(2n 2n ) = 2nφ( 2n )).
q 1 1
Hence φ( n ) = 2 and as 2 is not an integer, we have a contradiction. Hence such a homomorphism does not
exist. 
 The 3rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 13th -14th April 1993
Category I

Problem 2 Let A be a real magic matrix, i.e. there exists a nonzero real number S such that the sum of each
row is equal to S, the sum of each column is equal to S, the sum of the elements of the main diagonal is equal
to S and the sum of the elements of the secondary diagonal is equal to S.

1. Prove that if A is invertible then A−1 is magic.

2. Show that
S S S
 
3+u 3 −u+v 3 −v
A =  S3 − u − v S
3
S
3 + u + v ,
S S S
3 +v 3 +u−v 3 −u

where u and v are arbitrary numbers. Further show that A is not singular if and only if u2 6= v 2 .
Solution First we prove that
S S S
 
3 +u 3 −u+v 3 −v
A =  S3 − u − v S
3
S
3 + u + v .
S S S
3 +v 3 +u−v 3 −u

Let  
a11 a12 a13
A = a21 a22 a23 
a31 a32 a33
be a magic matrix. From this we get

a22 = S − a11 − a33 , (1)

a22 = S − a13 − a31 , (2)
a22 = S − a12 − a32 , (3)
a22 = S − a21 − a23 . (4)

Hence a22 = S3 . From (1) we obtain a11 = S3 + u and a33 = S3 − u. From (2) we get a31 = S3 + v and a13 = S3 − v.
From a11 + a12 + a13 = S we get a12 = S3 − u + v and a11 + a21 + a31 = S we obtain a21 = S3 − u − v. From (3)
and (4) we get a32 = S3 + u − v and a23 = S3 + u + v. We have matrix A in the form
S S S
 
3 +u 3 −u+v 3 −v
A =  S3 − u − v S
3
S
3 + u + v .
S S S
3 +v 3 +u−v 3 −u

The determinant of the matrix A is 3S(v 2 − u2 ), so if v 2 = u2 then the matrix A is singular; otherwise is regular.
Now suppose that A is regular. Then
 u S−v2 +u2 v S+v 2 −u2

− (3 v2 −3 u2 ) S − (3S−v−u
v+3 u) S (3 v 2 −3 u2 ) S
A−1 =  (3S+v−u 1
− (3S−v+u
 
v−3 u) S 3S v−3 u) S 
2 2 2 2
u S+v −u
− (3v vS−v +u
2 −3 u2 ) S
S+v+u
(3 v+3 u) S (3 v 2 −3 u2 ) S

and it is easy to check that this matrix is also magic. 

 The 3rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 13th -14th April 1993
Category I

Problem 3 Does there exist an injective function f : R → R satisfying the inequality

1
f (x2 ) − (f (x))2 ≥
4
for all x ∈ R?
Solution Let f : R → R be an injective function such that
1
f (x2 ) − (f (x))2 ≥
4
for all x ∈ R. For x = 0, we have
1
f (0) − (f (0))2 ≥
4
1
(f (0))2 − f (0) + ≤ 0
4
 1 2
f (0) − ≤0
2
and from this we get f (0) = 12 .
But on the other hand for x = 1 we get
1
f (1) − (f (1))2 ≥
4
1
(f (1))2 − f (1) + ≤ 0
4
 1 2
f (1) − ≤0
2
and from this we get f (1) = 12 . Hence f (0) = f (1) = 21 and because the function f is an injective function we
get a contradiction, so such a function does not exists. 
 The 3rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 13th -14th April 1993
Category I

Problem 4 Let a0 = 61992 , a1 = 3·61991 , . . . be a geometric progression and b0 = 465·31985 , b1 = 466·31985 , b2 =

467 · 31985 , . . . be an arithmetic progression. Find n such that an = bn .
Solution We have an = 3n · 61992−n and bn = (465 + n) · 31985 . We solve

3n · 61992−n = (465 + n) · 31985

31992 · 21992−n = (465 + n) · 31985
37 · 21992−n = 465 + n . (1)

For n > 1992 the term 37 · 21992−n is not an integer and equation (1) does not hold for any such n. Hence we
have two possibilities.
1. For n = 1992 we get 37 < 465 + 1992.
2. For n < 1992 we get 37 · 21992−n > 465 + n.
Hence we cannot find n ∈ N such that an = bn . 
 The 3rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 13th -14th April 1993
Category II

Problem 1 Decide if
1. Q[x]/(x2 − 1) ' Q[x]/(x2 − 4)
2. Q[x]/(x2 + 1) ' Q[x]/(x2 + 2x + 2),

where Q[x] is the ring of polynomials with rational coefficients and (f (x)) is the prime ideal in Q[x] generated
by f (x).
 The 3rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 13th -14th April 1993
Category II

Problem 2 Let n ≥ 1 be and mi be natural numbers such that mi < pn−i (0 ≤ i ≤ n − 1), where pk is
kth-prime. Prove that if m0 /pn + . . . + mn−1 /2 is a natural number then m0 = . . . = mn−1 = 0.
 The 3rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 13th -14th April 1993
Category II

Problem 3 Let P (4) (x) = x6 + x2 + 1. Prove that P (x) does not have ten distinct roots.
Solution The polynomial P (x) has at least ten roots then the polynomial P 0 (x) has at least nine roots, so
P (4) (x) has at least six roots. P (4) (x) = x6 + x2 + 1 and after substitution x2 = y we get the polynomial
H(y) = y 3 + y + 1. Because H 0 (y) = 3y 2 + 1 does not have real roots, we obtain that P (x) does not have ten
real roots. 
 The 3rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 13th -14th April 1993
Category II

Problem 4 Prove that if f : R → R fulfill the inequalities

f (x) ≤ x , f (x + y) ≤ f (x) + f (y)

for all x, y ∈ R, then f (x) = x for all x ∈ R.

Solution Let ∃ x ∈ R : f (x) < x. Hence

f (0) = f (x − x) ≤ f (x) + f (−x) < x + f (−x) .

And because f (−x) ≤ −x we have

f (0) < x − x ⇒ f (0) < 0 .
Hence
f (x) = f (0 + x) ≤ f (0) + f (x) < 0 + f (x) .
We obtained f (x) < f (x) which is contradiction, so f (x) = x for all x ∈ R. 
 The 4th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 1994
Category I

Problem 1 Prove that an arbitrary integer can be written as a sum of five cube powers of integers.
Solution For each n we have

6n = (n + 1)3 + (−n)3 + (−n)3 + (n − 1)3 .

Hence an arbitrary integer can be written in one of the following forms:

6n + 1 = 6n + 13
6n + 2 = 6(n − 1) + 23
6n + 3 = 6(n − 4) + 33
6n + 4 = 6(n + 2) + (−2)3
6n + 5 = 6(n + 1) + (−1)3
6n = (n + 1)3 + (−n)3 + (−n)3 + (n − 1)3 + 03 .


 The 4th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 1994
Category I

Problem 2 Prove that for the roots x1 , x2 of the polynomial

1
x2 − px − ,
2p2
where p ∈ R and p 6= 0, the following inequality holds:
√
x41 + x42 ≥ 2 + 2.

Solution According Vieta’s formula we have

x1 + x2 = p ,
1
x1 x2 = − .
2p2
Hence we use the relationship between the arithmetic mean and geometric mean and we get

x41 + x42 = (x1 + x2 )4 − 2x1 x2 (2(x1 + x2 )2 − x1 x2 )

1 1 1
= p4 + 2 (2p2 + 2 ) = 2 + p4 + 4
p 2p 2p
√
r
1
≥ 2 + p4 4 = 2 + 2 .
2p

Solution The roots of the polynomial
1
x2 − px −
2p2
are q q
2 2
p+ p2 + p2 p− p2 + p2
x1 = and x2 = .
2 2
Hence
1
x41 + x42 = 2 + p4 + .
2p4
1
Now we find the minimum of the function f (x) = x4 + 2x1 4 . The minimum occurs at the points x = ±2− 8 .
1 √
Hence f (±2− 8 ) = 2 and we obtain √
x41 + x42 ≥ 2 + 2 .

 The 4th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 1994
Category I

Problem 3 Prove that for all n ∈ N,

n 
Y 1
1+ < 3.
i=1
2i
Solution We prove that
n 
Y 1
1+ < 2.
i=2
2i
1
We know that 1 + x < 1−x for 0 < x < 1. Hence
 1  1  1 1
1+ 1+ ··· 1 + n < 1 1 1 ,
4 8 2 (1 − 4 )(1 − 8 ) · · · (1 − 2n )

and because (1 − x)(1 − y) > 1 − x − y for 0 < x, y < 1, we obtain

 1  1  1 1 1 1 1
1− 1− ··· 1 − n > 1 − − − ... − n ≥ .
4 8 2 4 8 2 2

 The 4th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 1994
Category I

Problem 4 Decide whether there exists a non-constant function f : R → R satisfying

2
f (x) − f (y) ≤ |x − y|3 (1)

for all x, y ∈ R.
Solution From (1) we get
f (x) − f (y) 2
( ) ≤ |x − y| .
x−y
Thus
f (x) − f (y)
lim =0
y→x x−y
and we have f 0 (x) = 0. Hence f (x) is constant. 
 The 4th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 1994
Category II

Problem 1 Find a triple of integers x, y, z, each greater then 50 and satisfying

x2 + y 2 + z 2 = 3xyz . (1)

Solution Let x ≤ y ≤ z. If (x, y, z) is a solution of the equation (1) then it is easy to check that (y, z, 3yz − x)
and (x, z, 3xz − y) solve the equation too. The triple (1, 1, 1) solve the same equation and hence is easy to find
the triple (x, y, z) greater then 50 which solve the equation (1). 
 The 4th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 1994
Category II

Problem 2 Prove that for an arbitrary n ∈ N, the number

 3 + √17 n  3 − √17 n
+
2 2
is an odd integer.
√
Solution The numbers λ1,2 = 3±2 17 are the solutions of the equation x2 −3x−2 = 0, which is the characteristic
equation of the recurrence yn+2 = 3yn+1 + 2yn . We have a0 = 2 and a1 = 3. Then for n ≥ 1

an+2 = 3an+1 + 2an ≡ 1an+1 + 0an = an+1 (mod 2) .

Hence for all n ≥ 1 the number

 3 + √17 n  3 − √17 n
an = +
2 2
is an odd integer. 
 The 4th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 1994
Category II

Problem 3 Let the function f : R → R satisfy

f (x) + f (y)
f (xy) = (1)
x+y

for all x, y ∈ R, x + y 6= 0. Is there x ∈ R such that f (x) 6= 0?

Solution For y = 1 we have
f (x) + f (1)
f (x) = (x 6= −1) (2)
x+1
and for y = 0 we have
f (x) + f (0)
f (0) = (x 6= 0) . (3)
x
From this equation we obtain f (x) = f (0)(x − 1) and for x = 1 we get f (1) = 0. From (2) we obtain xf (x) = 0
and we have f (x) = 0 for all x 6= 0, −1. Now if we put x = 2, y = 0 into (1) we get f (0) = 0 and for x = 0, y = −1
we obtain f (−1) = 0. 
 The 4th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 1994
Category II

Problem 4 How many real roots does the polynomial

x2 x3 xn
1+x+ + + ... +
2 3 n
have?
Solution Let
x2 x3 xn
f (x) = 1 + x + + + ... + .
2 3 n
We have two possibilities.
1. For n odd it is easy to check that f 0 (x) > 0 for all x ∈ (−∞, ∞). The function f (x) is continuous and
limx→−∞ f (x) = −∞ and limx→∞ f (x) = ∞, so we have one root.
2. For n even we obtain that f 0 (x) < 0 for x ∈ (−∞, −1), f 0 (x) = 0 for x = −1 and f 0 (x) > 0 for x ∈ (−1, ∞).
The function f (x) has a minimum at the point x = −1, but f (−1) > 0 so f (x) has no roots.

Hence the function f (x) has at most one real root. 

 The 5th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th – 26th April 1995
Category I

Problem 1 Discuss the solvability of the equations

λx + y + z = a
x + λy + z = b
x + y + λz = c

for all numbers λ, a, b, c ∈ R.

Solution The characteristic polynomial of this system is (λ − 1)(λ − 1)(λ + 2). Hence we have three possibilities
of solution:
1. For λ = 1 we obtain that if a = b = c then the system has infinity many solutions in the form x =
a − u − v, y = u, z = v, where u, v are parameters. Otherwise the system has no solution.
2. For λ = −2 we get if a + b + c = 0 the system has infinitely many solutions. Otherwise the system has no
solution.
3. For λ ∈ R \ {1, −2} the system has just one solution.

 The 5th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th – 26th April 1995
Category I

Problem 2 Let f (x) be an even twice differentiable function such that f 00 (0) 6= 0. Prove that f (x) has a local
extremum at x = 0.
Solution
f (h) − f (0) f (h) − f (0)
lim = lim− = f 0 (0)
h→0+ h h→0 h
and because f (h) = f (−h) we have

f (h) − f (0) f (h) − f (0)  f (h) − f (0) f (−h) − f (0) 

lim + lim = lim + =0
h→0+ h h→0− h h→0+ h −h
2f 0 (0) = 0 .

Hence f 0 (0) = 0 and because f 00 (0) 6= 0, we obtain that f (x) has an extremum at the point x = 0. 
 The 5th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th – 26th April 1995
Category I

Problem 3 Let f (x) and g(x) be mutually inverse decreasing functions on the interval (0, ∞). Can it hold
that f (x) > g(x) for all x ∈ (0, ∞)?
Solution The answer is yes. For example, the functions f (x) = − 21 x − 1 and g(x) = −2x − 2 are mutually
inverse functions and f (x) > g(x) for all x ∈ (0, ∞). 
 The 5th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th – 26th April 1995
Category II

Problem 1 Prove that the systems of hyperbolas

x2 − y 2 = a (1)
xy = b (2)

are orthogonal.
√ √
Solution An arbitrary point lying on hyperbola (1) has the form H1 = [x, x2 − a] and H1̄ = [x, − x2 − a]
and points lying on hyperbola (2) have the form H2 = [x, xb ]. The tangent vectors to hyperbola (1) are
H10 = (1, √xx2 −a ) and H1̄0 = (1, √x−x
2 −a
) and the tangent vector to hyperbola (2) is H20 = (1, −b
x2 ].
The point H is the point of intersection if and only if
p b b p
H = [x, x2 − a] = [x, ] ⇒ = x2 − a (3)
x x
or p b b p
H = [x, − x2 − a] = [x, ] ⇒ = − x2 − a . (4)
x x
Hyperbolas are orthogonal at the intersection point if and only if the scalar product of the tangent vectors at
this point is equal to 0. From (3) and (4) we have

b
H10 · H2 = 1 − √ =0
x x2 − a
and
b
H1̄0 · H2 = 1 − √ = 0.
x x2 − a
So the systems of hyperbolas are orthogonal. 
 The 5th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th – 26th April 1995
Category II

Problem 2 Let f = f0 + f1 z + f2 z 2 + . . . + f2n z 2n and fk = f2n−k for each k. Prove that f (z) = z n g(z + z −1 ),
where g is a polynomial of degree n.
Solution Let
 
n
f2n = f0 = an
0
 
n−1
f2n−1 = f1 = an−1
0
   
n n−2
f2n−2 = f2 = an + an−2
1 0
   
n−1 n−3
f2n−3 = f3 = an−1 + an−3
1 0
     
n n−2 n−4
f2n−4 = f4 = an + an−2 + an
2 1 0
..
.
P
 k/2 n−2i


i=0 an−2i k2 −i
for k even
f2n−k = fk = P k−1 n−2i−1

i=0 an−2i−1 k−1 −i for k odd
 2
2

Then we get f (z) in the form

n  k   
X k
f (z) = ak (z n+k + z n−k ) + (z n+k−2 + z n−k+2 ) + . . .
0 1
k=0
k
(
(z + z −1 ) for k even


k

+ k

2
−1
k−1 (z + z ) for k odd
2
n
X  k    
n k −k k 
=z ak (z + z ) + (z k−2 + z 2−k + . . .)
0 1
k=0
n 
 k     
n
X
k k k−2 k −k 
=z ak z + z + ... + z )
0 1 k
k=0
Xn
= zn ak (z + z −1 )k
k=0
= z g(z + z −1 ) ,
n

where g denotes the polynomial of degree n. 

 The 5th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th – 26th April 1995
Category II

Problem 3 Let f : R → R be a continuous function. Do there exist continuous functions g : R → R and

h : R → R such that f (x) = g(x) sin x + h(x) cos x holds for every x ∈ R?
Solution Let g(x) = f (x) sin x and h(x) = f (x) cos x. Both functions g(x) and h(x) are continuous function
and the condition
f (x) = g(x) sin x + h(x) cos x
holds. 
 The 5th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th – 26th April 1995
Category II

Problem 4 Let {xn }∞

n=1 be a sequence such that x1 = 25, xn = arctan xn−1 . Prove that this sequence has a
limit and find it.
Solution The sequence {xn }∞ n=1 is a decreasing sequence because 0 <
arctan x
x < 1 and {xn }∞
n=1 is bounded
by 0. So there exists a limit of this sequence. Let limn→∞ xn = L. Then L = arctan L, hence L = 0 and
limn→∞ xn = 0. 
 The 6th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 3rd April 1996
Category I

2 2
Problem 1 On the ellipse xa2 + zb2 = 1 find the point T = [x0 , z0 ] such that the triangle bounded by the axes
of the ellipse and the tangent at that point has the least area.
Solution The equation of the tangent at point T is
xx0 zz0
+ 2 =1
a2 b
2 2
and its intersection with the axes occurs at the points P = [ xa0 , 0] and Q = [0, zb0 ]. The area S of triangle OP Q
is
b2 a 2 b3 a
S= = p .
2z0 x0 2z0 b2 − z02
We will find the extreme point of the function S(z0 ). Note that

b3 a  1 1 
S 0 (z0 ) = 3 − p .
2 (b2 − z02 ) 2 z02 b2 − z02

Hence z0 = √b and x0 = √a . At the points z0 = b and z0 = 0, where the derivative does not exist, the function
2 2
S(z0 ) does not have an extremum, because we do not have a triangle. And because S 00 √b2 > 0 we have a

minimum at that point. If we use the same process to find the additional points, we get the other three points
[− √a2 , √b2 ], [ √a2 , − √b2 ] and [− √a2 , − √b2 ]. 
 The 6th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 3rd April 1996
Category I

Problem 2 Let {an }∞

n=0 be the sequence of integers such that a0 = 1, a1 = 1, an+2 = 2an+1 − 2an . Decide
weather
[n/2] 
X n
an = .
2k
k=0

Solution No. If
[n/2]  
X n
an = ,
2k
k=0
n−1
then we get the sequence an = 2 for n ≥ 1. But if an+2 = 2an+1 − 2an holds, then we have the another
sequence an = 12 (1 + i)n + 12 (1 − i)n for n ≥ 0. 
 The 6th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 3rd April 1996
Category I

Problem 3 Prove that the equation

z y x 1
+ + =
1 + z2 1 + y2 1 + x2 1996
has finitely many solutions in positive integers.
Solution Let x ≤ y ≤ z. Then
z y x
≤ ≤ . (1)
1 + z2 1 + y2 1 + x2
Hence
3x 1
≥
1 + x2 1996
and from this x2 − 5988x + 1 ≤ 0 so we have finitely many x ∈ N. Further we can write
z y 1
2
+ 2
= ,
1+z 1+y Cx
1 x
where Cx = 1996 − 1+x2 is bounded number (for particular x). From (1) we get

2y 1
2
≥
1+y Cx

and from this y 2 − 2Cx y + 1 ≤ 0 and y is also a bounded number. Then

z 1 x y
= − − ,
1 + z2 1996 1 + x2 1 + y2
1 x y
where the term on the right side has finitely many values. For each particular value of 1996 − 1+x2 − 1+y 2 we
obtain two distinct (or the same) numbers z. So we have only finitely many numbers z.
Hence the equation
z y x 1
2
+ 2
+ 2
=
1+z 1+y 1+x 1996
has finitely many solutions. 
 The 6th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 3rd April 1996
Category II

Problem 1 Is it possible to cover the plane with the interiors of a finite number of parabolas?
Solution Suppose that there exists a finite system S of parabolas, which cover the plane. The number of
parabolas is n. Take two parabolas from S which intersect. These parabolas have at most four intersection
points. We choose another parabola from S which covers at least one of the intersection points. Hence we have
either a new intersection point or part of a parabola which is not covered. We do this iteration process for all
parabolas from S. At the end of this iteration process we get one intersection point or part of a parabola which
is not covered. So we have a contradiction. 
 The 6th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 3rd April 1996
Category II

Problem 2 Let {xn }∞ n=0 be the sequence such that x0 = 2, x1 = 1 and xn+2 is the remainder of the number
xn+1 + xn divided by 7. Prove that xn is the remainder of the number
[n/2]
X n
4n 2 5k
2k
k=0

divided by 7.
Solution

 The 6th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 3rd April 1996
Category II

1997
Problem 3 Let cif(x) denote the sum of the digits of the number x in the decimal system. Put a1 = 19971996 ,
an+1 = cif(an ) for every n > 0. Find limn→∞ an .
Solution For the function cif(x) we have cif(x) ≡ x (mod 9). For x ≥ 10 we obtain cif(x) < x and for x ≤ 9
we obtain cif(x) = x. Hence there exists N such that 1 ≤ an+1 < an for all n ≤ N − 1 and 1 ≤ an+1 = an for
all n ≥ N . This implies that
lim an = aN ≡ a1 (mod 9) .
n→∞

We have 1997 1997

19971996 ≡ (−1)1996 =1 (mod 9) .
From 1 ≤ limn→∞ an ≤ 9 we obtain limn→∞ an = 1. 
 The 7th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 9th April 1997
Category I

Problem 1 Let a be an odd positive integer. Prove that if d | (a2 + 2) then d ≡ 1 (mod 8) or d ≡ 3 (mod 8).
Solution If p | (a2 +
2) and p is the
prime, then −2 is a quadratic residue modulo p. It follows that for the
Legendre symbol −2 p = 1, where ·· is the Legendre symbol. Using the properties of Legendre symbol we
obtain  −2   −1   2  p−1 p2 −1
= · = (−1) 2 + 8 .
p p p
p−1 p2 −1
Thus the number 2 + 8 is even and so

p2 + 4p ≡ 5 (mod 16). (1)

On the other hand p is an odd prime, thus p is of the form 8k + 1, 8k + 3, 8k + 5 or 8k + 7. This and (1) yield
that the prime p is of the form 8k + 1 or 8k + 3. The product of an even number of numbers of the form 8k + 3
is a number b fulfilling b ≡ 1 (mod 8), and the product of an odd number of numbers of the form 8k + 3 is c
fulfilling c ≡ 3 (mod 8). The product e of numbers of the form 8k + 1 satisfies e ≡ 1 (mod 8). This yields the
assertion. 
 The 7th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 9th April 1997
Category I

Problem 2 Let α ∈ (0, 1] be a given real number and let the real sequence {an }∞
n=1 satisfy the inequality

an+1 ≤ αan + (1 − α)an−1 for n = 2, 3, . . .

If {an } is bounded prove that then it must be convergent.

Solution Since (an ) is bounded there exists both lim inf an = l and lim sup an = L. We shall prove that L = l.
Let suppose the contrary L > l. By the definition of lim sup and lim inf we know that there exist subsequencies
(ank ) and (amk ) of (an ) which converge to l and L.
Since L = lim sup an we have:

(∀ε > 0)(∃n0 ∈ N)(∀n ≥ n0 )an < L + ε.

On the other hand

(∀ε > 0)(∃n1 ∈ N)(∀n ≥ n1 )l − ε < an .
We choose mk0 such that: mk0 > n1 , l − ε < amk0 < l + ε. By the inequality we obtain:

amk0 +1 ≤ (1 − α)amk0 + αamk0 −1 < (1 − α)(l + ε) + α(L + ε) = ε + (1 − α)l + αL

and

amk0 +2 ≤ (1 − α)amk0 +1 + αamk0 < (1 − α)(ε + (1 − α)l + αL) + α(l + ε) == ε + ((1 − α)2 + α)l + (1 − α)αL.

It is not difficult to demonstrate that we can choose ε such that:

ε
ε + (1 − α)l + αL < L − ,
2
ε
ε + (1 − α + α2 )l + (α − α2 )L < L − ,
2
using the functions: f (x) = ε + (1 − x)l + xL and g(x) = ε + (1 − x − x2 )l + (x − x2 )L, x ∈ [0, 1].
From that we have: amk0 +1 ≤ L − 2ε , amk0 +2 ≤ L − 2ε . From that we obtain:
 ε  ε ε
amk0 +3 ≤ (1 − α)amk0 +1 + amk0 < (1 − α) L − +α L− =L− ,
2 2 2
therefore by induction we get:
ε
an < L − , ∀n ≥ mk0 +1 .
2
The last formula yields:
ε
lim sup an ≤ L − < L,
2
which is a contradiction. Thus L = l. 
 The 7th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 9th April 1997
Category I

Problem 3 Let c1 , c2 , . . . , cn be real numbers such that

ck1 + ck2 + · · · + ckn > 0 for k = 1, 2, . . . (1)

Let us put
1
f (x) = .
(1 − c1 x)(1 − c2 x) . . . (1 − cn x)
Show that f (k) (0) > 0 for all k = 1, 2, . . . .
Solution Put g(x) := log f (x). Then
n n
X X (k − 1)!ckj
g(x) = − log(1 − cj x) and g (k) (x) = .
j=1 j=1
(1 − cj x)k

Assumption (1) and the above result give

g (k) (0) = ck1 + ck2 + · · · + ckn > 0.

Now observe that if all the derivatives of g at the origin are positive, then eg has the same property. For this
show by induction that
(eg )(k) = eg · S ,
where S is a finite sum of terms of the form

a(g (l1 ) )m1 (g (l2 ) )m2 . . . (g (lr ) )mr ,

where a, lj and mj are positive integers. For instance:

(eg(x) )0 = eg(x) g 0 (x)

(eg(x) )00 = eg(x) ((g 0 (x))2 + g 00 (x))
(eg(x) )000 = eg(x) ((g 0 (x))3 + 3g 0 (x)g 00 (x) + g 000 (x))
(eg(x) )(4) = eg(x) ((g 0 (x))4 + 6(g 0 (x))2 g 00 (x) + 3(g 00 (x))2 + 4g 0 (x)g 000 (x) + g (4) (x))


 The 7th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 9th April 1997
Category I

Problem 4-M Find all real numbers a > 0 for which the series
∞
X af (n)
n=1
n2

is convergent, where f (n) denotes the number of zeros in the decimal expansion of n.
Solution For n = 0, 1, . . . let us consider k’s fulfilling the inequality:

10n ≤ k < 10n+1 . (1)

For a fixed 0 ≤ j ≤ n there are

 
n n−j+1
9 k’s fulfilling (1) such that f (k) = j.
j

Consequently,

n  
! 10n+1 −1 n  
!
9(9 + a)n 9 X n j n−j X af (k) 9 X n j n−j 9(9 + a)n
2n+2
= 2n+2 a 9 ≤ 2
≤ 2n a 9 = .
10 10 j=0
j n
k 10 j=0
j 102n
k=10

This implies that

∞ ∞ ∞ 
9 X  9 + a n X af (n) X 9 + a n
≤ 2
≤ 9 .
100 n=0 100 n=1
n n=0
100
From the last inequalities we conclude that our series is convergent exactly when
9+a
<1
100
or
0 < a < 91.

 The 7th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 9th April 1997
Category I

Problem 4-I Let us have declared:

const N_MAX = 255;
type
tR = array[1..N_MAX] of real;
tN = array[1..N_MAX] of integer;
and function random without parameters which returns real random values distributed uniformly in [0, 1).
You need to choose K integer numbers (1 < K < N, N M AX ≥ N ) without repetitions under the condition
N
P
that the probability of the choice of a number i equals a given Pi , Pi = 1.
i=1
Write the procedure in Pascal that returns such K integer numbers in the first K elements of the vector of
tN type. Input parameters of the procedure are K, N and the vector of Pi .
Solution

 The 7th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 9th April 1997
Category II

Problem 1 Decide whether it is possible to cover a 3-dimensional Euclidean space with lines which are pairwise
skew (i.e. not coplanar).
Solution There is (for example) the following covering:

lines pa,b = {Aa,b = [a, b, 0]; sa,b = (−b, a, 1)},

where Aa,b is point and sa,b is vector of pa,b . We show that

1. if [a, b] 6= [c, d] then pa,b ∩ pc,d = ∅; and
2. for each X = [x, y, z] there is [a, b] such that X ∈ pa,b .

1. Let [a, b] 6= [c, d] and pa,b ∩ pc,d = ∅. Then (from the parametric expression)

a − tb = c − rd
b + ta = d + rc
t=r

for some real t and r. From the third equality t = r and from the first and the second we have a−c = t(b−d)
and b − d = −t(a − c). By linear combination (first times (a − c) plus second times (b − d)) we have

(a − c)2 + (b − d)2 = 0.

But this contradicts [a, b] 6= [c, d].

x+yz y−xz
2. For X = [x, y, z] put a = 1+z 2 ; b= 1+z 2 . Then X = Aa,b + z · sa,b .

This implies that it is possible to cover 3-dimensional Euclidean space with lines which are pairwise skew. 
 The 7th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 9th April 1997
Category II

Problem 2 Let f : C → C be a holomorphic function such that |f (z)| = 1 for all z ∈ C with |z| = 1. Prove
that there are θ ∈ R, k ∈ {0, 1, 2, . . . } such that

f (z) = eiθ z k for any z ∈ C.

Solution We know that

|f (z)| = 1 for all |z| = 1. (1)
Let ∆ := {z ∈ C : |z| < 1}. Let {z1 , . . . , zn } be zeros (counted with multiplicity) of the function f belonging
to ∆. The number of zeros is finite because of the identity principle and (1).
Let us define the following function:

f (z)
g(z) := Qn z−zj , for all z with z z̄j 6= 1.
j=1 1−z̄j z

¯ (even in C \
From the choice of zj we certainly have that g is holomorphic in some neighbourhood of ∆
{1/z̄1 , . . . , 1/z̄n }).
Because of (1) and the fact that  z−z 
j 
 = 1 for all |z| = 1

1 − z̄j z


we have that
|g(z)| = 1 for all |z| = 1. (2)
Moreover,
g(z) 6= 0 for all z ∈ ∆. (3)
In view of (2) and (3), the minimum and maximum principles applied to the mapping g imply that

|g| ≡ 1 on ∆,

which gives us that

g ≡ eiθ for some θ ∈ R.
Therefore,
n
Y z − zj
f (z) = eiθ , z ∈ C.
j=1
1 − z̄j z

And the last inequality is possible iff z1 = · · · = zn = 0. 

 The 7th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 9th April 1997
Category II

Problem 3 Let u ∈ C 2 (D̄), u = 0 on ∂D, where D is an open unit ball in R3 . Prove that the following
inequality Z Z Z
1
|grad u|2 dV ≤ ε (∆u)2 dV + u2 dV
D D 4ε D
holds for all ε > 0.
Solution Since the inequality must hold for every ε > 0 it will hold q too for those ε for which the function
g(ε) = ε D (∆u) dV + 4ε D u dV takes a minimum, i.e., for ε = 2 IIul (g 0 (ε) = Il − 4ε12 Iu = 0). That
2 1 1
R R 2
q √
minimum is g( 21 IIul ) = Iu Il . Therefore it suffices to show that

Z sZ Z
2
|grad u| dV ≤ (∆u)2 dV u2 dV .
D D D

By placing (P, Q, R) = u(ux , uy , uz ) in the Gauss-Ostrogradski formula we get:

Z Z Z
2 ∂u
|grad u| dV + u∆u dV = − u dS = 0,
D D ∂D ∂n

where n is inner normal. From this by the Cauchy-Schwartz inequality and the conditions of that statement we
obtain sZ
Z Z Z
2
|grad u| dV = − u∆u dV ≤ (∆u)2 dV u2 dV .
D D D D


 The 7th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 9th April 1997
Category II

Problem 4-M Prove that

∞ 2 6
n2 1 X X cos   2πj   2πjk 
X 2πj


= 3 e 7 · cos sin + .
n=1
(7n)! 7 j=0
7 7
k=1

Solution
∞ ∞ 2 ∞
X n2 1 X 7n(7n − 1) + 7n X X 1
= =
n=1
(7n)! 49 n=1 (7n)! n=1
(7n − k)!
k=1

We have
2 X
6 2 X
6 2 X
6 ∞ 2πijk 2 X∞ 6
X 2πij
+ 2πijk
X 2πijk
2πij X 2πijk X e 7 X 1 X 2πij(n+k)
A= ee 7
7 = e 7 ee 7
= e 7 = e 7 .
n=0
n! n=0
n! j=0
k=1 j=0 k=1 j=0 k=1 j=0 k=1

If 7 | N , then
6
X 2πijN
e 7 = 7.
j=0

If 7 - N , then
6
X 2πijN e2πiN − 1
e 7 = 2πijN = 0.
j=0 e 7 −1
It follows that
∞
2 X
X 1
A=7 .
n=1
(7n − k)!
k=1

Thus
∞ 2 6 2 6
X n2 1 X X e 2πij
7 + 2πijk 1 X X cos( 2πj )+i(sin( 2πj )+( 2πjk ))
= 3 e 7 = 3 e 7 7 7 =
n=1
(7n)! 7 j=0
7 j=0
k=1 k=1
2 6
1 X X cos( 2πj )   2πj   2πjk 
= 3 e 7 · cos sin + .
7 j=0
7 7
k=1


 The 7th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 9th April 1997
Category II

Problem 4-I Problem Div3 is specified as follows:

Input: any program P ,
Output: a finite set D(P ) of strings of 0’s and 1’s,

where it holds that the program P solves problem Div3 iff it outputs the correct answers for inputs from D(P ).
Theorem For any program T RAN SF which transforms programs in some way (i.e. for any given program P
it constructs some program P 0 , denoted by P 0 = T RAN SF (P )) there is a program P0 whose input/output
behaviour is not changed by the transformation (i.e. P0 and T RAN SF (P0 ) yield the same outputs for the same
inputs).
Solution Suppose there is some GEN-TEST-DATA with the property described above. Now construct a
program T RAN SF which works as follows:
When given a program P , it constructs D(P ) by using GEN-TEST-DATA and then constructs P 0 whose
behaviour can be described as follows:

s:= input;
if s ∈ D(P );
then output YES or NO according to whether or not s is
the binary code of a number divisible by 3;
else output YES;

Due to the Recursion Theorem there is a program P0 s.t. its input/output behaviour is the same as the
behaviour of T RAN SF (P0 ), and it can be described as follows:

s:= input;
if s ∈ D(P0 );
then output YES or NO according to whether or not s is
the binary code of a~number divisible by 3;
else output YES;

Such a program P0 obviously violates the condition supposed for GEN-TEST-DATA.

Hence we have to conclude that there is no desired program GEN-TEST-DATA. 
 The 8th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 1st April 1998
Category I

Problem 1 Let a and d be two positive integers. Prove that there exists a constant K such that every set of
K consecutive elements of the arithmetic progression {a + nd}∞
n=1 contains at least one number which is not
prime.
Solution Let p be a prime number such that p - d. Let for some n all elements an , . . . , an+p are prime numbers.
Clearly for i = 1, . . . , p we have an+i = an +id. Thus an > p, because an+an = an +an d it is not a prime number.
Consider now the system of the reminders an mod p, . . . , an+p (mod p). Because all are primes greater than
p, this system does not contain 0. Thus there exist two numbers r1 < r2 such that ar1 ≡ ar2 (mod p). This
yields an + r2 d ≡ an + r1 d (mod p), and so r2 d ≡ r1 d (mod p). But (p, d) = 1 therefore r2 ≡ r1 (mod p) which
is a contradiction. 
 The 8th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 1st April 1998
Category I

Problem 2 Find the limit
n n

1 + n1

lim .
n→∞ e
Solution Recall that  1 n
lim 1 ± = e±1 ,
n→∞ n
and put
1
f (x) = (1 + x) x forx > 0 ,
f (0) = lim+ f (x) = e .
x→0

0
It is easy to verify that f+ (0) = − 2e . Hence

(1 + n1 )n 1 1
f (x) = e(1 − 2x) + o(x) and =1− +o
e 2n n
and
n n
1 + n1
  1 n
 1 1
lim = lim 1 − +o =√
n→∞ e n→∞ 2n n e

 The 8th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 1st April 1998
Category I

Problem 3 Give an example of a sequence of continuous functions on R converging pointwise to 0 which is

not uniformly convergent on any nonempty open set.
for n ∈ N a continuous function gn by gn (t) = 0 for t ≤ 0 or t ≥ n2 , gn n1 = 1 and gn is linear


Solution
 1  Define
on 0, n , n1 , 2n
1

. Put
∞
X
fn (t) = 2−k gn (t − rk ) , (1)
k=0

where rk is an enumeration of rationals. By the Weierstrass theorem, the fn are continuous. To prove that fn
∞
2−k < 2ε . By (1) and the definition of
P
tends pointwise to 0, fix t ∈ R and ε > 0. Choose m ∈ N such that
k=m
gn , we can choose n ∈ N such that for l ≥ n
m−1
X ε
2−k gl (t − rk ) < ,
2
k=0

which proves our claim. Now suppose that there exists an open, nonempty interval I such that fn → 0 uniformly
on I. Choose k0 ∈ N with rk0 ∈ I. Put tn = rk0 + n1 . Then for n sufficiently large,

1
sup|fn (t)| ≥ fn (tn ) ≥ 2−k0 gn (tn − rk0 ) = 2−k0 gn = 2−k0 ;
t∈I n

a contradiction. 
 The 8th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 1st April 1998
Category I

Problem 4-M Prove the inequality

n−1
r
nπ 1 1 X k2 nπ
−√ ≤ + 1− 2
≤ (1)
4 8n 2 n 4
k=1

for every integer n ≥ 2.

Solution We obtain the inequalities by approximating the area of the first quarter of the unit circle with the
sum of the areas of inscribed and tangent trapezoids (see the figure).

k−1 k
n n

Obviously the sum of the area of the inscribed trapezoids with vertices
r r
k − 1  k  k k2   k − 1 (k − 1)2 
,0 , ,0 , , 1− 2 , , 1−
n n n n n n2
π
(from k = 1 to k = n) estimates 4 (the area of the quarter of the unit circle) from below. The areas of such a
trapezoid is q q
(k−1)2 k2
1 1− n2 + 1− n2
,
n 2
thus after summation we obtain the inequality
n−1
r
1 1X k2 π
+ 1− ≤ , (2)
2n n n2 4
k=1

which immediatelly implies the second inequality in (1).

The set of the points of the unit circle with positive ordinate and with abscissa between 2k−1 2k+1
2n , and 2n is a
2k−1 2k+1
subset of the trapezoid determined by the horizontal axis, the vertical lines with abscissas 2n and 2n , and
 q  q
2 2
the tangent line of the unit circle at the point nk , 1 − nk 2 . The area of this trapezoid is simply n1 1 − nk 2 .
1
We still have to cover the points of the first quarter of the unit circle with abscissas less than 2n or greater than
2n−1 1
2n . The former part is simply covered by a rectangle of area 2n , while the latter part is again covered by a
tangent trapezoid of area Tn , which satisfies
r  4n − 1 2 r
1 1 8n − 1 1 1
Tn = 1− = ≤ √ .
2n 4n 2n 16n2 2n 2n

The sum of these areas estimates the area of the quarter of the unit circle from above, therefore we have
n−1
r
π 1 1X k2 1 1
≤ + 1− 2 + √ . (3)
4 2n n n 2n 2n
k=1

Obviously inequality (3) is equivalent to the first inequality in (1). 

 The 8th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 1st April 1998
Category I

Problem 4-I Prove that there exists a program in standard Pascal which prints out its own ascii code. No
disc operations are permitted.
Solution Let g be a computable Gödel numbering over pascal programs. One can assume that every number
is used in this numbering, i.e. every number is g(K) for some program K. Let Kn be the program whose Gödel
number is n and let k(σ) be the following program, σ ∈ N:
Program K ;
const N = ∆σ ;
procedure PrintOut ( P : integer ) ;
var ... end ; { PrintOut }
begin
PrintOut ( N ) ;
end .
where ∆σ is the decimal representation of σ, and PrintOut is the procedure which given a number n prints the
program whose Gödel number is n. We will show that there exists σ satisfying

g(k(σ)). (1)

Indeed, let
Df
Π = hλs : g(k(s))i
and let
Df
δ = hλs : the output of Ks running with s as the inputi.
Both the functions Π and δ are computable, hence the function Π ◦ δ is computable. Let L be a program
Df
computing the latter and let M be the Gödel number of L. Then σ = δ(M ) satisfies (1). Indeed, we have:

σ = δ(M ) = hthe value of the program KM on data M i =

= hthe value of the program L on data M i = (Π ◦ δ)(M ) =
= Π(δ(M )) = Π(σ) = g(k(σ)).

So the desired program is k(σ). To see that, let us notice that k(σ) prints Kσ , i.e. it prints Kg(k(σ)) = k(σ).
Hence it prints itself. 
 The 8th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 1st April 1998
Category II

Problem 1 Let H be a complex Hilbert space. Let T : H → H be a bounded linear operator such that
|(T x, x)| ≤ kxk2 for each x ∈ H. Assume that µ ∈ C, |µ| = 1, is an eigenvalue with the corresponding eigenspace
E = {φ ∈ H : T φ = µφ}. Prove that the orthogonal complement E ⊥ = {x ∈ H : ∀φ ∈ E : (x, φ) = 0} of E is
T -invariant, i.e., T (E ⊥ ) ⊆ E ⊥ .
Solution Suppose x ∈ E ⊥ , kxk = 1. This means that (x, φ) = 0 for each φ ∈ E. Then for any t ∈ C and
φ ∈ E, kφk = 1, we have

1 + |t|2 = kφ + txk2 ≥ |T (φ + tx), φ + tx|

= |(T φ, φ) + t(T x, φ) + t̄(T φ, x) + |t|2 (T x, x)|
= |µ + t(T x, φ) + |t|2 (T x, x)|

because (T φ, x) = µ(φ, x) = 0. If (T x, x) 6= 0 we can take

|(T φ, x)|2 (T x, x)
t= 6= 0
|(T x, x)|2 (T x, φ)

to obtain 1 + |t|2 ≤ |µ + 0| ≤ 1, a contradiction. In the case (T x, x) = 0 one can take t = −µ/(T x, φ) yielding
a contradiction again. This implies that (T x, φ) = 0 for any φ ∈ E, i.e. T x ∈ E ⊥ . 
 The 8th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 1st April 1998
Category II

Problem 2 Decide, whether there is a member in the arithmetic progression {an }∞ n=1 with first member
a1 = 1998 and common difference d = 131 which is a palindrome (a palindrome is a number whose decadic
expression is symmetric, e.g. 7, 33, 43334, 2135312 and so on).
Solution We know that 10131−1 ≡ 1 (mod 131) (because 131 is a prime). We shall construct the following
number which shall be a member of our arithmetic progression:

. . 0} 1 |0 .{z
A = 8991 |0 .{z . . 0} 1 |0 .{z
. . 0} 1 . . . . . . . . . 1 |0 .{z
. . 0} 1 |0 .{z
. . 0} 1998
126-times 130-times 130-times 130-times 126-times

(there is a suitable number of ones inside of the number - we shall show precisely later). We can express the
number in a more precise manner:
r
X
A = 8991 · 10126 · 10130·r + 10130·i + 1998
i=1

Now we can find out for which walues of r is the number a member of our arithmetic progression (A is a
member of the sequence if 131 divides A − 1998):
r
X r
X
A − 1998 = 8991 · 10126 · 10130·r + 10130·i ≡ t · 1r + 1i ≡ t + r (mod 131),
i=1 i=1

where t is the remainder on dividing 8991 · 10126 by 131. Now if t + r is a multiple of 131 then A will be a
member of the sequence. But there are infinitely many such values of r which solve the problem. 
 The 8th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 1st April 1998
Category II

Problem 3 Show that the roots of the real polynomial

P (z) = a0 z n + a1 z n−1 + · · · + an−1 z + an ,

where 0 < a0 < · · · < an , satisfy |z| > 1.

Solution Observe that for any z 6∈ (0, 1] on the unit disc

Xn
|(1 − z)P (z)| = | aj (z n−j − z n+1−j )|
j=0
n−1
X
> |an | − ( |aj − aj+1 | · |z n−j |)
j=0
n−1
X
≥ |an | − |aj − aj+1 |
j=0
n−1
X
= an − ( (aj+1 − aj ) + a0 ) = 0.
j=0

The strict inequality follows from the fact, that if

n−1
X n−1
X
| (aj − aj+1 )z n−j − a0 z n+1 | = |(aj − aj+1 )z n−j | + |a0 z n+1 |
j=0 j=0

then
|(a0 − a1 )z n | + |(a1 − a2 )z n−1 | = |(a0 − a1 )z n + (a1 − a2 )z n−1 |
and
|(a0 − a1 )z| + |(a1 − a2 )| = |(a0 − a1 )z + (a1 − a2 )|.
Consequently, z > 0. Since |z| ≤ 1, z ∈ (0, 1]; a contradiction.
Hence for z 6= 1, P (z) 6= 0. Since P (z) > 0 for z ∈ (0, 1], the result is proved. 
 The 8th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 1st April 1998
Category II

Problem 4-M A function f : R → R has the property that for every x, y ∈ R there exists a real number t
(depending on x and y) such that 0 < t < 1 and


f tx + (1 − t)y = tf (x) + (1 − t)f (y) . (1)

Does this imply that

x + y f (x) + f (y)
f = (2)
2 2
for every x, y ∈ R?
Solution As is justified by the following counterexample, the answer is negative. Since the sets R \ {0} and R2
have the same cardinality, there exists a surjective mapping γ 7→ (aγ , bγ ) from R \ {0} onto R2 . Let us define

φγ (x) = aγ x + bγ (x ∈ R) for every γ ∈ R \ {0} and φ0 (x) = |x| (x ∈ R).

There also exists a family {Aγ : γ ∈ R} of pairvise disjoint sets such that Aγ is a dense subset of R for every
γ ∈ R \ {0} and A0 = {−1, 0, 1}. Now we can define
(
φγ (x) if x ∈ Aγ , γ ∈ R ,
f (x) = S (3)
0 if x ∈ R \ γ∈R Aγ .

For every x, y ∈ R with x 6= y there exists β ∈ R \ {0} such that

f (y) − f (x) yf (x) − xf (y)

= aβ and = bβ .
y−x y−x

Since Aβ is a dense set, there exists w ∈ Aβ strictly between x and y. Then f (u) = φβ (u) for u ∈ {x, y, w},
hence (1) is satisfied with t = y−u
y−x . On the other hand

 −1 + 1  φ0 (−1) + φ0 (1) f (−1) + f (1)

f = f (0) = φ0 (0) = 0 6= 1 = = .
2 2 2

 The 8th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 1st April 1998
Category II

Problem 4-I Let us consider the first order language L with just one 3-ary predicate P LU S; hence (well-
formed) formulas of L contain symbols for variables, logical connectives, quantifiers, brackets, and the predicate
symbol P LU S: (∃x1 )(∀x2 ) : (Plus(x2 , x1 , x2 ) ∧ (∀x3 ) : ¬Plus(x1 , x3 , x3 )) is example of such formula). Recall
that a formula is closed iff each variable symbol occurs within the scope of a quantifier.
Show that there is an algorithm which decides whether or not a given closed formula of L is true for the
set N of natural numbers ({0, 1, 2, . . . }) where Plus(x, y, z) is interpreted as x + y = z.
Solution We start by showing that there is a finite automaton recognizing the language described in the Hint,
let us denote it L1 . But this is obvious when we consider the reverse image (L1 )R (the respective automaton
just checks correctness of the binary addition), and recall that the class Reg of languages recognizable by finite
automata is closed under the reverse operation.
The desired algorithm works as follows:
The given formula is transformed into the prenex form

(Q1 x1 )(Q2 x2 ) . . . (Qn xn )F(x1 , x2 , . . . , xn )

(where Qi is either ∃ or ∀).

By generalizing the above construction of finite automaton, we construct an automaton An which accepts
exactly those words which represent n-tuples a1 , a2 , . . . , an s.t. F(a1 , a2 , . . . , an ) is true. This can be done by
using the fact that Reg is closed under union and complement (as well as intersection).
When Qn = ∃ we can, from An , construct An−1 which accepts exactly the words representing (n − 1)-tuples
a1 , a2 , . . . , an−1 s.t. (∃xn )F(a1 , a2 , . . . , an−1 , xn ) is true as follows: An−1 will simulate An by nondeterministi-
cally guessing the appropriate bits of the n-th number; it can also start by a sequence of ε-moves (not reading
input) which capture the possibility that the n-th number has more significant bits than the other numbers. Of
course, An−1 can be made deterministic by a standard construction.
If Qn = ∀ we realize that (∀xn )F(a1 , a2 , . . . , an−1 xn ) is equivalent to

¬(∃xn )¬F(a1 , a2 , . . . , an−1 , xn )

and proceed similarly (Reg is closed under complement).

Thus we successively construct automata An , An−1 , . . . , A1 , A0 . In the end, it remains to check if A0 accepts
the empty word. 
 The 9th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 24th March 1999
Category I

Problem 1 Find the limit

n 
Y 
1999 k 
lim (e n − 1) ln .
n→∞ k+n
k=1

Solution
n 
Y  1999 n 
Y 
1999 k  (e n − 1) 1999 k 
lim (e n − 1) ln = lim 1999 ln
n→∞ k+n n→∞
n
n k+n
k=1 k=1
n
1X  k 
= 1999 lim ln
n→∞ n k+n
k=1
n
1  1 
X
= 1999 lim ln
n→∞ n 1 + nk
k=1
Z 1 
x 
= 1999 ln dx = −3998 ln 2
0 x+1


 The 9th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 24th March 1999
Category I

Problem 2 Find all natural numbers n ≥ 1 such that the following implication holds

(a, b - natural, 11 | an + bn ) ⇒ (11 | a and 11 | b) .

Solution Observe that for odd natural number n we have 11 | 101 + 11 and

102k+1 + 1 = 9 · 11 · 102k−1 + 102k−1 + 1

for any natural number k. Hence our assertion follows easily by an induction argument. Consequently, the
required implication is false odd natural numbers.
We prove it for all even neutral numbers. If n = 2m, then an = (am )2 and bn = (bm )2 are squares
of natural numbers. Hence am = 11e + f and bm = 11g + h for some natural numbers e, g and f, h ∈
{0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10}. Therefore standard computation gives

an = 11A + C and bn = 11B + D,

for some natural numbers A, B and C, D ∈ {0, 1, 3, 4, 5, 9}. By simple calculation one can check that 11 | C + D
if and only if C = D = 0. This implies that 11 | an and 11 | bn . This yields the assertion. 
 The 9th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 24th March 1999
Category I

Problem 3 Suppose that we have countable set A of balls and a unit cube in R3 . Let us also assume that
for every finite set B, which is a subset of A, it is possible to put all the balls from B into the cube in such a
way that they have disjoint interiors. Show that it is possible to arrange all the balls in the cube that all have
pairwise disjoint interiors.
Solution

 The 9th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 24th March 1999
Category I

Problem 4 Show that for complex numbers x, y the following implication folows:
x + y, x2 + y 3 , x3 + y 3 , x4 + y 4 are integers, then for all natural n the numbers xn + y n are also integers.
Solution Notice that (x + y)2 − (x2 + y 2 ) = 2xy ∈ Z, moreover −(x4 + y 4 ) + (x2 + y 2 )2 = 2x2 y 2 ∈ Z.
So it follows that xy is of the shape n2 for n an integer. From the second relation mentioned above we infer
2
that n2 is integer. Hence n is even and xy is integer.
So we arrive at x + y, xy ∈ Z. The rest of the solution is by induction. Namely for the n < 5 the validity
is granted. So, assume that for some natural k > 4 the numbers xm + y m are integers for all m < k. Now we
k k k
consider xk + y k . If k is even then xk + y k = (x 2 + y 2 )2 − 2(xy) 2 is integer. Otherwise it k is odd then it is
divisible by a prime p, and
k k k k k
xk + y k = (x p )p + (y p )p = (x p + y p )(xkp(p−1) − xkp(p−2) y p + . . . ) = . . .

This ends the proof. 

 The 9th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 24th March 1999
Category II

Problem 1 Find the minimal k such that every set of k different lines in R3 contains either 3 mutually parallel
lines or 3 mutually intersecting lines or 3 mutually skew lines.
Solution
1. Let us show that k > 8:
Let ABCDA0 B 0 C 0 D0 be a cube and let K, L, M, N be the centres of the edges A0 B 0 , B 0 C 0 , C 0 D0 , D0 A0 ,
respectively. The lines AB, BC, CD, DA, KL, LM, M N and N K form an 8-tuple which does not contain
any triple either of parallel or intersecting or skew lines.
2. Let us show that k ≤ 9:
Lemma If we have 5 different lines among which no 2 are parallel then choosing any other line causes
that we will get 3 intersecting lines or 3 skew lines.
Proof Let us consider the graph G = (V, E) and its colouring f : E → {1, 2, 3} having the following
properties:
V = {the chosen lines}, E = {[p1 , p2 ]; pi ∈ V } and

f (p1 , p2 ) = 1 if p1 is parallel to p2
f (p1 , p2 ) = 2 if p1 is intersecting to p2
f (p1 , p2 ) = 3 if p1 is skew to p2

If we have the 5-vertex complete graph coloured by two colours (2 and 3) then either it already contains
a single-colour triangle (and we have 3 intersecting or 3 skew lines) or our graph is isomorphic to the
following one:

In this case, adding of a 6-th vertex causes that a triangle of colour 2 and 3 will appear. 

Parallelness is a transitive property, i.e. if f (p1 , p2 ) = f (p1 , p3 ) = 1 then 3 parallel lines exist. Let us
now consider a 9-vertex graph with the above mentioned colouring. Then it is clear that there is at least
one edge of colour 1 in each 5-tuple. Since the maximal number of edges of colour 1 is four, the proof is
finished.

 The 9th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 24th March 1999
Category II

Problem 2 Let a, b ∈ R, a ≤ b. Assume that f : [a, b] → [a, b] satisfies

|f (x) − f (y)| ≤ |x − y|

for every x, y ∈ [a, b]. Choose an x1 ∈ [a, b] and define

xn + f (xn )
xn+1 = , n = 1, 2, 3, . . . .
2
Show that {xn }∞
n=1 converges to some fixed point of f .
Solution Let
W = {w ∈ [a, b] : xnk → w for some subsequence {nk }} .
It is clear that W is nonempty, compact subset of [a, b]. Let g : W → R+ be defined by

g(w) = |w − f (w)| .

First we show that e = inf g(w) = 0. If not, let

w∈W

W1 = {w ∈ W : g(w) = e} .

Since W is compact and g is continuous, W1 6= ∅. Set

A = {w ∈ W1 : e = w − f (w)}
wA +f (wA )
and B = W1 \ A. Suppose A 6= ∅. Let wA = min A. Put w1 = 2 . It is clear that w1 ∈ W . Observe that

g(w1 ) ≤ |f (w1 ) − f (wA )| + |f (wA ) − w1 | ≤

 wA + f (wA ) 
≤ |w1 − wA | + f (wA ) − = wA − f (wA ) = e .
2
Consequently, w1 ∈ W and since w1 < wA , w1 ∈ B. But then f (wA ) < w1 < wA < f (w1 ). Since f satisfied
6 ∅, taking xB = max B and reasoning in the same
the Lipschitz condition, this leads to a contradiction. If B =
manner, we get a contradiction.
Consequently e = 0. Take any w ∈ W1 . Then

w + f (w)
|xn+1 − w| = |xn+1 − | ≤ |xn − w| .
2
Hence xn → w and w is obviously a fixed point of f . 
 The 9th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 24th March 1999
Category II

Problem 3 Suppose that we have a countable set A of balls and a unit cube in R3 . Assume that for every finite
subset B of A it is possible to put all balls of B into the cube in such a way that they have disjoint interiors.
Show that it is possible to arrange all the balls in the cube so that all of them have pairwise disjoint interiors.
Solution We number the balls (as the set A is countable) as P1 , P2 , . . . . It is possible for every n to arrange
the balls P1 , . . . , Pn in the cube in such a way that they have disjoint interiors. Let the p1,n , . . . , pn,n be the
centres of the balls P1 , . . . , Pn in the mentioned arrangement. So we have following sequence:

p1,1 p1,2 p2,2 p1,3 p2,3 p3,3 . . . . . . . . .

Now as the points p1,n lie in a compact set there is a subsequence p1,n1 , p1,n2 , . . . convergent to some point R1 .
Then we choose the subsequence p2,nj1 , p2,nj2 , . . . converging to a point R2 , and so forth. In this way we get a
sequence of points R1 , R2 , . . . . If we put the ball Pi in a position with the centre in Ri then this arrangement
satisfies the requirements concerning the interiors. 
 The 9th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 24th March 1999
Category II

Problem 4 Let u1 , u2 , . . . , un ∈ C([0, 1]n ) be nonnegative and continuous functions, and let uj do not depend
on the j-th variable for j = 1, . . . , n. Show that
Z n
Y n−1 n Z
Y
uj ≤ un−1
j .
[0,1]n j=1 j=1 [0,1]n

Solution
!2
Z 1 Z 1 Z 1
u1 (y, z)u2 (z, x)u3 (x, y) dx dy dz =
0 0 0
! !2
Z 1 Z 1 Z 1
u1 (y, z) u2 (z, x)u3 (x, y) dy dz ≤
0 0 0
! Z !2 !
Z 1 Z 1 1 Z 1 Z 1
u21 (y, z) dy dz u2 (z, x)u3 (x, y) dy dz ≤
0 0 0 0 0
! Z ! Z ! !
Z 1 Z 1 1 Z 1 Z 1 1
u21 (y, z) dy dz u22 (z, x) dx u23 (x, y) dx dy dz =
0 0 0 0 0 0
! Z !
Z 1 Z 1 1 Z 1 Z 1 Z 1 Z 1 Z 1
u21 (y, z) dy dz u22 (z, x) dx dy dz u23 (x, y) dx dy dz =
0 0 0 0 0 0 0 0
! Z ! Z !
Z 1 Z 1 1 Z 1 1 Z 1
u21 (y, z) dy dz u22 (z, x) dz dy u23 (x, y) dx dy .
0 0 0 0 0 0

Both inequalities use the Hölder inequality for p = 2. Then we use Fubini’s theorem (which is elementary for
continuous functions on a compact interval). When passing to the last line, we use the fact that an integral
over the interval [0, 1] with integrand not depending on the integration variable can be omitted. 
 The 10th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 5th April 2000
Category I

Problem 1 Is there a countable set Y and an uncountable family F of its subsets, such that for every two
distinct A, B ∈ F their intersection A ∩ B is finite?
Solution The answer is yes.
Put all natural numbers N in a coordinate system like in the picture.

10 p

6 9 ..
.

3 5 8 12

1 2 4 7 11
x

For each ray p we put in the set Ap all numbers, which p intersects (intersects its square), and we place all such
sets Ap into F. For each ray p we assign the angle αp , which is between p and x. Since all αp form the interval
(0, π2 ), the set of rays p is uncountable. Furthermore, for different rays p1 and p2 , the intersection Ap1 ∩ Ap2 is
finite, because there exists distance dp1 p2 from the origin of the coordinate system where rays are far enough,
so they will not pass through the same square any more. The problem is solved. 
Solution Yes. Let Y = N and denote S set of all infinite sequences {an } of 0 and 1. To each sequence {an }
assign the set C{an } ∈ F in the following way:

n k
X o
C{an } = 2k + an · 2n−1 , for k = 1, 2, . . . .
n=1

Suppose we have two distinct sets C{an } , C{bn } for some distinct {an }, {bn } ∈ F.
Suppose now
k1
X k2
X
k1 n−1 k2
2 + an · 2 =2 + bn · 2n−1 for some k1 , k2 ∈ N,
n=1 n=1
 k1
X   k2
X 
where 2k1 + an · 2n−1 ∈ C{an } and 2k2 + bn · 2n−1 ∈ C{bn } .
n=1 n=1

This is like equality of two numbers written in binary system. For the number
k1
X
2k1 + an · 2n−1
n=1

is the first digit regarding to the 2k1 th digit 1, and the next digits are ak1 −1 , ak1 −2 , . . . , a1 . Analogous are the
digits for the other number. So these numbers are equal if and only if k1 = k2 and ai = bi for all i < k1 . So if
the sets C{an } , C{bn } contain infinitely many of the same numbers then ai = bi for all i ∈ N and C{an } = C{bn } ,
a contradiction. 
 The 10th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 5th April 2000
Category I

Problem 2 Let f : N → R be given by

τ (n)
f (n) = n 2 ,
n ∈ N = {1, 2, . . . }, τ (n) - the number of divisors of n. Show that f is injective into N.
Solution Recall that every natural number n greater than 1 can be written uniqely (up to order) as a product

n = pa1 1 pa2 2 . . . pakk ,

where the pi are different primes and the ai are natural numbers. Then

τ (n) = (1 + a1 )(1 + a2 ) . . . (1 + ak ) .

Now we show that for every natural number n, f (n) is also a natural number:
• f (1) = 1 is natural,
τ (n)
• if n > 1 and τ (n) is even, then 2 is natural and so is f (n),
• if n > 1 and τ (n) is odd, then all ai are even so n is a square of a natural number and f (n) is natural too.

Now it is easy to see that prime number p divides the natural number n if and only if it divides f (n). We use
this fact to prove injectivity of f .
Let f (m) = f (n). Then m and n are divisible by the same primes and we can write n = pa1 1 pa2 2 . . . pakk ,
m = pb11 pb22 . . . pbkk . As f (n) = f (m), it is true that ai τ (n) = bi τ (m), i = 1, 2, . . . k. We can assume that
τ (n) ≤ τ (m) (if not we change m and n). Using this we get ai ≥ bi and τ (n) ≥ τ (m). From this follows that
τ (n) = τ (m), ai = bi and n = m. 
αk
Solution Let us write n in the form n = pα 1 α2
1 p2 . . . pk , where p1 , p2 , . . . , pk are prime divisors (this represen-
tation is unique), and let 1 = d1 < d2 < · · · < dτ (n) = n be all its divisors. Then
q p τ (n)
d1 d2 . . . dτ (n) = (d1 dτ (n) ) · (d2 dτ (n)−1 ) . . . (dτ (n) d1 ) = nτ (n) = n 2 = f (n),

since dk dτ (n)−k+1 =n. So f (n) is natural, because it can be expressed as multiple of natural numbers d1 , . . . ,
τ (m) τ (n)
dτ (n) . Suppose now f (n) = f (m) for distinct natural numbers m, n. From m 2 = f (m) = f (n) = n 2 it
τ (n)
follows that m = n τ (m) . This implies that m and n have the same set of prime divisors, so m can be writen
τ (n)
as m = pβ1 1 pβ2 2 . . . pβkk .
The obtained relation m = n τ (m) implies αβ11 = αβ22 = · · · = αβkk = c. Without loss of
generality we can suppose that m > n. Then c > 1, and, since all αi and βi are positive integers, m is divisible
by n. So the set of all divisors of n is a subset of the set of all divisors of m, and τ (m) ≥ τ (n). From m > n we
τ (m) τ (n)
can conclude that m 2 > n 2 , a contradiction. 
 The 10th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 5th April 2000
Category I

Problem 3 Let a1 , a2 , . . . be a bounded sequence of reals. Is it true that

N N
1 X 1 X an
lim an = b and lim =c
N →∞ N N →∞ log N n
n=1 n=1

imply b = c?
N
1
P
Solution We prove that if N an → b for any (not even necessarily bounded) sequence of reals then
n=1

N
1 X an
→ b.
log N n=1 n

Assume that
N
1 X
an → b .
N n=1
Define
N
X
hN = an .
n=1

(We have h0 = 0.)

Then by our assumption we have hN → b and by definition we get

aN = N hN − (N − 1)hN −1 = N (hN − hN −1 ) + hN −1 .

Thus
N N N
X an X hn−1 X hn−1
= hn − hn+1 + = hN + .
n=1
n n=1
n n=1
n
Therefore
N N
1 X an hN X hn
= + .
log N n=1 n log N n=1 n
N N
1
P 1 1
P an
Since hN converges the first term goes to 0. As log N n → 1 and hN → b we get that log N n → b. 
n=1 n=1
n n
1
P 1
P ai
Solution Denote n ai = bn , n ≥ 1, b0 = 0 and log n i = cn , n ≥ 1. The sequence {bn }n∈∞ converges to
i=1 i=1
b.
Since an = nbn − (n − 1)bn−1 , we obtain
n n
1 X ai 1 X ibi − (i − 1)bi−1
cn = =
log n i=1 i log n i=1 i
n n−1
1 X i−1
1 X bi

= bi − bi−1 = bn − .
log n i=1 i log n i=1
i+1

Let us write bn = b + εn , where ε → 0. Then

Xn n−1  Pn Pn−1 εi
1 1 X εi
i=1 i=1 i+1 εn
cn = b+ + εn = b+ + .
log n i=1
i i=1
i+1 log n log n log n
Pn 1
εn
Easily we see i=1 i
log n → 1, log n → 0. For the rest we have
Pn−1 εi Pk εi Pn−1 εi
i=1 i+1 i=1 i+1 i=k+1 i+1
= +
log n log n log n
and it follows that
Pk εi
i=1 i+1
→0 for n → ∞,
log n
 Pn−1 εi  Pn−1 1
 i=k+1 i+1 
  ≤ sup{|εj |} · i=k+1 i
≤ sup{|εj |} for all n ≥ k.
 log n  j≥k log n j≥k

Since sup{|εj |} → 0 for k → ∞ and our choice of k can be arbitrary large, we got cn → b, and the problem is
j≥k
solved. 
 The 10th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 5th April 2000
Category I

Problem 4 Let us choose arbitrarily n vertices of a regular 2n-gon and colour them red. Remaining vertices are
coloured blue. We arrange all red-red distances into a nondecreasing sequence and do the same with blue-blue
distances. Prove that the sequences are equal.
Solution Let 1 ≥ m ≥ n be the combinatorial distance (i.e. the Euclidean distance dm between the first and
(m + 1)-st vertices corresponds to m). For a given m denote by k the gcd of 2n and m. The original 2n-gon
can be decomposed into k disjoint 2nk -gons of edge lenght dm . Assume that there are ri red and bi blue vertices
on the i-th 2nk -gon, and denote ci to be the number of oriented red → blue changes between the neighboring
vertices. The number of neighboring blue pairs equals ri −P ci , while the number
P of neighboring blue pairs equals
bi − ci . The total number of neighoring red pairs will be (ri − ci ) = n − ci which is the same as the total
P P i i
number (bi − ci ) = n − ci of neighboring blue pairs. 
i i
Solution It is enough to show, that the number of each particular distance is same among both red–red and
blue–blue pairs. Choose one of the used distances, let us say d. Denote drr to be the number of red–red pairs
with distance d and analogously dbb number of blue–blue pairs and drb number of red–blue pairs with distance
d. Clearly 2drr + drb = 2n, since 2× number of red–red pairs plus 1× red–blue pairs gives 2× number of red
vertices. Analogously 2dbb + drb = 2n. By subtracting one relation from the other we get 2dbb = 2drr , which is
the end. 
 The 10th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 5th April 2000
Category II

Problem 1 Let p be a prime of the form p = 4n − 1 where n is a positive integer. Prove that
p
Y
(k 2 + 1) ≡ 4 (mod p) .
k=1

Qp Qp−1
Solution Consider the polynomials P (x) = k=1 (k 2 − x2 ) and p(x) = k=1 (k − x). Easily we see that
P (x) = p(x) · p(−x) · (p2 − x2 ).
Since p(x) is of degree p − 1 and has roots 1, 2, . . . , p − 1, p(x) is congruent modulo p to the polynomial
q(x) = xp−1 − 1, which has also roots 1, 2, . . . , p − 1. Therefore

P (x) = p(x)p(−x)(p2 − x2 ) ≡ q(x)q(−x)(p2 − x2 ) ≡ q(x)q(−x)(−x2 ) (mod p)

and it follows that

p
Y
(k 2 + 1) = P (i) ≡ q(i) · q(−i) · (−i2 ) = (ip−1 − 1) · (−i)p−1 − 1

k=1
= (i4n−2 − 1) · (−i)4n−2 − 1 = (−2) · (−2) = 4


(mod p) .


 The 10th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 5th April 2000
Category II

Problem 2 If we write the sequence AAABABBB along the perimeter of a circle, then every word of the
length 3 consisting of letters A and B (i.e. AAA, AAB, ABA, BAB, ABB, BBB, BBA, BAA) occurs exactly
once on the perimeter. Decide whether it is possible to write a sequence of letters from a k-element alphabet
along the perimeter of a circle in such a way that every word of the length l (i.e. an ordered l-tuple of letters)
occurs exactly once on the perimeter.
Solution Let us denote the alphabet  by P . Let us form  the directed graph G = (V, E), where V =
{[a1 , . . . , al−1 ]; ai ∈ P } and E = { [a1 , . . . , al−1 ], [b1 , . . . , bl−1 ] ; a2 = b1 , a3 = b2 , . . . , al−1 = bl−2 }.
First, considering any two vertices [a1 , . . . , al−1 ] and [b1 , . . . , bl−1 ], we find that there must be at least one
oriented path between them:

[a1 , . . . , al−1 ] → [a2 , . . . , al−1 , b1 ] → [a3 , . . . , al−1 , b1 , b2 ] → · · · → [b1 , . . . , bl−1 ]

(some vertices and arcs can repeat in the sequence). This implies that the graph is strongly connected.
Second, we realize that every vertex [a1 , . . . , al−1 ] has exactly k outgoing and k ingoing arcs: the outgoing arcs
are directed to the vertices [a2 , . . . , al−1 , o], where o goes through the the whole alphabet P , and the ingoing
come from vertices [i, a1 , . . . , al−2 ], where i also goes through the whole alphabet. That means that the inner
and outer degrees of every vertex are identical and the graph is an Euler (directed) graph. As a consequence,
there exists an Eulerian cycle in it, i.e. a cycle containing all arcs going through every arc exactly once. We
can now form the searched for cyclic sequence as follows:
Let us start with an arbitrary vertex and write down its sequence a1 , . . . , al−1 .
Let us follow the Eulerian cycle and add the last letter of every vertex to the sequence until we reach the starting
vertex again. Now we delete the last l − 1 letters (which are necessarily the same as the starting ones).
Since there is a bijection between the set of all l-letter words and the set of arcs V :
 
[a1 , . . . , al ] ←→ [a1 , . . . , al−1 ], [a2 , . . . , al ] ,

it is clear that the sequence has the demanded properties. 

 The 10th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 5th April 2000
Category II

Problem 3 Let m, n be positive integers and let x ∈ [0, 1]. Prove that

n
(1 − xn )m + 1 − (1 − x)m ≥ 1 .

Solution We will prove that (1 − xn )m ≥ 1 − (1 − (1 − x)m )n .

Take an m × n chessboard. The probability, that one particular square is black, is x ∈ [0, 1], the probability
of being white is x − 1. Assume this for all squares. Then
(1 − x)m is the probability that the whole row is white
1 − (1 − x)m is the probability that there is at least one black in the row
(1 − (1 − x)m )n is the probability that in each row there is at least one black
1 − (1 − (1 − x)m )n is the probability that at least one row does not contain a black.
Denote by A the last event, in which some row does not a contain black (it is all white). We continue:
1 − xn is the probability that the column contains at least one white
(1 − xn )m is the probability that each column contains at least one white.
Denote by B the event in which each column contains at least one white. It is clear that A ⊂ B, because if
one row is white then each column contains some white. Therefore is P (b) ≥ P (A), written in the other form:
(1 − xn )m ≥ 1 − (1 − (1 − x)m )n . 
 The 10th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 5th April 2000
Category II

S

Problem 4 Let B be a family of open balls in Rn and c < λ B where λ is the n-dimensional Lebesgue
measure. Show that there exists a finite family of pairwise disjoint balls {Ui }ki=1 ⊆ B such that
k
X c
λ(Uj ) > .
j=1
3n

Solution Suppose first sup λ(U ) < ∞ for U ∈ B. In other case we just take a large enough ball U0 for which
U ∈B
c
λ(U0 ) > 3n .
S

B
S  λ
Take ε > 0. We first construct the infinite sequence {Uk }∞
k=1 of disjoint balls for which λ Uk ≥ (3+ε)n .
The procedure is the following. In the k-th
step count sup λ(U ) through all the rest of the balls in B, then
3 n
choose Uk such that λ(Uk ) ≥ sup λ(U ) · 3+ε and remove from B all balls U which intersect ball Uk . Continue
to infinity by the (k + 1)-th step. It is clear, that the balls in the constructed sequence {Un } are disjoint.
Further, if we increase each ball Uk from {Uk }∞k=1 (3 + ε)-times, than they will contain all of the set B. This
holds, because ball Ui increased (3 + ε)-times covers all balls intersecting Ui removed from B in i-th step. It
S  λ S B
S

follows that λ Uk ≥ (3+ε)n . If λ B < ∞, then there is k0 ∈ N for which

k0 S

λ B
[ 
λ Uk ≥ .
(3 + 2ε)n
k=1

S

λ B
> 3cn (this is always possible) we solve the problem. In case λ
S

By choosing ε so that (3+2ε)n B = ∞ we just
choose k0 big enough to exceed the finite constant c in the desired inequlity. 
 The 11th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2001
Category I

Problem 1 Prove that for an arbitrary prime p ≥ 5 the number

X p
2p
k
0<k< 3

is divisible by p2 .
1 p 2p
is congruent to (−1)k−1 k1 modulo p. Hence it is sufficient to show


Solution The number p k for 0 < k < 3
that the element X X 1
(1) = (−1)k−1
k
0<k<2p/3

is 0 in a finite field Fp . The sum

X
X1 1 1 X 1 1 1 
2 = − − + − +
p k 2k 2k p p k 2k p − 2k
k< 6 6 <k< 3

1 1
P
P

is 0 in Fp . But 1 = 2 . In fact the terms of the shape 2k+1 are evidently the same. As to a term 2k
P
p 1 p


in the 2 for 2k < 3 it appears with the coefficient −1, what is
O.K. The
term 2k for 2k < 3 appears in
P
P P
2 twice with the sign “−” and once with the sign “+”. So 1 = 2 . 
 The 11th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2001
Category I

Problem 2 Let n ≥ 2 be a natural number. Prove that

n √
Y n!
ln k < .
n
k=2

Solution Consider f : [1, ∞) → R defined by

1
f (t) = 2 ln t − t + .
t
We have
(t − 1)2
f 0 (t) = − ,
t2
hence f 0 is negative on (1, ∞). Therefore f is strictly decreasing. Since f (1) = 0, the function f has negative
values on (1, ∞). So
 1 1
(∀t ∈ (1, ∞)) 2 ln t − t + < 0.
t t2 − 1
Putting x = t2 in the above inequality we obtain that
ln x 1
(∀t ∈ (1, ∞)) <√ .
x−1 x

Hence
n n
Y ln k Y 1 1
< √ =√
k−1 k n!
k=2 k=2

and √
n
Y (n − 1)! n!
ln k < √ = .
n! n
k=2


√ k−1
Solution
√ We prove that ln k < k· k for k ∈ N and ≥ 2. Let us consider the functions f (x) = ln x and
g(x) = x · x−1
x . It is

1
f (1) = g(1) = 0 and f 0 (x) = ,
x
1 x−1 √ 1
g 0 (x) = √ + x 2,
2 x x x
1 √
1 √
2
g 0 (x) − f 0 (x) = √ x−1+2−2 x = √ x−1 ,
2 x 3 2 x 3

which is = 0 for x√= 1 and > 0 for x > 1. It proves that g(x) > f (x) for x > 1, as we needed. Now by
multiplying ln k < k · k−1
k over k = 2, 3, . . . , n we get

n n √ n−1 √
Y 
Y 1 2 1
ln k < k · · · ... · = n! · ,
2 3 n n
k=2 k=2

the problem solved. 

 The 11th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2001
Category I

Problem 3 Let A, B, C be sets in Rn . Suppose that A is nonempty and bounded, that C is closed and convex,
and that A + B ⊆ A + C. Show the inclusion B ⊆ C.
We remind you that
E + F = {e + f : e ∈ E, f ∈ F }
and D ⊆ Rn is convex when
∀x, y ∈ D ∀t ∈ [0, 1] : tx + (1 − t)y ∈ D .
Solution We will use the following lemma.
Lemma Let a1 , . . . , am be points of a convex set D ⊆ Rn . Let λ1 , . . . , λm ≥ 0 with λ1 + · · · + λm = 1. Then
λ1 a1 + · · · + λm am ∈ D.
Proof We argue by induction on m. When m = 1 the assertion is trivial. Suppose that the assertion holds
when m is some positive integer k. Let

x = λ1 a1 + · · · + λk+1 ak+1 ,

where a1 , . . . , ak+1 ∈ D and λ1 , . . . , λk+1 ≥ 0 with λ1 + · · · + λk+1 = 1. At least one λi must be less than 1, say
λk+1 < 1. Write
λ1 λk
y= a1 + · · · + ak ,
λ λ
where
λ = λ1 + · · · + λk = 1 − λk+1 > 0 .
By the induction hypothesis, y ∈ D. Since D is convex and contains both y and ak+1 the equation x =
λy + λk+1 ak+1 shows that x ∈ D. This completes the proof by induction. 

Let a0 ∈ A. If b ∈ B, then a0 + b ∈ A + B ⊆ A + C, and so there exists a1 ∈ A, c1 ∈ C such that

a0 + b = a1 + c1 . Similarly, there exist a2 , . . . , ai ∈ A and c2 , . . . , ci ∈ C with

a1 + b = a2 + c2 , . . . , ai−1 + b = ai + ci .

We add the i equations above together to deduce that

a0 + ib = a1 + c1 + · · · + ci .

Since C is convex, the point xi defined by the equation

1
xi = (c1 + · · · + ci )
i
lies in C (Lemma). Now
1
kb − xi k =
kai − a0 k → 0 as i → ∞ ,
i
since A is bounded. Thus xi → b as i → ∞. But C is closed, whence b ∈ C and B ⊆ C. 
Solution For contradiction suppose there is b ∈ B which ∈ / C. Since C is convex and closed, there existS
(n − 1)-dimensional hyperplane H such that it separates b and C. Denote ~n the normal vector of H orientated
in direction of point b. Now every point x of space Rn can be expressed as x = hx + a~n, where hx ∈ H and
a ∈ R. From this define linear function f (x) = a. It is clear that f (b) > 0 and f (C) < 0. Take now sup f (a) (it
a∈A
is finite since A is bounded) and point a0 such that f (a0 ) > sup f (a) − f (b). Then clearly, since function f is
a∈A
linear, it holds
 
f (a0 + b) = f (a0 ) + f (b) > f (a) − f (b) + f (b) > f (a) + f (c) = f (a + c) ,

for all a ∈ A and c ∈ C. But it is contradiction with f (A + B) ⊆ f (A + C) (which follows from A + B ⊆ A + C).

 The 11th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2001
Category I

Problem 4 Let A be a set of positive integers greater than 0 such that for any x, y ∈ A, x > y,
xy
x−y ≥ .
25
Find the maximal possible number of elements of the set A.
Solution For x > y ≥ 25 we have
xy
x−y <x≤ .
25
Hence A contains at most one element greater than 24. Let A = {x1 , x2 , . . . , xn } where x1 < x2 < · · · < xn ,
xn−1 < 25. For the differences dj = xj+1 − xj , 1 ≤ j ≤ n − 1, we get

xj+1 xj (xj + dj )xj

dj ≥ = ,
25 25
which yields
x2j
dj ≥ .
25 − xj
x2
Since the function g(x) = 25−x is increasing in the interval [0, 25), we obtain succesively

x5 ≥ 5, d5 ≥ g(5) > 1,
x6 ≥ 7, d6 ≥ g(7) > 2,
x7 ≥ 10, d7 ≥ g(10) > 6,
x8 ≥ 17, d8 ≥ g(17) > 36,
x9 ≥ 54.

So, we get n ≤ 9. Simultaneously, we can see that the set with 9 elements

A = {1, 2, 3, 4, 5, 7, 10, 17, 54}

satisfies all the conditions. 

 The 11th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2001
Category II

Problem 1 Let n ≥ 2 be an integer and let x1 , x2 , . . . , xn be real numbers. Consider N = n2 sums xi + xj ,

1 ≤ i < j ≤ n and denote them by y1 , y2 , . . . , yN (in arbitrary order). For which n are the numbers x1 , x2 , . . . , xn
uniquely determined by the numbers y1 , y2 , . . . , yN ?
Solution The answer is n 6= 2p .
Denote the kth symmetric polynomial in x1 , x2 , . . . , xn by σk . Further denote
n
X N
X
sk = xki , tk = yik .
i=1 i=1

The numbers x1 , x2 , . . . , xn are uniquely determined by the numbers σ1 , σ2 , . . . , σn and these are uniquely
determined by the numbers s1 , s2 , . . . , sn since we have the following identity:

sk − sk−1 σ1 + sk−2 σ2 − · · · + (−1)k−1 s1 σk−1 = (−1)k−1 kσk .

So we will try to show that s1 , s2 , . . . , sn are determined by the numbers t1 , t2 , . . . , tn . We have

n X n n X n X k  
X X k r k−r
2tk + 2k sk = (xi + xj )k = x x =
i=1 j=1 i=1 j=1 r=0
r i j
k−1
X 
k
= 2nsk + sr sk−r .
r=1
r

For n 6= 2k−1 , we get

k−1
X k 
1  
sk = k
2tk − sr sk−r .
2n − 2 r=1
r
Using induction with respect to k, we can conclude that for n 6= 2p , the numbers t1 , t2 , . . . , tn determine
uniquely the numbers s1 , s2 , . . . , sn .
For n = 2 the numbers from the sets A2 = {0, 3} and B2 = {1, 2} have the same sums. Suppose that we
have two disjoint sets An , Bn , every with n elements, which have the same sums of all possible couples. Then
the sets A2n = An ∪ (c + Bn ) and B2n = Bn ∪ (c + An ) for c large enought are disjoint with 2n elements and
have the same sums of all possible couples. 
 The 11th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2001
Category II

Problem 2 Let f : [0, 1] → R be a continuous function of function {fn }, fn : [0, 1] → R. Define the sequence
in the following way: Z x
fn+1 (x) = ft , n = 0, 1, 2, . . . .
0

Prove that if fn (1) = 0 for all n, then f (x) ≡ 0.

Solution Using induction on k, we prove that for any n, k ≥ 0 integers
Z 1
(1 − t)k fn (t) = k! · fn+k (1) . (1)
0

This is trivial for k = 0. For greater k,

Z 1 Z 1
(1 − t)k fn (t) = [(1 − t)k fn+1 (t)]1t=0 + k (1 − t)k−1 fn+1 (t) =
0 0
= 0 + k · (k − 1)! · f(n+1)+(k−1) (1) = k! · fn+k (1) .
R1
From (1) it follows for an arbitrary polynomial p, that 0 p · f = 0.
By Weierstrass’ theorem, for an arbitrary ε > 0 there exists a polynomial pε such that |pε (t) − f (t)| < ε for
all t ∈ [0, 1]. This implies
Z 1 Z 1 Z 1 Z 1 Z 1
2 2
f = f − pε · f = (f − pε )f ≤ ε |f | .
0 0 0 0 0
R1
This holds for any ε, thus 0
f 2 = 0. This implies f ≡ 0. 
 The 11th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2001
Category II

Problem 3 Let f : (0, +∞) → (0, +∞) be a decreasing function, satisfying

Z ∞
f (x) < ∞ .
0

Prove that lim xf (x) = 0.

x→∞
Solution As first we prove that lim inf xf (x) = 0. Let lim inf xf (x) = c > 0, that implies ∃x0 ∀x > x0 :
x→∞ x→∞
c0
xf (x) > c0 > 0, or f (x) > x, and we get:
∞ ∞ ∞
c0
Z Z Z
f (x) > f (x) > = ∞,
0 x0 x0 x

a contradiction.
Now, let us suppose lim sup xf (x) = c > 0. It implies ∀y∃x > y : xf (x) ≥ 2c . We have also constructed a
x→∞
sequence {xn }∞
n=1 , satisfying:

c c
xn → ∞, and xn f (xn ) ≥ > 0, which is the same as f (xn ) ≥ .
2 2xn
Since f is decreasing: f (x) > f (xn ), for x ∈ (xn−1 , xn ] and
Z ∞ ∞
X
∞> f (x) > (xn − xn−1 )f (xn ) ≥
0 n=1
∞ ∞
c X xn − xn−1 c X xn−1 
≥ = 1− .
2 n=1 xn 2 xn
n=D1

∞  
xn−1
So we have a sequence {xn }∞
P
n=1 such that xn → ∞ and 1− xn < ∞.
n=1
xn−1
To make a proof clearer, we will do a substitution bn = 1 − xn . Sequence {bn }∞
n=1 satisfies:

∞ ∞
X Y x0
bn < ∞ and (1 − bn ) = lim = 0.
n=1 n=1
n→∞ xn

∞
Second condition for a sequence {bn }∞
P
n=1 is the same as − ln(1 − bn ) = ∞.
n=1
From the ratio criterion for convergence of the infinity sums, if
∞
X ∞
X
bn < ∞ and − ln(1 − bn ) = ∞,
n=1 n=1
− ln(1 − bn )
lim =∞
n→∞ bn
∞
P
should hold. But above limit is equal to 1, as can be easy checked by many ways. (From bn < ∞, we get
n=1
bn → 0, and
1
− ln(1 − bn ) − ln(1 − bn ) L’H 1−bn
lim = lim = lim = 1 .)
n→∞ bn bn →0 bn bn →0 1
This yields to contradiction.
As a conclusion we have lim inf xf (x) = 0 and lim sup xf (x) = 0. 
x→∞ x→∞
Solution For contradiction assume that lim xf (x) = 0 is not true. Then it must exist increasing sequence
x→∞
{xi }∞
i=1 , xi → ∞, such that exists ε > 0 that xi f (xi ) > ε for all xi . Moreover, we can choose subsequence
{yi } ⊂ {xi }, such that yi+1 ≥ 2yi . Then following inequalities hold (inequality (∗) holds, because f is decreasing
function):
Z ∞ (∗) X∞ ∞ ∞
1X 1X
f (x) ≥ (yn − yn−1 )f (yn ) ≥ yn f (yn ) ≥ ε = ∞.
0 n=2
2 n=2 2 n=2
R∞
This contradicts the assumption that f (x) < ∞. 
0
 The 11th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2001
Category II

Problem 4 Let R be an associative non commutative ring and let n > 2 be a fixed natural number. Assume
that xn = x, ∀x ∈ R. Prove that xy n−1 = y n−1 x holds ∀x, y ∈ R.
Solution Let a = xn−1 , then

a2 = (xn−1 )2 = x2n−2 = xn xn−2 = xxn−2 = xn−1 = a.

We show that if r2 = 0 then r = 0. Indeed r = rn = rn−2 r2 = 0. If e2 = e then for every x ∈ R:

(ex − exe)2 = (ex − exe)(ex − exe) =

= exex − exexe − exe2 x + exe2 xe = exex − exexe − exex + exexe = 0

and similarly
(ex − exe)2 = 0
so ex − exe = 0 and xe − exe = 0, so for every x ∈ R and every e ∈ R, such that e2 = e we have:

ex = xe

and since for every y ∈ R, (y n−1 ) = y n−1 , we get:

xy n−1 = y n−1 x

for every x, y ∈ R. 
Solution Since R is an integral domain and

y(xy n−1 − y n−1 x)y = yxy n − y n xy = yxy − yxy = 0 ,

it is either xy n−1 − y n−1 x = 0 or y = 0, but that also implies xy n−1 = y n−1 x. The end. 
 The 12th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 10th April 2002
Category I

Problem 1. Differentiable functions f1 , . . . , fn : R → R are linearly

independent. Prove that there exist at least n−1 linearly independent
functions among f10 , . . . , fn0 . [10 points]

Problem 2. Let p > 3 be a prime number and n = (22p − 1)/3.

Show that n divides 2n − 2. [10 points]

Problem 3. Positive numbers x1 , . . . , xn satisfy

1 1 1
+ + ··· + = 1.
1 + x1 1 + x2 1 + xn

Prove that
 
√ √ √ 1 1 1
x1 + x2 + · · · + xn ≥ (n − 1) √ + √ + · · · + √ .
x1 x2 xn

[10 points]

Problem 4. The numbers 1, 2, . . . , n are assigned to the vertices

of a regular n-gon in an arbitrary order. For each edge compute
the product of the two numbers at the endpoints and sum up these
products. What is the smallest possible value of this sum?
[10 points]
 The 12th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 10th April 2002
Category II

Problem 1. Find all complex solutions to the system

(a + ic)3 + (ia + b)3 + (−b + ic)3 = −6 ,

(a + ic)2 + (ia + b)2 + (−b + ic)2 = 6 ,
(1 + i)a + 2ic = 0 .

[10 points]

Problem 2. A ring R (not necessarily commutative) contains at

least one zero divisor and the number of zero divisors is finite. Prove
that R is finite. [10 points]

Problem 3. Let E be the set of all continuous functions u: [0, 1] → R

satisfying
Z t
u2 (t) ≤ 1 + 4
 
su(s) ds , ∀t ∈ [0, 1] .
0

Let ϕ: E → R be defined by
Z 1
u2 (x) − u(x) dx .


ϕ(u) =
0

Prove that ϕ has a maximum value and find it. [10 points]

Problem 4. Prove that

!
1 √ π2
Z
2 n
lim n 1+ xn dx − 1 = .
n→∞ 0 12

[10 points]
Problem j12-I-1/j12-I-4. Differentiable functions f1 , . . . , fn : R → R are linearly indepen-
dent. Prove that there exist at least n − 1 linearly independent functions among f10 , . . . , fn0 .
(Eötvös Loránd University, Budapest)
Solution. Select a maximal independent set from the derivatives. Without loss of
generality, it can be assumed that this set is f10 , . . . , fm
0
, where m ≤ n. If m ≤ n − 2, then
fn−1 and fn can be expressed as a linear combination of f10 , . . . , fm
0 0 0
; hence, there exist real
numbers a1 , . . . , am , b1 , . . . , bm such that

m m
!0
X X
ai fi0 − fn−1
0
= ai fi − fn−1 =0
i=1 i=1
and
m m
!0
X X
bi fi0 − fn0 = bi fi − fn = 0.
i=1 i=1

Pm Pm
This implies that functions i=1 ai fi − fn−1 and i=1 bi fi − fn are constant. Eliminating
these constants, a linear combination of f1 , . . . , fn is found which vanishes. 

j12-I-1/j12-I-4-1
 2p
Problem j12-I-2/j12-I-9. Let p > 3 be a prime number and n = 2 3−1 . Show that
n divides 2n − 2. (Jagiellonian University in Kraków)
2p
Solution. n = 2 3−1 = 4p−1 +4p−2 +· · ·+1. Hence, in the binary system, n = 1010 . . . 101
(number of 1’s is p). Therefore, in the binary system,

(∗) 2n − 2 = 1111 . . . 110 (number of 1’s is n − 1),

(∗∗) 3n = 1111 . . . 111 (number of 1’s is 2p).

Now if we prove that 2p divides n − 1, then by (∗), (∗∗) and by the rules of multiplication
in the binary system, we will get that 3n divides 2n − 2 — just what we need. But now
observe:
2p | (n − 1) ⇐⇒ (n is odd†) ⇐⇒ p | (n − 1) ⇐⇒
  22p − 1    22p − 4 
⇐⇒ p  − 1 ⇐⇒ p  ⇐⇒
 
3 3
⇐⇒ (p > 3 and prime) ⇐⇒ p | 22p − 4 ⇐⇒


  22p − 4 
⇐⇒ (p > 3 and prime) ⇐⇒ p  ⇐⇒

4
⇐⇒ p | 22p−2 − 1 .

But now from Fermat’s small theorem (p prime and p does not divide a, then ap−1 − 1 ≡ 0
(mod p)), we have 2p−1 ≡ 1 (mod p), hence (2p−1 )2 ≡ 12 (mod p) and finally 22p−2 ≡ 1
(mod p). 

† The sentences in parentheses serve only as justifications of the stated equivalences here.
Thus, e.g., 2p | (n − 1) ⇔ (n is odd) ⇔ p | (n − 1) should be read as “2p divides (n − 1) if
and only if p divides (n − 1) because n is odd” and so on.

j12-I-2/j12-I-9-1
Problem j12-I-3/j12-II-59. Positive numbers x1 , . . . , xn satisfy

1 1 1
+ + ··· + = 1. (1)
1 + x1 1 + x2 1 + xn

Prove that
 
√ √ √ 1 1 1
x1 + x2 + · · · + xn ≥ (n − 1) √ + √ + · · · + √ .
x1 x2 xn

(University of Ostrava)
Solution. It is sufficient to prove that
       
√ 1 √ 1 √ 1 1 1 1
x1 + √ + x2 + √ + ··· + xn + √ ≥ n √ + √ + ··· + √
x1 x2 xn x1 x2 xn

or equivalently (see (1))

    
1 + x1 1 + xn 1 1 1 1 1
√ + ··· + √ + ··· + ≥ n √ + √ + ··· + √ . (2)
x1 xn 1 + x1 1 + xn x1 x2 xn
√
Consider the function f (x) = x + √1x = x+1√ , x ∈ (0, +∞). It is easy to check that the
x
function f is non-decreasing on [1, +∞) and that

f (x) = f ( x1 ) (3)

holds for each x > 0.

Further, it follows from (1) that only x1 can be less than 1 (i.e. xk ≥ 1, k = 2, 3, . . .)
1 1 x1
and 1+x 2
≤ 1 − 1+x 1
= 1+x 1
. Hence
1
x2 ≥ (4)
x1
(a contradiction otherwise). It is now apparent directly (if x1 ≥ 1) or from (3) and (4) (if
x1 < 1) that
f (x1 ) = f ( x11 ) ≤ f (x2 ) ≤ · · · ≤ f (xn ).
n on
This means that the sequence 1+x √ k
xk is non-decreasing. Thus (2) holds according to
k=1  1 n
the well-known Chebyshev’s inequality since the sequence 1+x k k=1
is decreasing.
The equality in (2) holds if and only if

1 1 1 1 + x1 1 + x2 1 + xn
= = ··· = or √ = √ = ··· = √ ,
1 + x1 1 + x2 1 + xn x1 x2 xn

which implies x1 = x2 = · · · = xn . Then we obtain from (1) that x1 = x2 = · · · = xn =

n − 1. 

j12-I-3/j12-II-59-1
Problem j12-I-4/j12-I-5. The numbers 1, 2, . . . , n are assigned to the vertices of a regular
n-gon in an arbitrary order. For each edge compute the product of the two numbers at the
endpoints and sum up these products. What is the smallest possible value of this sum?
(Babeş-Bolyai University, Cluj-Napoca)
Solution. Due to the (a−b)2 = a2 −2ab+b2 identity, it is sufficient to find the maximum
of the sum
n
X
2
σ(k + 1) − σ(k)
k=1

where σ(k) denotes the number from the k th vertex and σ(n + 1) = σ(1). We will give an
inductive algorithm to find an optimal arrangement and so we can find the maximal sum
(or the minimal for the initial problem). Suppose we have an arbitrary arrangement with n
numbers and construct an arrangement with n + 2 numbers in the following way:
  
• Find the maximum
 of σ(k + 1) − σ(k). For such a k, denote x = min σ(k + 1), σ(k)

and y = max σ(k + 1), σ(k) .
• Increase each number by 1.
• Insert the numbers 1 and n + 2 as in figure 1.
If we denote by sn+2 and sn the corresponding distance sums, we have:

2
sn+2 = sn − (x − y)2 + (n + 1) − x + (n + 1)2 + y 2
= sn + 2(n + 1)2 + 2xy − 2x − 2nx.

On the other hand, from the obvious inequalities x ≥ 1 and n + 1 − y ≥ 1, we have

x(n + 1 − y) ≥ 1 and this implies 2xy − 2x − 2nx ≤ −2. Hence

sn+2 = sn + 2(n + 1)2 − 2n = 2n(n + 2).

If yn is the maximal sum, we have yn+2 = yn + 2n(n + 2) (because for

 n = 3 in the maximal

arrangement x = 1, y = 3 and in each step the maximal distance σ(k + 1) − σ(k) occurs
at x = 1 and y = n). For n = 2 and n = 3, we have y2 = 2 and y3 = 6 so from the obtained
recurrence relation we can deduce y2n = 2 + 38 (n − 1)n(n + 1) and thus
Pn
2 k=1 k 2 − y2n 4n3 + 6n2 + 5n − 3
x2n = =
2 3
where xn denotes the minimal sum for the initial problem. Analogously we have

4n3 + 12n2 + 14n + 3

x2n+1 = .
3

For n ∈ {6, 7, 8, 9, 10}, we have illustrated the optimal arrangements on figures 2, 3, 4, 5

(in the exterior we have written the arrangement’s numbers, inside the circle the product of
any two adjacent number and in the inside circle the sum of these products).
Pn Remark. For p >
p 1, the above arrangements will give the maximum of the sum
k=1 σ(k + 1) − σ(k) , and this can be proved by the same method using the inequality

(n + 1)p + (n + 1 − x)p + y p − (y − x)p ≤ np + (n + 1)p + np − (n − 1)p .

j12-I-4/j12-I-5-1
j12-I-4/j12-I-5-2
Problem j12-II-1/j12-II-56. Find all complex solutions of the system
(a + ic)3 + (ia + b)3 + (−b + ic)3 = −6 ,
(a + ic)2 + (ia + b)2 + (−b + ic)2 = 6 ,
(1 + i)a + 2ic = 0 .

(P. J. Šafárik University in Košice)

Solution. Let us notice that the third equation can be written as
(a + ic) + (ia + b) + (−b + ic) = 0;
that is why a natural substitution is
x = a + ic, y = ia + b, z = −b + ic.
Then, our system is
x3 + y 3 + z 3 = −6
x2 + y 2 + z 2 = 6
x +y +z =0

Using symmetric polynomials, we get

x + y + z = σ1 ,
x + y + z = (x + y + z)2 − 2(xy + yz + xz) = σ12 − 2σ2 ,
2 2 2

x3 + y 3 + z 3 = (x + y + z)3 − 3(xy + yz + xz)(x + y + z) + 3xyz = σ13 − 3σ1 σ2 + 3σ3 .

It is a well-known fact that x, y, z must be roots of the cubic polynomial
f (t) = t3 − σ1 t2 + σ2 t − σ3 .
Since σ1 = 0, σ12 − 2σ2 = 6, σ13 − 3σ1 σ2 + 3σ3 = −6, we have
σ1 = 0, σ2 = −3, σ3 = −2.
Rational roots of the polynomial f (t) = t3 − 3t + 2 can only be from the set {−2, −1, 1, 2}.
Trying these, it turns out that t = 1 and t = −2 are roots. Decomposition of the polynomial
then reveals that 1 is a double root.
Thus, we have 
(x, y, z) ∈ (1, 1, −2), (1, −2, 1), (−2, 1, 1) .
Returning back, we solve the system
a + ic = x
ia + b =y
−b + ic = z

Its determinant is  
1 0 i 

|A| =  i 1 0  = i + 1 6= 0,
 0 −1 i
so for each (x, y, z) there is exactly one solution. It is easy to get the inverse matrix:
 
1 + i −1 − i −1 − i
1
A−1 =  1 − i 1 + i −1 + i  .
2
−1 − i 1−i 1−i
Multiplying this matrix by the vectors (x, y, z) gives three solutions (a, b, c):
(1 + i, 2 − i, −1), (1 + i, −1 − i, −1), (−2 − 2i, −1 + 2i, 2).
One can easily verify that all three satisfy the system. 

j12-II-1/j12-II-56-1
Problem j12-II-2/j12-II-52. A ring R (not necessarily commutative) contains at least
one zero divisor and the number of zero divisors is finite. Prove that R is finite.
(Eötvös Loránd University, Budapest)
Solution. Let m be the number of zero divisors and u, v ∈ R two non-zero elements such
that uv = 0.
We generate more zero divisors in the following way. For an arbitrary x ∈ R, the element
xu is either 0 or also a zero divisor, since (xu)v = x(uv) = 0.†
If xu = yu for some different elements x, y ∈ R, then (x − y)u = 0, and x − y is a zero
divisor. This implies that 0 or an arbitrary zero divisor can be obtained at most m + 1 times
in the form xu.‡
Thus, each of 0 and the m zero divisors is obtained at most m times and the number of
elements of R cannot exceed (m + 1)2 . 

† The set {xu; x ∈ R} is finite, its cardinality being ≤ m + 1.

‡ Define an equivalence relation: x ∼ y iff xu = yu. In each class of equivalence, there
are (m + 1) elements at most. Finally, the number of the classes of equivalence is equal to
the cardinality of the set {xu; x ∈ R}, which is finite.

j12-II-2/j12-II-52-1
Problem j12-II-3/j12-II-53. Let E be the set of all continuous functions u: [0, 1] → R
satisfying Z t
u2 (t) ≤ 1 + 4
 
su(s) ds, ∀t ∈ [0, 1].
0

Let ϕ: E → R be defined by
Z 1
u2 (x) − u(x) dx.


ϕ(u) =
0

Prove that ϕ has a maximum value and find it. (Babeş-Bolyai University, Cluj-Napoca)
Solution. Let Z t  
v(t) = 1 + 4 su(s) ds, ∀t ∈ [0, 1].
0

We have s
Z t p
v 0 (t) = 4tu(t) ≤ 4t
   
1+4 su(s) ds ≤ 4t v(t)
0

so
t t
v 0 (s)
p Z Z
v(t) − 1 = p ds ≤ 2s ds = t2
0 2 v(s) 0

therefore  p
u(t) ≤ v(t) ≤ t2 + 1.


If we consider ϕ, we have
 2
u (t) − u(t) = u(t)u(t) − 1 ≤ (t2 + 1)(t2 + 2),
   
Z 1 Z 1
 2 16
(t2 + 1)(t2 + 2) dt =
  
ϕ(u) ≤ u (t) − u(t) dt ≤ .
0 0 5

Equality can be achieved if

u(t) = t2 + 1 u(t) − 1 = t2 + 2.
   
and

This is the case of u(t) = −t2 − 1, which belongs to E. 

j12-II-3/j12-II-53-1
Problem j12-II-4/j12-II-62. Prove that
!
1 √ π2
Z
2 n
lim n 1+ xn dx − 1 = .
n→∞ 0 12

(Sofia University St. Kliment Ohridski)

Solution. We will prove that
∞
!
1 √ (−1)k−1 π2
Z X
2 n
lim n 1+ xn dx − 1 = = .
n→∞ 0 k2 12
k=1

R1 √
P∞ α
k
Let an = n2 0 n 1 + xn dx − 1 . It  αis
kwidely known that (1 + t)α
= k=0

k t  for
any t ∈ [0, 1] and α ∈ (0, 1). Moreover,  k t  = (−1)k−1 αk tk and  αk tk  ≥  αk tk+1  for


k ≥ 1, t ≥ 0 and α ∈ (0, 1). Thus, the following inequalities hold:

2p   2p+1
X 
X α k α α k
t ≤ (1 + t) ≤ t .
k k
k=0 k=0

1
Let us put t = xn and α = n. Integrating on [0, 1], we obtain

2p  1 2p+1
√
 Z X 
X 1/n 1 n 1/n 1
≤ 1+ xn dx ≤ .
k nk + 1 0 k nk + 1
k=0 k=0

Hence,
2p    
X 1/n 1 1/n 1
0 ≤ an − n2 ≤ n2 .
k nk + 1 2p + 1 n(2p + 1) + 1
k=1

A simple calculation gives the following estimation:

 
2 1/n 1 1
n ≤ .
2p + 1 n(2p + 1) + 1 (2p + 1)2

Consequently, as n tends to infinity,

2p
!
X (−1)k 1
0 ≤ lim sup an − ≤
n→∞ k2 (2p + 1)2
k=1

and
2p
!
X (−1)k 1
0 ≤ lim inf an − ≤ .
n→∞ k2 (2p + 1)2
k=1

Letting p → ∞, we obtain the desired result. 

j12-II-4/j12-II-62-1
 The 13th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 2nd April 2003
Category I

Problem 1. Let d(k) denote the number of all natural divisors of a

natural
 number
∞ k. Prove that for any natural number n0 the sequence
d(n2 + 1) n=n0 is not strictly monotone. [10 points]

Problem 2. Let A = (aij ) be an m × n real matrix with Pn at least one

non-zero element. For each i ∈ {1, . . . , m}, let Ri = j=1 aij be the
sum ofPthe i-th row of the matrix A, and for each j ∈ {1, . . . , n}, let
m
Cj = i=1 aij be the sum of the j-th column of the matrix A. Prove
that there exist indices k ∈ {1, . . . , m} and l ∈ {1, . . . , n} such that

akl > 0 , Rk ≥ 0 , Cl ≥ 0 ,
or
akl < 0 , Rk ≤ 0 , Cl ≤ 0 .

[10 points]

Problem 3. Find the limit

s r q √
lim 1+2 1+3 · · · + (n − 1) 1 + n .
n→∞

[10 points]

Problem 4. Let A and B be complex Hermitian 2×2 matrices having

the pairs of eigenvalues (α1 , α2 ) and (β1 , β2 ), respectively. Determine
all possible pairs of eigenvalues (γ1 , γ2 ) of the matrix C = A + B. (We
recall that a matrix A = (aij ) is Hermitian if and only if aij = aji for
all i and j.) [10 points]
 The 13th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 2nd April 2003
Category II

Problem 1. Two real square matrices A and B satisfy the conditions

A2002 = B 2003 = I and AB = BA. Prove that A + B + I is invertible.
(The symbol I denotes the identity matrix.) [10 points]

Problem 2. Let {D1 , D2 , . . . , Dn } be a set of disks in the Euclidean

plane. (A disk is a set of points whose distance from the given centre
is less than or equal to the given radius.) Let aij = S(Di ∩ Dj ) be
the area of Di ∩ Dj . Prove that the inequality
n X
X n
aij xi xj ≥ 0
i=1 j=1

holds for any real numbers x1 , x2 , . . . , xn . [10 points]

Problem 3. Let {an }∞ n=0 be the sequence of real numbers satisfying

a0 = 0, a1 = 1 and
an
an+2 = an+1 + n
2
for every n ≥ 0. Prove that
∞
X 1
lim an = 1 + Qn .
n→∞
n=1
2n(n−1)/2 k=1 (2
k − 1)

[10 points]

Problem 4. Let f, g: [0, 1] → (0, +∞) be two continuous functions

such that f and fg are increasing. Prove that

1
Rx Z 1
f (t) dt
Z
f (t)
R0x dx ≤ 2 dt .
0 0
g(t) dt 0 g(t)

[10 points]
Problem j13-I-1/j13-I-15. Let d(k) be the number of
all natural divisors of a number
∞
k ∈ N. Prove that for any n0 ∈ N the sequence d(n2 + 1) n=n is not strictly monotone.
0
(Vilnius University)
Solution. Note that d(n2 + 1) < n for all even n. Indeed, the number n2 + 1 is not
square and so it is possible to split the set of all its divisors into pairs { d, (n2 + 1)/d } where
d < n and d is odd. The number of divisors in all such pairs does not exceed n.
Let us assume that starting from some n0 ∈ N, the sequence is strictly monotone. For
d(n2 + 1) is always even, we get

d (n + 1)2 + 1 ≥ d n2 + 1 + 2

or, in general,
d (n + k)2 + 1 ≥ d n2 + 1 + 2k

for any natural numbers n ≥ n0 and k ≥ 1. Let N ≥ n0 (e.g., N = n0 ). Taking any

s ≥ N − d(N 2 + 1) (such that N + s is even), we get

d (N + s)2 + 1 ≥ d N 2 + 1 + 2s ≥ N + s,

which is a contradiction with d (N + s)2 + 1 < N + s. 

j13-I-1/j13-I-15-1
Problem j13-I-2/j13-I-19. Let A = [ai,j ] be anPm × n real matrix with at least one
n
non-zero element. For each i ∈ {1, . . . , m} let Ri := j=1 ai,j (the sum of the i-th row of A)
Pm
and for each j ∈ {1, . . . , n} let Cj := i=1 ai,j (the sum of the j-th column of A). Prove
that there exist indices k ∈ {1, . . . , m} and l ∈ {1, . . . , n} such that

ak,l > 0 , Rk ≥ 0 , Cl ≥ 0 ,
or
ak,l < 0 , Rk ≤ 0 , Cl ≤ 0 .

(University of Zagreb)
Solution. Consider the following sets of indices (some of them may be empty):

I + :=

i ∈ {1, . . . , m} | Ri ≥ 0 ,
I−

:= i ∈ {1, . . . , m} | Ri < 0 ,
J+

:= j ∈ {1, . . . , n} | Cj > 0 ,
J−

:= j ∈ {1, . . . , n} | Cj ≤ 0 .

Suppose that the statement of the problem does not hold. Then (but not equivalently) we
+ + − −
have ai,j ≤ 0 for every
P (i, j) ∈ I × J and we have ai,j ≥ 0 for every (i, j) ∈ I × J . Let
us write the sum (i,j)∈I − ×J + ai,j in two different ways:

X n
X X X  X X
ai,j = ai,j − ai,j = Ri − ai,j ≤ 0 ,
(i,j)∈I − ×J + i∈I − j=1 j∈J − i∈I − (i,j)∈I − ×J −
X m
X X X  X X
ai,j = ai,j − ai,j = Cj − ai,j ≥ 0 .
(i,j)∈I − ×J + j∈J + i=1 i∈I + j∈J + (i,j)∈I + ×J +

P
Therefore, ai,j = 0 and we have only equalities in the two formulae above.
(i,j)∈I − ×J +P
This is only possible if i∈I − Ri = 0 and j∈J + Cj = 0, so I − = ∅ and J + = ∅,† which
P
means Ri ≥ 0 for all i = 1, . . . , m and Cj ≤ 0 for all j = 1, . . . , n. Moreover, from
m
X X n
m X n X
X m n
X
0≤ Ri = ai,j = ai,j = Cj ≤ 0,
i=1 i=1 j=1 j=1 i=1 j=1

we conclude Ri = 0 for i = 1, . . . , m and Cj = 0 for j = 1, . . . , n. Since A is a non-zero

matrix, there are indices k and l such that ak,l 6= 0, but Rk = 0 and Cl = 0, which leads to
a contradiction with the assumption that the statement of the problem is false. 

† If I − 6= ∅, then (i,j)∈I − ×J + ai,j ≤ i∈I − Ri < 0 — a contradiction. We can argue

P P

similarly to show J + = ∅.

j13-I-2/j13-I-19-1
Problem j13-I-3/j13-I-9. Find the limit
s r q √
lim 1+2 1+3 · · · + (n − 1) 1 + n .
n→∞

(Dr. Moubinool Omarjee, Paris†)

Solution. Let
s r q √
um,n = 1 + m 1 + (m + 1) · · · + (n − 1) 1 + n .

We have
u2m,n = 1 + mum+1,n ,
u2m,n − (m + 1)2 = m um+1,n − (m + 2) .

Using the equality |a − b| = |a2 − b2 |/|a + b| and inequality um,n + m + 1 ≥ m + 2, we get

 
um,n − m − 1 ≤ m  
um+1,n − (m + 2).
m+2

We deduce that
2 3 n−1
|u2,n − 3| ≤ · · ··· · · |un−1,n − n|,
4 5 n+1
√
q   
6 1
|u2,n − 3| ≤ 1 + (n − 1) 1 + n − n = O .
n(n + 1) n

So we get
lim u2,n = 3.
n→∞

† This problem is formally proposed by the University of Ostrava.

j13-I-3/j13-I-9-1
Problem j13-I-4/j13-I-12. Let A and B be complex hermitian 2 × 2 matrices with pairs
of eigenvalues (α1 , α2 ) and (β1 , β2 ), respectively. Determine all possible pairs (γ1 , γ2 ) of
eigenvalues of the matrix C = A + B. (A matrix A = [ai,j ] is hermitian if and only if
ai,j = aj,i for all i, j.) (Charles University in Prague)
Solution. Recall that all eigenvalues of a hermitian matrix are real numbers and that
there exists an orthonormal basis consisting of eigenvectors of the matrix. As we can add a
suffitiently large multiple of the identity matrix to both matrices A and B, we can suppose
wlog that α1 , α2 , β1 , β2 > 0 and also γ1 , γ2 > 0.
Let us also wlog suppose α1 ≥ α2 , β1 ≥ β2 , γ1 ≥ γ2 and α1 − α2 ≥ β1 − β2 . By easy
arguments, we can see

γ1 + γ2 = Tr C = Tr A + Tr B = α1 + α2 + β1 + β2 .

Further, it holds that

γ1 ≤ α1 + β1 , γ2 ≥ α2 + β2 .
(The first inequality can be seen if we rewrite it slightly: γ1 = kCk ≤ kAk + kBk = α1 + β1 .
The second inequality follows if we consider the equality above and the first inequality
together. — Alternatively, γ1 = max(Cx, x)/(x, x) ≤ max(Ax, x)/(x, x) + max(Bx, x)/
(x, x) = α1 + β1 and γ2 = min(Cx, x)/(x, x) ≥ min(Ax, x)/(x, x) + min(Bx, x)/(x, x) =
α2 + β2 .) Later we will also prove the inequalities

γ1 ≥ α1 + β2 , γ2 ≤ β1 + α2

(in fact, it suffices to prove only the first one because the second one follows if we use the
equality given above).
From these inequalities, we can see that γ1 ∈ [α1 + β2 , α1 + β1 ]. (The value of γ2 has
to be “complementary” to obtain the right value of the sum γ1 + γ2 . It also worths noting
that even if γ1 = α1 + β2 , then still γ1 ≥ γ2 = β1 + α2 . This follows from the assumption
α1 − α2 ≥ β1 − β2 .) We will show that γ1 can assume any value from the given interval
[α1 + β2 , α1 + β1 ]. Consequently, the set of all possible pairs (γ1 , γ2 ) of eigenvalues of the
matrix C = A + B is

(γ1 , γ2 ) : α1 + β2 ≤ γ1 ≤ α1 + β1 , γ1 + γ2 = α1 + α2 + β1 + β2 .

To see this, let us put

     
α1 0 β1 0 cos t sin t
A= , B= , P (t) = .
0 α2 0 β2 − sin t cos t

The matrix A obviously has eigenvalues (α1 , α2 ). The matrix B(t) = P −1 (t)BP (t) obviously
has eigenvalues (β1 , β2 ). If we note that P −1 (t) = P T (t) and define the matrix C(t) =
A + B(t), we have
   
α1 + β1 0 α1 + β2 0
C(0) = A + B = , C( π2 ) = .
0 α2 + β2 0 α2 + β1

The matrix C(0) has the eigenvalue γ1 (0) = α1 + β1 . (Note that γ1 (0) ≥ γ2 (0) = α2 + β2 .)
The matrix C(π/2) has the eigenvalue γ1 (π/2) = α1 + β2 . (Note that γ1 (π/2) ≥ γ2 (π/2) =
α2 + β1 .) As both eigenvalues (γ1 , γ2 ) of a matrix C depend continuously on the coefficients
of the matrix, we deduce that γ1 (t) is a continuous function. Consequently, it assumes every
value from the interval [α1 + β2 , α1 + β1 ], which we wanted to demonstrate.
Now it only remains to prove the inequality γ1 ≥ α1 + β2 for any two complex hermitian
matrices A and B. Let us recall that we still wlog suppose α1 ≥ α2 > 0, β1 ≥ β2 > 0
and γ1 ≥ γ2 > 0. Let v1 and v2 denote the eigenvectors of the matrix A corresponding
to the eigenvalues α1 and α2 , respectively, and let w1 and w2 denote the eigenvectors of B
corresponding to the eigenvalues β1 and β2 , respectively. We can suppose that the bases

j13-I-4/j13-I-12-1
 u11 u12


{v1 , v2 } and {w1 , w2 } are orthonormal. So there exists some unitary matrix U = u21 u22
such that
v1 = u11 w1 + u12 w2 , w1 = u11 v1 + u21 v2 ,
and
v2 = u21 w1 + u22 w2 , w2 = u12 v1 + u22 v2 .
We will estimate γ1 in the following way. First,

γ1 = sup kCxk : kxk = 1 ≥ kCv1 k

where k·k denotes the Euclidean norm. (Let us justify the formula. Recall that γ1 =
maxkxk=1 (Cx, x). Obviously, γ12 is the greater eigenvalue of C 2 . Consequently, it follows
that γ12 = maxkxk=1 (C 2 x, x). As C is hermitian, we have (C 2 x, x) = x∗ CCx = x∗ C ∗ Cx =
(Cx, Cx) = kCxk2 .) Second,

Cv1 = (A + B)v1 = α1 v1 + β1 u11 w1 + β2 u12 w2 = (α1 + β2 )v1 + (β1 − β2 )u11 w1 =

= α1 + β2 + (β1 − β2 )u11 u11 v1 + (β1 − β2 )u11 u21 v2 .

As the vectors v1 and v2 are orthonormal and (β1 − β2 )u11 u11 ≥ 0, we conclude
q 2  2
γ1 ≥ kCv1 k = α1 + β2 + (β1 − β2 )u11 u11  + (β1 − β2 )u11 u21  ≥
q 2
≥ α1 + β2 + (β1 − β2 )u11 u11  ≥ α1 + β2 .

j13-I-4/j13-I-12-2
Problem j13-II-1/j13-II-51. Two real square matrices A and B satisfy the conditions
A2002 = B 2003 = I and AB = BA. Prove that A + B + I is invertible. (The symbol I
denotes the identity matrix.) (University of Belgrade)
Solution. Let (A + B + I)x = 0 for some vector x, i.e., (B + I)x = −Ax. Then
we have −A2 x = A(B + I)x = (B + I)Ax = −(B + I)2 x, and, continuing in this way,
(B + I)k x = (−1)k Ak x. As A2002 = I, we get (B + I)2002 x = x, i.e.,

(B + I)2002 − I x = B 2003 − I x = 0.

(Recall B 2003 = I.) In other words, taking that p(t) = (t + 1)2002 − 1 and q(t) = t2003 − 1
are polynomials, we have just got

p(B)x = q(B)x = 0.

But, since 2003 is a prime, q(t)/(t − 1) is a primitive polynomial for all its roots, and
therefore none of them is a root of the another monic polynomial p(t) of degree 2002; further,
the remained root t = 1 of q(t) is not a root of p(t), which implies that p(t) and q(t) are
coprime.†
Since there exist non-zero polynomials r(t) and s(t) such that r(t)p(t) − s(t)q(t) = 1
(recall the Euclidean algorithm), we can conclude that x = r(B)p(B)x − s(B)q(B)x = 0,
and so A + B + I must be invertible indeed. 

† The polynomials p(t) and q(t) are really coprime (i.e. relatively prime). Here is another
argument: Every polynomial (of degree ≥ 1) can be written as a product of factors of
Q2002
degree 1. In particular, p(t) = (t + 1)2002 − 1 = k=1 (t − zp,k ) and q(t) = t2003 − 1 =
Q2003
k=1 (t − zq,k ), where zp,1 , . . . , zp,2002 and zq,1 , . . . , zq,2003 are the roots of the polynomial
p and q, respectively. Obviously, the polynomials p and q are relatively prime iff they have
no root in common.
It is easy to see that the roots of q lie on the unit circle in the complex plane. Similarly,
it is easy to see that all roots of p are on the circle with radius 1 and its centre at the
point −1.
Thus, the intersections of the two circles,
√ √
2 2
= cos ± 3π 3π π π



2 ±i 2 2 + i sin ± 2 = −1 + cos ± 2 + i sin ± 2 ,

are the only possible common roots of q and p. But none of these two points is a root of q.
It follows that p and q are coprime.

j13-II-1/j13-II-51-1
Problem j13-II-2/j13-I-17. Let {D1 , D2 , . . . , Dn } be a set of disks (a disk is a circle with
its interior) in the Euclidean plane and aij = S(Di ∩ Dj ) be the area of Di ∩ Dj . Prove that
for any numbers x1 , x2 , . . . , xn ∈ R the following inequality holds:
n X
X n
aij xi xj ≥ 0.
i=1 j=1

(Warsaw University)
Solution. Let χDi : R2 → {0, 1} be the characteristic function of the set Di :

1, if (x, y) ∈ Di ,
χDi (x, y) =
0, if (x, y) ∈
/ Di .

We have:
χDi ∩Dj = χDi χDj ,
Z Z
S(Di ) = χDi (x, y) dx dy = χ2Di (x, y) dx dy,
R2 R2
Z Z
S(Di ∩ Dj ) = χDi ∩Dj (x, y) dx dy = χDi (x, y)χDj (x, y) dx dy.
R2 R2

Thus,

X n
n X Z n X
X n
aij xi xj = xi χDi (x, y)xj χDj (x, y) dx dy =
i=1 j=1 R2 i=1 j=1
Z

2
= x1 χD1 (x, y) + · · · + xn χDn (x, y) dx dy ≥ 0.
R2

j13-II-2/j13-I-17-1
Problem j13-II-3/j13-II-70. A sequence (an )∞ n=0 of real numbers is defined recursively
by
an
a0 := 0, a1 := 1, an+2 := an+1 + n , n ≥ 0.
2
Prove that
∞
X 1
lim an = 1 + n(n−1) Qn .
n→∞
n=1 2 2 · k=1 (2k − 1)
(University of Zagreb)
Remark. In fact, we will prove the following:
∞
(a) The sequence (an )P
n=0 is convergent. Q
∞ n

(b) limn→∞ an = 1 + n=1 1/ 2n(n−1)/2 · k=1 (2k − 1) .
(c) The limit limn→∞ an is an irrational number.
Solution. (a) Obviously, an ≥ 0 for every n ≥ 0. The sequence (an )∞ n=0 is increasing
since an+2 − an+1 = an /2n ≥ 0 for every n ≥ 0. It suffices to show that (an )∞
n=0 is bounded
from above. For each n ≥ 0, we have an+2 ≤ an+1 + an+1 /2n = an+1 (1 + 1/2n ). Using the
inequality between geometric and arithmetic mean, for every n ≥ 1 we obtain
n   n    n !n !n
Y 1 Y 1 1 X 1 n+1
an+2 ≤ 1+ k =2 1+ k ≤2 n+ ≤2 ≤ 2e.
2 2 n 2k n
k=0 k=1 k=1
P∞ p √
(b) Consider the power series n=0 an z n . Since lim supn→∞ n |an | ≤ limn→∞ n 2e = 1,
its radius of convergence is R ≥ 1. Therefore, on the P open unit disc, with center at the
∞ n
origin, it converges
Pn to a holomorphic function f (z) := Pa∞n z . Inductively,
n=0 we

obtain
an+2 = 1 + k=0
ak /2 for any n ≥ 0. So limn→∞ an = 1 + k=0 ak /2k = 1 + f 21 and we
k

have to find f 21 .
Now we use the recurrent relation for (an )∞
n=0 to obtain a functional equation for f . We
multiply an+2 := an+1 + an /2n by z n+2 and sum over all n ≥ 0 to get
∞ ∞ ∞
z n
X X X
n+2 n+1 2


an+2 z =z an+1 z +z an 2 ,
n=0 n=0 n=0
that is
f (z) − z = zf (z) + z 2 f z


2 ,
or
(1 − z)f (z) = z 2 f z


2 +z for |z| < 1. (1)

We substitute z = 1/2n for n = 1, . . . , N (where N ≥ 1 is a fixed number) into (1), then

multiply the n-th equality by some constant sn > 0 and finally sum up those N equalities:

2
1 − 21 f 12 = 12 f 14 + 12 ,




| · s1 ,
2
1 − 14 f 14 = 14 f 18 + 14 ,





| · s2 ,
..
.

2
1 − 2n f 2n = 21n f 2n+1
1

1
1
+ 21n ,


| · sn ,
1

1

1

2 1

1
1 − 2n+1 f 2n+1 = 2n+1 f 2n+2 + 2n+1 , | · sn+1 ,
..
.

2
1 − 2N f 2N = 21N f 2N1+1 + 21N ,
1

1


| · sN ,
s1 1

sN 1

XN sn
f 2 = f 2N +1
+ .
2 22N n=1 2n
1


To obtain the given result (namely, to achieve cancelling of the terms with f 2n for n =
2, . . . , N ), we had to choose the numbers sn so that
1

2
sn+1 = 21n sn ,


1 − 2n+1 for n ≥ 0. (2a)

j13-II-3/j13-II-70-1
Let us put
s0 := 1. (2b)
It follows that s1 = 2. Equalities (2b) and (2a) lead to

n−1 n−1 1 2 n−1

Y sk+1 Y
2k
Y 1 1
sn = = 1 = k−1 k+1
= n(n−1) Qn
sk 1 − 2k+1 2 (2 − 1) −n k
k=0 k=0 k=0 2 2 k=1 (2 − 1)

for every n ≥ 1. Finally, we have

N N
f 2N1+1


1

sN 1

X sn X 1
f 2 = 2N f 2N +1
+ n
= N (N −1) QN + n(n−1) Qn
.
2 n=1
2 2 2 +N (2k − 1) n=1 2 2 (2k − 1)
k=1 k=1

The first term tends to 0 when N → ∞, so

∞
1

X 1
f 2 = n(n−1) Qn . (3)
k
2 k=1 (2 − 1)
2
n=1

(c) The proof of limn→∞ an ∈ R \ Q is based on the fact that the series in (3) converges
“very rapidly”. Suppose that its sum equals pq for some positive integers p and q. For each
integer N ≥ 1, denote

N N
N (N −1) Y X 1
qN := 2 2 (2k − 1), pN := qN n(n−1) Qn .
2 k − 1)
k=1 (2
2
k=1 n=1

Obviously, pN and qN are positive integers. We manage to estimate pqN − qpN . We have

N N
N (N −1) Y N (N −1) Y 2
qN = 2 2 (2k − 1) < 2 2 2k = 2N
k=1 k=1

and
∞ ∞
p pN X 1 X 1
− = n(n−1) n
≤ n(n−1) Qn
=
q qN Q k k−1
n=N +1 2 k=1 (2 − 1) n=N +1 2 k=1 2
2 2

∞ ∞
X 1 X 1 1 1
= ≤ = N 2 +N −1 < N −1 .
2n(n−1) 2m 2 2 qN
n=N +1 m=N (N +1)
q
Thus, 0 < pqN − qpN < 2N −1 ,† so (pqN − qpN )N ≥1 is a sequence of positive integers that
converges to 0. This is a contradiction and we are done. 

† It is easy to see from the definition of the numbers pN that the sequence pqN


N
is strictly
p pN p
increasing to the limit q . Hence qN < q , qpn < pqN , and 0 < pqN − qpN . As the difference
is integer, we have even 1 ≤ pqN − qpN .

j13-II-3/j13-II-70-2
Problem j13-II-4/j13-I-18. Let f, g: [0, 1] → (0, +∞) be continuous functions such that
f and fg are increasing. Prove that

1
Rx Z 1
f (t) dt
Z
0 f (t)
Rx dx ≤ 2 dt.
0 0
g(t) dt 0 g(t)

(University of Zagreb)
Solution. First, we estimate the expression inside the integral sign on the left side of the
given inequality. By the Chebycheff’s inequality for integrals applied to increasing functions
f and fg on the segment [0, x] (where x ∈ (0, 1] is fixed), we get
 Z x  Z x
1 x
 Z
1 1 g(t)
f (t) dt dt ≤ g(t) dt,
x 0 x 0 f (t) x 0
that is,
Rx
f (t) dt x
R0x ≤ R x g(t) (1)
0
g(t) dt dt
0 f (t)

for every x ∈ (0, 1]. From the integral form of the Cauchy-Schwarz inequality on the segment
[0, x], we have
x x 2 x 2
x4
Z Z  Z
g(t) t f (t)
dt dt ≥ t dt = ,
0 f (t) 0 g(t) 0 4
or
Z x 2
1 4 t f (t)
Rx ≤ dt. (2)
g(t)
dt x4 0 g(t)
0 f (t)

From (1) and (2) we obtain

Rx Z x 2
f (t) dt 4 t f (t)
R0x ≤ 3 dt. (3)
0
g(t) dt x 0 g(t)

Finally, it remains to integrate (3) over x ∈ (0, 1] and to reverse the order of integration.
Z 1 Rx Z 1 Z x 2  Z 1 Z 1 2 
0
f (t) dt 4t f (t) 4t f (t)
Rx dx ≤ 3
dt dx = dx dt =
0 0
g(t) dt 0 0 x g(t) 0 t x3 g(t)
Z 1 2 Z 1  Z 1 2  
4t f (t) dx 4t f (t) 1 1
= 3
dt = − dt =
0 g(t) t x 0 g(t) 2t2 2
Z 1 Z 1
f (t) 2

f (t)
=2 1 − t dt ≤ 2 dt.
0 g(t) 0 g(t)

(Remark. The constant 2 on the right hand side of the given inequality is optimal, i.e.,
the least possible. Consider f (t) := 1 and g(t) := t + ε for some fixed ε > 0. Then
Z 1 Rx Z 1 Z 1
f (t) dt x dx
R0x dx = 1 2 dx = 2 = 2 ln(1 + 2ε) − 2 ln 2 − 2 ln ε
0 0
g(t) dt 0 2 x + εx 0 x + 2ε

and Z 1 Z 1
f (t) dt
dt = = ln(1 + ε) − ln ε.
0 g(t) 0 t+ε
The quotient of these two expressions can be made arbitrarily close to 2 since

2 ln(1 + 2ε) − 2 ln 2 − 2 ln ε − ln(1+2ε)

ln ε + ln 2
ln ε + 1
lim = 2 lim = 2.
ε&0 ln(1 + ε) − ln ε ε&0 − ln(1+ε)
ln ε +1

Therefore, the constant 2 is the best possible one.) 

j13-II-4/j13-I-18-1
 The 14th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2004
Category I

Problem √ 1 Suppose that f : [0, 1] → R is a continuously differentiable function such that f (0) = f (1) = 0 and
f (a) = 3 for some a ∈ (0, 1). Prove that there exist two tangents to the graph of f that form an equilateral
triangle with an appropriate segment of the x-axis.
Solution Let k : [0, 1] → R be defined as k(x) = f 0 (x). We know that k is continuous. √ Let b ∈ (0,√1) be the
0
point where f reaches √ its maximum value. Then k(b) = f (b) = 0 and f (b) ≥ √ 3 > 0√(as f (a) = 3). Since
0
f (0) = 0 and f (b) ≥ 3, there exists an x0 ∈ [0, b) such √ that f (x 0 ) = k(x 0 ) ≥ 3/b > 3. As k is continuous,
we can find an y1 ∈ (x0 , b) such
√ that k(y1 ) = f 0 (y1 ) = 3. Using the same argument, there exists an y2 ∈ (b, 1]
such that k(y2 ) = f 0 (y2 ) = − 3. It is easy to see that the tangents to the graph of f at y1 and y2 with the
appropriate segment of the x-axis form an equilateral triangle. 
 The 14th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2004
Category I

Problem 2 Evaluate the sum

∞  
X 1
arctan .
n=0
1 + n + n2
Solution Using the formula for the diference of two arcustangents
u−v
arctan u − arctan v = arctan ,
1 + uv
(this formula is valid whenever uv 6= −1) we obtain
∞   ∞  
X 1 X (n + 1) − n
arctan = arctan(1) + arctan =
n=0
1 + n + n2 n=1
1 + n(n + 1)
∞  
π X (n + 1) − n
= + arctan =
4 n=1 1 + n(n + 1)
∞    
π X 1 1 π
= + arctan − arctan = .
4 n=1 n n+1 2


 The 14th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2004
Category I

Problem 3 Denote by B(c, r) the open disc in the plane of center c and radius r. Prove or disprove that there
exists a sequence {zn }∞ 2
n=1 in R such that the open discs B(zn , 1/n) are pairwise disjoint and the sequence
∞
{zn }n=1 is convergent.
Solution Both statemets are true, it is sufficient to prove (b) (in general version, it is obviously not true in R1
due to measure argument, and in Rk , if k ≥ 2, it is sufficient to prove the positive answer for k = 2). We will
prove the crucial Lemma:
Lemma Let a > 0 and let C = (0, a)2 . Then there exists an na ∈ N and a sequence {tn }n≥na ⊂ C such that
the open balls B(tn , 1/n) ⊂ C for n ≥ na are mutually disjoint.
Proof Consider the numbers n ∈ N and c > 0 which will be specified concretely later and suppose that n > 4
and n(c − 1) > 1. Let us denote nk = bnck c. We have n0 = n and as n(c − 1) > 1, we also have that the
sequence {nk } is strictly increasing. For every k ≥ 0, we denote
2 2
sk = + ··· + .
nk nk+1 − 1
By easy integral estimate, we have

nck+1 c
sk ≤ 2 ln(nk+1 − 1) − ln(nk − 1) ≤ 2 ln k = 2 ln =: d(c, n) .
nc − 2 1 − 2/n
Now we easilly see that for every k ≥ 0, we can find mutually disjoint open balls of diameters 1/nk , . . . ,
1/(nk+1 − 1) inside the rectangle of width d(c, n) and height 2/nk . Since
∞ ∞ ∞
X 2 X 1 X 1 4c
=2 k
≤ 4 k
= ,
nk bnc c nc n(c − 1)
k=0 k=0 k=0

it is clear that we can find mutually disjoint

open balls of diameters 1/n, 1/(n + 1), . . . inside the rectangle of
width d(c, n) and height 4c n(c − 1) .1 For the given a, now it is sufficient to set c > 1 so that 2 ln c < a2 and
consequently set na = n so that n > 4, n(c − 1) > 1 and
 
4c
max d(c, n), < a.
n(c − 1)
The lemma is proved. 

Now by means this Lemma, we can construct the sequence zn easilly. Let ai = 2−i and let Ci = (ai , 2ai )2 .
By the lemma, we can find an increasing sequence of the numbers ni = nai such that we can place open balls of
the diameters 1/n0 , 1/(n0 + 1), . . . , 1/(n1 − 1) inside C0 in such a way that they are mutually disjoint. Then
we place open balls of the diameters 1/n1 , . . . , 1/(n2 − 1) inside C1 so that they are mutually disjoint, etc.
The first n0 − 1 balls can be placed anywhere else. It is clear that the sequence of the centres of thse balls is
convergent (having the limit 0). 
1 1 1
Solution Take squares with lenghts of sides 4, 2, 1, 2 , . . . and make a row of them. Put discs with radii 1, 2 , 3
to first square. It is possible since it can be divided into 4 smaller squares with side 1. Than put disc with radii
1 1
4 , . . . , 15 to second square. Thanks to dividing to 16 smaller squares it is again possible. Continuing by this
way we put all discs into sequence of convergent squares, therefore also sequence of discs is convergent.

radii 1, 12 , 31

radii 14 , . . . , 15
1

1 1
16 , . . . , 63

| {z }| {z }| {z }
4 2 1
1 Recall that n = n0 . Place the balls of diameters 1/n0 , . . . , 1/(n1 − 1) into a rectangle of width d(c, n) and height 2/n0 . Place
the balls of diameters 1/n1 , . . . , 1/(n2 − 1) into a rectangle of width d(c, n) and height 2/n1 . Etc.
2 In order that it is possible to ensure d(c, n) < a by the choice of large n.
 
1
Solution Take rectangle 4 × 2 and divide it into rectangles 2 × 2, 2 × 1, 1 × 1, 1 × 2,
. . . . To first rectangle
put disc with radius 1, to second rectangle discs with radii 12 and 13 , to third one discs with radii 14 , . . . , 17 , and
so on. Sequence of rectangles is again convergent and therefore also sequence of discs.

...
...


√
bP nc n
6 3
π2
P
Solution Let zn = (xn , yn ), where xn = i2 and yn = √ 2 i
. From real analysis we know lim xn =
i=1 n→∞
i=b nc
∞ 2
and by small counting lim yn = 0, so {zn }n=1 converges to (π , 0).
n→∞
1
Let’s prove that Bn (zn , n1 ) and Bm (zm , m ) are disjont for all m 6= n. The discs are grouped in columns. In
each column are discs from Bn2 to Bn2 +2n , so each column of discs is included in one of zones
 Xn n 
6 1 X 6  1
An = (x, y) : x ∈ − , + .
i=1
i2 n2 i=1 i2 n2

These An are pairwise disjont, because

n n+1
X 6 1 X 6 1
2
+ 2
< 2
−
i=1
i n i=1
i (n + 1)2

so all of the columns are pairwise disjoint. Discs in each column are pairwise disjoint (if Bn and Bn+1 are in
3
the same column, then |yn+1 − yn | ≥ n+1 ≥ n1 + n+1
1
). Result follows: there exists desired sequence. 
 The 14th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2004
Category I

Problem 4 Find all pairs (m, n) of positive integers such that m + n and mn + 1 are both powers of 2.
Solution The set of such pairs is

L = (2t − 1, 2t + 1), (2t + 1, 2t − 1), (2t − 1, 1), (1, 2t − 1)



where t runs through the set of all positive integers.

For the proof, let m, n, k, l be integers such that m, n ≥ 1 and

mn + 1 = 2l ,
m + n = 2k .

From m, n ≥ 1 we get 2l = mn + 1 ≥ 2, so l ≥ 1 and 2l is even. Hence m and n are both odd. Assume
m = n. Then m = n = 2k−1 must be odd, hence k = 1, which implies m = n = 1, and (1, 1) ∈ L. From now
on, we assume m < n, so y = 12 (n − m) is a positive integer.
If y = 1, then we get n − m = 2, m + n = 2k , so n = 2k−1 + 1 and m = 2k−1 − 1, which yields a pair in the
set L (if k > 1 as m is not positive otherwise).
Let y > 1. Then
 n + m 2
0 < y2 − 1 = − mn − 1 = 22k−2 − 2l .
2
This implies 2k − 2 > l, and hence y 2 − 1 = (y − 1)(y + 1) is divisible by 2l . The greatest common divisor of
y − 1 and y + 1 is 2, so one of these positive factors must be divisible by 2l−1 . In both cases, we get y + 1 ≥ 2l−1 .
Hence
22k−2 − 2l = (y + 1)(y − 1) ≥ 2l−1 (2l−1 − 2) = 22l−2 − 2l ,
so 2k − 2 ≥ 2l − 2, which implies k ≥ l. But then

0 ≤ (m − 1)(n − 1) = (mn + 1) − (m + n) = 2l − 2k ≤ 0 ,

so we must have equality. Thus l = k and (since m < n) m = 1. From this we get n = 2k − 1, so (m, n) ∈ L. 
Solution Let us write m+n = 2a and mn+1 = 2b . Of course a, b ≥ 1 and b ≥ a since 2b −2a = (m−1)(n−1) ≥ 0.
If b = a, then (m − 1)(n − 1) = 0, so

m = 1 and n = 2a − 1 or m = 2a − 1 and n = 1.

So, suppose b > a. If a = 1, then m = n = 1. If a = 2, then {m, n} = {1, 3}. So, let also a ≥ 3. It is obvious,
that m and n are odd (since mn + 1 is even). If m ≡ n ≡ 3 (mod 4), then 0 ≡ 2a ≡ m + n ≡ 2 (mod 4),
contradiction. So thanks to symetry we can assume that m ≡ 1 (mod 4). Since m + n ≡ 0 (mod 4) we must
have n ≡ 3 (mod 4). So, let m = 4c + 1 and n = 4d + 3, where c, d ∈ {0, 1, . . . }. We have

(m + 1)(n + 1) = 2a + 2b = 2a (2b−a + 1)

thus
8(2c + 1)(d + 1) = 2a (2b−a + 1) .
Since 2c + 1 is odd, we have 2a−3 | d + 1. So let d + 1 = x · 2a−3 , with x ∈ {1, 2, . . . }. Thus, n = 4(d + 1) − 1 =
2a
x2a−1 − 1 and so m + x2a−1 − 1 = 2a . Since m ≥ 1, we have x ≤ 2a−1 = 1 and so x = 1 or x = 2. If x = 1,
a−1 a−1
then m = 2 + 1 and n = 2 − 1. If x = 2, then m = 1 and n = 2a − 1. So, we have proved that it is

either {m, n} = {1, 2a − 1} or {m, n} = {2a + 1, 2a − 1} , where a ∈ N.


 The 14th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2004
Category II

Problem 1 Are the groups (Q, +) and (Q+ , ·) isomorphic? (The symbol Q+ denotes the set of all positive
rational numbers.)
Solution Assume they are, and name the isomorphism ϕ. Then there exists an a ∈ Q such that ϕ(a) = 2. Let
ϕ a2 = b. Then


a a a a
2 = ϕ(a) = ϕ + =ϕ ·ϕ .
2 2 2 2

√
Therefore, b = ϕ a2 = 2, a contradiction. 
Solution Suppose (for contradiction) that there exists an isomorfism f : (Q, +) → (Q+ , ·). Then f (0) = 1,
because 0 and 1 are corresponding neutral elements
of agroups (Q, +) and (Q+ , ·). Let f (1) = ab , a, b ∈ N,
1
(a, b) = 1. Similarly, let for every n ∈ N, n ≥ 2, f n = bnn , where an , bn ∈ N, (an , bn ) = 1. Then it is
1 1 1    1 n  an n a
f (1) = f + + ··· + = f = = .
n n n n bn b


It is still (an )n , (bn )n = 1, so consequently a = (an )n and b = (bn )n . If there is m ∈ N such that am ≥ 2,
then a ≥ 2 and an ≥ 2 for all n. That follows a ≥ 2n for all n, contradiction. Thus for every n ∈ N it holds
an = 1, specially a = 1. Same argument gives b = 1. But then f (1) = ab = 1 = f (0), f is not injective,
contradiction. 
Solution Suppose there exists an isomorphism σ : (Q, +) → (Q+ , ·). Consider homomorphism τ : (Q, +) →
(Q, +) given by equation τ (q) = 2q for q ∈ Q, and corresponding homomorphism τ 0 : (Q+ , ·) → (Q+ , ·) given by
equation τ 0 (q) = q 2 for q ∈ Q+ . It follows that

2
σ ◦ τ (q) = σ(2q) = σ(q)

2
τ 0 ◦ σ(q) = τ 0 σ(q) = σ(q)

for all q ∈ Q. It is clear that homomorphism τ is isomorphism, so also τ ◦ σ = τ 0 ◦ σ is isomorphism, but this
contradicts the fact, that τ 0 is not injective (that is obvious) and therefore τ 0 ◦ σ cannot be isomorphism. 
 The 14th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2004
Category II

Problem 2 Find all functions f : R+ + +

0 × R0 → R0 such that

1. f (x, 0) = f (0, x) = x for all x ∈ R+

0,

2. f (f (x, y), z) = f (x, f (y, z)) for all x, y, z ∈ R+

0 and

3. there exists a real k such that f (x + y, x + z) = kx + f (y, z) for all x, y, z ∈ R+

0.

(The symbol R+
0 denotes the set of all non-negative real numbers.)
Solution Take x ≤ y. Then
f (x, y) = f (x + 0, x + (y − x)) = kx + f (0, y − x) = kx + (y − x) = (k − 1)x + y =
= f (x + (y − x), x + 0) = f (y, x) . (1)
Let x ≤ y. If z ≥ y + (k − 1)x, then
f (f (x, y), z) = f (y + (k − 1)x, z) = (k − 1)(y + (k − 1)x) + z .
Further, if z ≥ y and z + (k − 1)y ≥ x, then
f (x, f (y, z)) = f (x, (k − 1)y + z) = (k − 1)x + (k − 1)y + z .
If we take z so large that all the above inequalities are satisfied, then we have
(k − 1)(y + (k − 1)x) + z = (k − 1)x + (k − 1)y + z ,
or (k − 1)2 x = (k − 1)x, and the only possibilities are k = 1, 2. Substituting into (1), we obtain f (x, y) = x + y
for k = 2, and f (x, y) = max{x, y} for k = 1 (we may use (1) only if y ≥ x, whence the maximum). It is easy
to check, that both functions satisfy all conditions. 
Solution Consider two cases x ≤ y and x ≥ y:


x ≤ y : f (x, y) = f x, x + (y − x) = kx + f (0, y − x)
= kx + y − x = (k − 1)x + y


x ≥ y: f (x, y) = f y + (x − y), y = ky + f (x − y, 0)
= ky + x − y = (k − 1)y + x
So it is
f (x, y) = (k − 1) min{x, y} + max{x, y}.


Because f (x,
x) = kx ≥ 0, we have k ≥ 0. Consider now the second equality from the problem: f f (x, y), z =
f x, f (y, z) .
f f (1, 1), k = f (k, k) = k 2



f (1, 2k − 1) = (k − 1) min{1, 2k − 1} + max{1, 2k − 1}



 = (k − 1) + (2k − 1) for k ≥ 1
f 1, f (1, k) = 2 2 2
f (1, k − k + 1) = (k − 1) min{1, k − k + 1} + max {1, k − k + 1}


= (k − 1)(k 2 − k + 1) + 1 for k ≤ 1


Consider cases k ≥ 1 and k ≤ 1 separatelly.

k ≥ 1: k 2 = (k − 1) + (2k − 1)
k 2 − 3k + 2 = 0
(k − 2)(k − 1) = 0
k = 1, 2 or some of them
k ≤ 1: k 2 = (k − 1)(k 2 − k + 1) + 1
k 3 − 3k 2 + 2k = 0
k(k − 2)(k − 1) = 0
k = 0, 1, 2 or some of them
For k = 0, 1, 2 the functions have the following form.

k = 2: f (x, y) = (k − 1) min{x, y} + max{x, y} = x + y

k = 1: f (x, y) = max{x, y}
k = 0: f (x, y) = |x − y|

First two functions satisfy to all three conditions of the problem (easy to check), the last one does not fulfil the
second condition for triple (x, y, z) = (1, 2, 3). 
 The 14th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2004
Category II

∞
P
Problem 3 Let an be a divergent series with positive nonincreasing terms. Prove that the series
n=1
∞
X an
n=1
1 + nan

diverges.
Solution Let (dn ) be a sequence defined as follows:

1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, . . .

(a block of 2n ones is followed by a block of 2n zeros, and that block of 2n zeros is followed by a block of 2n+1
ones and so forth).
Define sets

D = { n ∈ N : dn = 1 } ,
C = 2D ,
E = 2N \ C ,
B = C ∪ (E + 1) ,
A = C ∪ B1 = 2N

and define sequences xn = χA (n) and yn = χB (n) where χS denotes the indicator function of a set S.
Then
x0 + x1 + · · · + xn 1 y0 + y1 + · · · + yn 1
lim = and lim =
n→∞ n+1 2 n→∞ n+1 2
and the sequence
x0 y0 + x1 y1 + · · · + xn yn
lim
n→∞ n+1
1
oscillates between 4 and 38 . 
 The 14th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2004
Category II

Problem 4 Let f : R → R be an infinitely differentiable function. Assume that for every x ∈ R there is an
n ∈ N (depending on x) such that
f (n) (x) = 0 .
Prove that f is a polynomial.
Solution Denote
x ∈ R ; f (n) (x) = 0 .

En :=
Note that each of the sets En is closed due to continuity of f (n) . The assumption implies
[
R= En .
n∈N

By Baire’s theorem, there is an interval I such that some En is dense in I.3 Since f (n) is a continuous function,
we obtain
f (n) (x) = 0 for all x ∈ I.
Therefore, the function f is a polynomial on the interval I. Now, we denote {Iλ }λ∈Λ the set of all maximal
open intervals Iλ such that f is a polynomial on Iλ . (Here, Λ is an index set.) We have already proved at least
one such interval Iλ exists. Moreover, all the sets Iλ are clearly mutually disjoint and their union
S is dense in R
(otherwise, we can repeat the previous argument using Baire’s theorem in the interval J ⊆ R \ λ∈Λ Iλ ).
Now, let us consider the set [
H := R \ Iλ .
λ∈Λ

First, we prove that this set has no isolated points. If there were some isolated point x ∈ H, then there would
be two intervals I1 , I2 ∈ {Iλ }λ∈Λ , one of them having x as the right end-point and the other one having x as the
left end-point. There would also be some n ∈ N such that the n-th derivate vanishes on the union I1 ∪ I2 . This
would mean that the function f is a polynomial on I1 ∪ {x} ∪ I2 and the intervals I1 and I2 are not maximal.
The set H is a closed subset of the complete space R. Therefore, if it is not empty, it is of second category.
Using Baire’s theorem again,4 we prove that there exists an interval J containing at least one point from H
and an index n ∈ N such that
f (n) (x) = 0 for all x ∈ J ∩ H.
Since every x ∈ J ∩ H is an accumulation point of H (and
is an accumulation point of J ∩ H therefore, and
since f (n+1) (x), i.e. the limit lim f (n) (x + h) − f (n) (x) /h does exist at every point x), we can calculate the
h→0
(n + 1)-st derivative of f on J ∩ H using just the points from the intersection J ∩ H. We obtain

f (n+1) (x) = 0 for all x ∈ J ∩ H

and, repeating this argument,

f (m) (x) = 0 for all x ∈ J ∩ H and m ≥ n.

Now, take any interval I ∈ {Iλ }λ∈Λ such that I ⊆ J.5 There is some index m ∈ N such that f (m) (x) = 0
for x ∈ I. Let us assume m > n. There are two cases: either f (m−1) vanishes (is zero) on the interval I or
f (m−1) is a linear function on I. Since the end-points of I belong to H, if follows that f (m−1) is null at these
end-points. Therefore, f (m−1) vanishes on the interval I. By induction (repeating this argument), we prove
that f (n) vanishes on the interval I in fact. This conclusion (i.e., f (n) ≡ 0 on I) is true for every interval
I ∈ {Iλ }λ∈Λ .
Choose an x ∈ J ∩ H. As we already know that J ∩ H cannot contain an interval, there must exist two
intervals Iλ1 , Iλ2 ∈ {Iλ }λ∈Λ such that x is the right end-point of Iλ1 and the left end-point of Iλ2 . Since
3 As
S
the sets En are closed and n En = R, one of the sets must have non-empty interior. Equivalently, there must exist an
index n and an interval I such that I ⊆ En .
4 We just repeat the above argument: As the sets E are closed and
S
n n En ⊇ H, one of the sets En must have non-empty
interior in H. That is, there exists an interval J whose intersection with H is non-empty such that J ∩ H ⊆ En .
5 If there were no such an interval I, it would mean that the set J ∩ H (hence the set H itself) contains a (closed) interval. But

we have already inferred f (n+1) (x) = 0 for x ∈ J ∩ H. Hence, f is a polynomial on that (closed) interval. That interval should
have been included in the collection {Iλ }λ∈Λ therefore.
f (n) is continuous and is zero on Iλ1 ∪ Iλ2 , it is zero at x as well. Hence, f is a polynomial on the interval
Iλ1 ∪ {x} ∪ Iλ2 ∈ {Iλ }λ∈Λ , which is a contradiction (all intervals from {Iλ }λ∈Λ must be mutually disjoint). This
contradiction can be inferred if H 6= ∅. Consequently, the set H is empty and f is a polynomial on the whole
real line. 
 The 15th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 2005
Category I

Im z
Problem 1 Let S0 = {z ∈ C : |z| = 1, z 6= −1} and f (z) = 1+Re z . Prove that f is a bijection between S0 and
R. Find f −1 .
Solution Using z = eit = cos(t) + i sin(t), we can interpret our function as

sin(t)
f (t) = , t ∈ (−π, π).
1 + cos(t)

We can find easily (using L’Hospital or trigonometrical identities) lim f (π− ) = ∞ and lim f (π+ ) = −∞, and by
the continuity the surjectivity follows. The injectivity can be deduced by
1
f 0 (t) = > 0.
1 + cos(t)

Since f (0) = 0,qf maps (0, π) to R+ and (−π, 0) to R− . For t > 0, sin t > 0 and y = f (t) > 0. Then
1−y 2
0 < y = f (t) = 1−cos t 2
1+cos t . We get cos t = 1+y 2 and finally t = arccos(−1 + 1+y 2 ), and similarly for t < 0 and
2
y < 0: t = − arccos(−1 + 1+y 2 ). 
 The 15th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 2005
Category I

Problem 2 Let f : A3 → A, where A is a nonempty set and f satisfies:

1. for all x, y ∈ A, f (x, y, y) = f (y, y, x) = x and
2. for all x1 , x2 , x3 , y1 , y2 , y3 , z1 , z2 , z3 ∈ A,


f f (x1 , x2 , x3 ), f (y1 , y2 , y3 ), f (z1 , z2 , z3 ) =


= f f (x1 , y1 , z1 ), f (x2 , y2 , z2 ), f (x3 , y3 , z3 ) .

Prove that for an arbitrary, fixed a ∈ A, the operation x + y = f (x, a, y) is an Abelian group addition.
Solution Neutral element:
a + x = f (a, a, x) = x = f (x, a, a) = x + a.
Associativity:

(x + y) + z = f (f (x, a, y), a, z) = f (f (x, a, y), f (a, a, a), f (a, a, z)) =

= f (f (x, a, a), f (a, a, a), f (y, a, z)) = f (x, a, f (y, a, z)) = x + (y + z).

Inverse element: Define −x to be f (a, x, a). Now,

x + (−x) = f (x, a, f (a, x, a)) = f (f (x, a, a), f (a, a, a), f (a, x, a)) =
= f (f (x, a, a), f (a, a, x), f (a, a, a)) = f (x, x, a) = a.

Commutativity:

x + y = f (x, a, y) = f (f (a, a, x), f (a, a, a), f (y, a, a)) =

= f (f (a, a, y), f (a, a, a), f (a, a, x)) = f (y, a, x) = y + x.


 The 15th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 2005
Category I

Problem 3 Find all reals λ for which there exists a non-zero polynomial P with real coefficients such that

P (1) + P (3) + P (5) + · · · + P (2n − 1)

= λP (n)
n
for all positive integers n. Find all such polynomials for λ = 2.
Solution Let P be a polynomial satisfying the given equation. Then, for all positive integers n,


P (2n + 1) = λ (n + 1)P (n + 1) − nP (n)


Hence P (2x + 1) − λ (x + 1)P (x + 1) − xP (x) is a polynomial with infinitely many zeros; therefore it must be
the zero polynomial. Conversely, if P is a polynomial satisfying


P (2x + 1) − λ (x + 1)P (x + 1) − xP (x) = 0, (1)

we get (by putting x = 0) P (1) = λP (1) and then by induction

P (1) + P (3) + · · · + P (2n − 1) = λnP (n)

for all positive integers n. We can therefore equally well consider (1).
The set of all real λ for which there is a non-trivial solution of (1) is given by the set
n 2k o n 4 o
= 1, , 2, . . .
k+1 k 3
where k runs through all non-negative integers. To prove this, let the pair (P ,λ) be a solution for (1), and
write P (x) = axk + Q(x) where a 6= 0 and Q is a polynomial of degree less than k. Then the polynomial
S(x) = (x + 1)Q(x + 1) − xQ(x) is of degree less than k as well, and hence

(x + 1)P (x + 1) − xP (x) = a (x + 1)k+1 − xk+1 + S(x)

= a(k + 1)xk + T (x)

for some polynomial T of degree less than k. On the other hand, the degree-k-coefficient of P (2x + 1) is 2k a,
and therefore we conclude from (1)
2k a = λa(k + 1)
k k
2 2
which is the same as λ = k+1 . Conversely, let λ be of the form k+1 for some non-negative integer k, and write Pn
for the (n + 1)-dimensional vector space of all real polynomials of degree at most n. By the computations above,
we have a linear map

Pk −→ Pk−1


P 7→ P (2x + 1) − λ (x + 1)P (x + 1) − xP (x)

Since the dimension of Pk−1 is less than the dimension of Pk , this map must have non-trivial kernel. Hence
there is some non-zero polynomial P satisfying (1).
2k
Now let λ = 2. If k is the degree of P , we have seen that k+1 = λ = 2, but this is only possible for k = 3.
Putting x = 1 in (1) we get P (1) = 2P (1), so P (1) = 0. For x = −1 we get


P (−1) − 2 −P (−1) = 0,

which is P (−1) = 0. Hence P (x) = (x − 1)(x + 1)(ax + b) for some reals a, b. From P (1) + P (3) = 4P (2) we
get P (3) = 4P (2), hence
2 · 4 · (3a + b) = 4 · 3 · (2a + b),
that is, b = 0. Thus we got P (x) = a(x3 − x). The set of polynomials we are looking for form a non-trivial
vector space contained in the one-dimensional vector space spanned by x3 − x. Therefore these are exactly the
solutions of (1) for λ = 2, and we are done. 
 The 15th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 2005
Category I

√
Problem 4 Let (xn )n≥2 be a sequence of real numbers such that x2 > 0 and xn+1 = −1 + n
1 + nxn for n ≥ 2.
Find
1. lim xn ;
n→∞

2. lim nxn .
n→∞

Solution
1. It follows by induction that the sequence (xn )n≥2 is correctly defined and that xn > 0 for n ≥ 2. From
the Bernoulli inequality follows

1 + nxn = (1 + xn+1 )n ≥ 1 + nxn+1 ,

i.e. this sequence decrease (i.e. it is convergent). If lim xn = x ≥ 0, then

n→∞

n(n − 1) 2
1 + nxn = (1 + xn+1 )n ≥ (1 + x)n ≥ 1 + nx + ·x ,
2
2(xn −x)
i.e. 0 ≤ x2 ≤ n−1 . Because lim (xn − x) = 0 and lim (n − 1) = ∞, we get x = 0, i.e. lim xn = 0.
n→∞ n→∞ n→∞
 
1 1
2. From the Stolz theorem, because xn increase and lim = ∞ (after 1), we get
n≥2 n→∞ xn

1 xn xn+1
lim sup nxn ≤ lim sup 1 1 = lim sup .
n→∞ n→∞ xn − xn+1 n→∞ xn − xn+1

It follows
n(n − 1) 2
1 + nxn = (1 + xn+1 )n ≥ 1 + nxn+1 + · xn+1 ,
2
n−1 nxn
i.e. xn − xn+1 ≥ 2 · x2n+1 and lim sup nxn ≤ lim sup n(n−1) . From the equality ln(1 + xn+1 ) =
n→∞ n→∞ 2 ·xn+1
ln(1+nxn )
n for n ≥ 2 and inequality x ≤ (1 + x) ln(1 + x) for x ≥ 0 follows

(1 + nxn ) ln(1 + nxn )

0 ≤ lim sup nxn ≤ lim sup n(n−1)
=
n→∞ n→∞ · xn+1
2
 1 + nx
n ln(1 + xn+1 )

= 2 · lim sup · = 0.
n→∞ n−1 xn+1

It follows 0 ≤ lim sup nxn ≤ lim sup nxn ≤ 0, i.e. lim nxn = 0.
n→∞ n→∞ n→∞

ln t
3. From 1 and 2 and lim = 1 follows
t→0 t

xn+1 xn+1 ln(1 + nxn )

lim = lim · = 1.
n→∞ xn n→∞ ln(1 + xn+1 ) nxn

 The 15th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 2005
Category II

Problem 1 For an arbitrary square matrix M , define

M M2 M3
exp(M ) = I + + + + ... .
1! 2! 3!
Construct 2 × 2 matrices A and B such that exp(A + B) 6= exp(A) exp(B).
Solution Note that if A and B commute then obviously exp(A + B) = exp(A) exp(B). So A and B should not
commute.


Let A = 00 10 and B = 01 00 . Then A2 = B 2 = 0,
   
1 1 1 0
exp(A) = I + A = , exp(B) = I + B = ,
0 1 1 1
 
2 1
exp(A) exp(B) = .
1 1

On the other hand,  

1 0
k is even



 0 1
(A + B)k =  
 0 1

 k is odd
1 0


so
∞ ∞
1 1
P P
   
(2k)! (2k+1)! cosh 1 sinh 1
k=0 k=0
exp(A + B) = ∞ ∞ = .
P 1
P 1 sinh 1 cosh 1
(2k+1)! (2k)!
k=0 k=0


 The 15th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 2005
Category II

Problem 2 Let (ai,j )ni,j=1 be a real matrix such that ai,i = 0 for i = 1, 2, . . . , n. Prove that there exists a set
J ⊂ {1, 2, . . . , n} of indices such that
n
X X 1 X
ai,j + ai,j ≥ ai,j .
2 i,j=1
i∈J i∈J
/
j ∈J
/ j∈J

Solution For every I ⊂ {1, 2, . . . , n} let F (I) be defined as follows:

X X
F (I) = ai,j + ai,j .
i∈I,j6∈I i6∈I,j∈I

Let J ⊂ {1,Pn2, . . . , n} satisfy the condition F (J ) = max{F (I) : I ⊂ {1, 2, . . . , n}}. We aim to prove, that
F (J ) ≥ 21 i,j=1 ai,j .
For k ∈ J we have
X X X X
0 ≤ F (J ) − F (J \ {k}) = ak,j + ai,k − ak,j − ai,k .
j6∈J i6∈J j∈J i∈J

Summing the above inequalities with k ∈ J we obtain

X X 1
0 ≤ F (J ) − 2 ai,j hence ai,j ≤ F (J ).
2
i∈J i∈J
j∈J j∈J

Similarily, for k 6∈ J we have

X X X X
0 ≤ F (J ) − F (J ∪ {k}) = ak,j + ai,k − ak,j − ai,k .
j∈J i∈J j6∈J i6∈J

Summing the above inequalities with k 6∈ J we obtain

X X 1
0 ≤ F (J ) − 2 ai,j hence ai,j ≤ F (J ).
2
i6∈J i6∈J
j6∈J j6∈J

Finally, we have:
n
1 X 1 X X 
ai,j = ai,j + ai,j + F (J ) ≤
2 i,j=1 2
i∈J i6∈J
j∈J j6∈J
11 1 
≤ F (J ) + F (J ) + F (J ) = F (J ).
2 2 2

 The 15th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 2005
Category II

Problem 3 Let f : [0, 1] × [0, 1] → R be a continuous function. Find the limit

 2 Z 1 Z 1
(2n + 1)!
n
lim xy(1 − x)(1 − y) f (x, y) dx dy.
n→∞ (n!)2 0 0

1 1


Solution Answer: f 2, 2 .
Proof: Set  (2n + 1)! 2 Z Z
n
Ln (f ) = xy(1 − x)(1 − y) f (x, y) dx dy .
(n!)2
Q

1 1 k 1 l



Step 1. lim Ln (xk y l ) = 2k+l
= 2 2 . This is a straightforward calculation:
n→∞

ZZ

n
Z 1 Z 1
k l n+k n
xy(1 − x)(1 − y) x y dx dy = x (1 − x) dx y n+l (1 − y)n dy =
0 0
Q
(n + k)!n! (n + l)!n!
(integrate by parts) = · .
(2n + k + 1)! (2n + l + 1)!

Thus
(n + 1)(n + 2) . . . (n + k)
lim Ln (xk y l ) = lim ·
n→∞ n→∞ (2n + 2)(2n + 3) . . . (2n + k + 1)
(n + 1)(n + 2) . . . (n + l) 1
· = k+l .
(2n + 2)(2n + 3) . . . (2n + l + 1) 2

Step 2. The desired result is satisfied for every polynomial P (x, y). Indeed, the limit and Ln are linear
operators.
Step 3. Fix an arbitrary ε > 0. A polynomial P (x, y) can be chosen such that |f (x, y) − P (x, y)| < ε for
every (x, y) ∈ Q . Then
|Ln (f ) − Ln (P )| ≤ Ln (|f − P |) < Ln (ε · I) = ε ,
where I(x, y) = 1, for every (x, y) ∈ Q.  1 1


According to step 2 there exists n0 such that Ln (P ) − P 2, 2
 < ε for n ≥ n0 . For these integers

Ln (f ) − f 1 , 1  ≤ |Ln (f ) − Ln (P )| + Ln (P ) − P 1 , 1  + f 1 , 1 − P 1 , 1  < 3ε ,
          
2 2 2 2 2 2 2 2
which concludes the proof. 
 The 15th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 6th April 2005
Category II

Problem 4 Let R be a finite ring with the folowing property: for any a, b ∈ R, there exists an element c ∈ R
(depending on a and b) such that a2 + b2 = c2 . Prove that for any a, b, c ∈ R, there exists an element d ∈ R
such that 2abc = d2 .
(Here 2abc denotes abc + abc. The ring R is assumed to be associative, but not necessarily commutative and
not necessarily containing a unit.)
Solution Let us denote S = {x2 : x ∈ R}. The property of R can be rewritten as S + S ⊆ S. For each y ∈ S
the function S → S, x 7→ x + y is injective, but since S is finite it is indeed bijective. Therefore, S is also closed
under subtraction, so S is an additive subgoup of R.
Now for any x, y ∈ R we have xy + yx = (x + y)2 − x2 − y 2 , so

xy + yx ∈ S .

We take arbitrary a, b, c ∈ R and substitute:

x = a, y = bc ⇒ abc + bca ∈ S (1)

x = c, y = ab ⇒ cab + abc ∈ S (2)
x = ca, y = b ⇒ cab + bca ∈ S (3)

If we add (1), (2) and subtract (3), we shall obtain 2abc ∈ S. 

 The 16th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 29th March 2006
Category I

Problem 1 Given real numbers 0 = x1 < x2 < · · · < x2n < x2n+1 = 1 such that xi+1 − xi ≤ h for 1 ≤ i ≤ 2n,
show that
n
1−h X 1+h
< x2i (x2i+1 − x2i−1 ) < .
2 i=1
2
Pn
Solution (by Stijn Cambie) Notice that i=1 (x2i+1 + x2i−1 )(x2i+1 − x2i−1 ) = x22n+1 − x21 = 1. Hence we have
to prove
 n
X  X n 
1 − 2 (x2i )(x2i+1 − x2i−1 ) =  (x2i+1 + x2i−1 − 2x2i )(x2i+1 − x2i−1 ) ≤ h
   
i=1 i=1

Now
Xn  X n  
 (x2i+1 + x2i−1 − 2x2i )(x2i+1 − x2i−1 ) ≤ x2i+1 + x2i−1 − 2x2i (x2i+1 − x2i−1 )
   
i=1 i=1
n
X
≤ h(x2i+1 − x2i−1 ) = h
i=1

because |x2i+1 + x2i−1 − 2x2i | ≤ max(x2i+1 − x2i , x2i − x2i−1 ). Equality can not occur, because we would need
x2i+1 − x2i or x2i − x2i−1 would have to be zero in that case. 
 The 16th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 29th March 2006
Category I

Problem 2 Suppose that (an ) is a sequence of real numbers such that the series
∞
X an
n=1
n

is convergent. Show that the sequence

n
P
aj
j=1
bn =
n
is convergent and find its limit.
Pn−1
Pn Ai
Solution (by Stijn Cambie) Write An = i=1 aii . Suppose this converges to A. We have bn = An − i=1 n .
This converges to zero as n −→ ∞. Indeed, for each  > 0 take some I0 such that |Ai − A| ≤ 3 for i ≥ I0 and
take n0 > I0 such that
0 −1
IX
 
(n0 − I0 + 1) + |A − Ai | < n0
3 i=1
2

and | nA0 | < 6 .

Then for n > n0 we have
 Pn−1    Pn−1

i=1 Ai 
A |A − Ai |   
|bn | = An −  ≤ |An − A| +   + i=1 < + +

 n  n n 3 2 6


 The 16th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 29th March 2006
Category I

Problem 3 Two players play the following game: Let n be a fixed integer greater than 1. Starting from number
k = 2, each player has two possible moves: either replace the number k by k + 1 or by 2k. The player who is
forced to write a number greater than n loses the game. Which player has a winning strategy for which n?
Solution (by Stijn Cambie)
Write n in base 4. We will prove that the second player B can only win when the representation contains
only 0 and 2s. The first player A wins in the other cases.
Claim 1 Person A wins for n odd.
Proof He just has to do k → k + 1 in each step, in each move he makes an even number odd. Next B can
make the number only even. As n is odd, A won’t ever make an even number smaller than n bigger than n by
adding one. Hence B has to do this and will lose. 

B wins for n = 2, trivial. When A chooses 2 → 4 in his first step, he wins for n = 4, 6. Person B can win
for 8 by multiply the number of A by 2 in his first step and then both have to add one each step and B reaches
8. Hence the base cases are correct.
Claim 2 When person X wins for n, he can also win for 4n and 4n + 2.
Proof At some step, the other person will get a number k bigger than n, next X makes 2k > 2n + 1 and by
alternating adding one, we see X will reach every number. 

Hence if A wins for some n, he wins also for 4n, 4n + 1, 4n + 2, 4n + 3 by both claims. So player B can only
win for 4n, 4n + 2 where n is a number that has the predicted representation (and so do 4n, 4n + 2). 
 The 16th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 29th March 2006
Category I

Problem 4 Let A = [aij ]n×n be a matrix with nonnegative entries such that
n X
X n
aij = n .
i=1 j=1

1. Prove that |det A| ≤ 1.

2. If |det A| = 1 and λ ∈ C is an arbitrary eigenvalue of A, show that |λ| = 1.

(We call λ ∈ C an eigenvalue of A if there exists a non-zero vector x ∈ Cn such that Ax = λx.)
Solution (by Stijn Cambie)

1. We prove that the statement is true and equality occur only for a permutation matrix.
We can prove this by induction.
For n = 1 this is trivial (wehave det
 A = 1).
a b
For n = 2 we have for A = that | det(A)| = |ad − bc| ≤ ( a+d 2 b+c 2
2 ) +( 2 ) ≤(
a+d+b+c 2
2 ) = 1.
c d
Equality occurs only when a = d = 1 or b = c = 1. So the induction hypothesis is proven for n ≤ 2.
Induction step:
Assume the sum of the entries in the (n + 1)-th row of A is x. The sum of all other entries is n + 1 − x.
By homogenizing and using the induction hypothesis, we have that for each minor the determinant of it
has absolute value ≤ ( |n+1−x|
n )n .
|n+1−x| n
n+1
Now | det A| ≤ |a(n+1)j || det M(n+1)j | ≤ nx ) ≤ nx+n(n+1−x)
P
n ( n n(n+1) = 1 by AM − GM.
Equality could only occur when x = 1 and each | det M(n+1)j | = 1 when |a(n+1)j | > 0 so each minor has
to be a permutation matrix, which is possible only once.
Hence A is also a permutation matrix.

2. If λ is an eigenvalue and an eigenvector is (x1 x2 · · · xn )T , then λxi = xj for some i, j, such that xi 6= 0.
Repeating this, we get cycles such that xi = λm xi for some m and hence λm = 1, hence |λ| = 1.

 The 16th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 29th March 2006
Category II

Problem 1
1. Let u and v be two nilpotent elements in a commutative ring (with or without unity). Prove that u + v
is also nilpotent. (An element u is called nilpotent if there exists a positive integer n for which un = 0.)

2. Show an example of a (non-commutative) ring R and nilpotent elements u, v ∈ R such that u + v is not
nilpotent.
Solution (by Stijn Cambie)
1. As u, v are nilpotent, there exist n, m such that un = 0 P= v m . This t s

means u = 0 for all t ≥ n and v = 0
n+m n+m i n+m−i
for all s ≥ m. Next u + v is nilpotent as (u + v) = i uv = 0,
because each term ui v n+m−i = 0 as i ≥ n or n + m − i ≥ m so ui or v n+m−i = 0 in each summand.
   
1 −1 −1 −1
2. Take the ring R = (Z2∗2 , +, ·) with elements u = and v = .
1 −1 1 1
0 (−2)n
     
2 2 0 0 0 −2 n
Then u = v = while u + v = is not nilpotent as (u + v) = .
0 0 2 0 2n 0

 The 16th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 29th March 2006
Category II

Problem 2 Let (G, ·) be a finite group of order n. Show that each element of G is a square if and only if n is
odd.
Solution (by Stijn Cambie)
1. If n is odd, we know by Lagrange’s theorem that for every element g ∈ G : |g| = ord(g) divides |G| = n
and hence |g| is also odd. Write t = |g|+1 t
2 , then g is the square of g . As g was taken arbitrary, it holds
for every element of G.
2. If n is even, we have to find at least one element which isn’t a square.
Claim There exist some element with order 2.
Proof Suppose the contrary. We know that the inverse in a group is unique and 1 is its own inverse. For every
other element g, we would have g 6= g −1 as g 2 = 1 means |g| = 1, 2 and |g| = 1 is only possible for 1. Now look
at the sets {g, g −1 }. A {1, 1} contains only one element and every other set contains 2 elements, we would have
split up G in one one-element-set and two-element-sets, which is impossible as 2 | ord(G).
Hence there is at least yet one element g such that g = g −1 and hence g 2 = 1. 

Because G is finite, we can write all orders of the different elements. Take the maximum m > 0 of
{v2 (|g|) | g ∈ G}. Next, choose an element h ∈ S such that 2m | ord(h) = 2t. Suppose h is a square, we
have h = k 2 for some k ∈ G. Then we have h2t = k 4t = 1, so ord(k) | 4t. Next k 2t = ht 6= 1 as the order of h
is 2t. This means ord(k) | 4t and ord(k) - 2t hence v2 (k) = v2 (4t) = m + 1. This is in contradiction with the
way we have chosen m. Hence h is an element of G which is not a square. So if n is even, not all elements are
squares.

 The 16th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 29th March 2006
Category II

Problem 3 For
a function f
: [0, 1] → R the secant of f at points a, b ∈ [0, 1], a < b, is the line in R2 passing
through a, f (a)
and b, f (b) . A function is said to intersect its secant at a, b if there exists a point c ∈ (a, b)
such that c, f (c) lies on the secant of f at a, b.

1. Find the set F of all continuous functions f such that for any a, b ∈ [0, 1], a < b, the function f intersects
its secant at a, b.
2. Does there exist a continuous function f ∈
/ F such that for any rational a, b ∈ [0, 1], a < b, the function f
intersects its secant at a, b?

Solution

 The 16th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 29th March 2006
Category II

Problem 4 Let f : [0, +∞) → R be a strictly convex continuous function such that

f (x)
lim = +∞ .
x→+∞ x

R +∞

Prove that the improper integral 0 sin f (x) dx is convergent but not absolutely convergent.
Solution

 The 17th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 28th March 2007
Category I

Problem 1. Can the set of positive rationals be split into two nonempty disjoint subsets
Q1 and Q2 , such that both are closed under addition, i.e. p + q ∈ Qk for every p, q ∈ Qk ,
k = 1, 2?
Can it be done when addition is exchanged for multiplication, i.e. p · q ∈ Qk for every
p, q ∈ Qk , k = 1, 2?
[10 points]

Problem 2. Alice has got a circular key ring with n keys, n ≥ 3. When she takes it out
of her pocket, she does not know whether it got rotated and/or flipped. The only way she
can distinguish the keys is by colouring them (a colour is assigned to each key). What is the
minimum number of colours needed? [10 points]

Problem 3. A function f : [0, ∞) → R \ {0} is called slowly changing if for any t > 1 the
limit lim ff(tx)
(x) exists and is equal to 1. Is it true that every slowly changing function has
x→∞
for sufficiently large x a constant sign (i.e., is it true that for every slowly changing f there
exists an N such that for every x, y > N we have f (x)f (y) > 0?) [10 points]

Problem 4. Let f : [0, 1] → [0, ∞) be an arbitrary function satisfying

f (x) + f (y) x + y
≤f +1
2 2

for all pairs x, y ∈ [0, 1]. Prove that for all 0 ≤ u < v < w ≤ 1,

w−v v−u
f (u) + f (w) ≤ f (v) + 2 .
w−u w−u

[10 points]

28-Mar-2007 11:33
Category I
Problem 1. Can the set of positive rationals be split into two nonempty disjoint subsets
Q1 and Q2 , such that both are closed under addition, i.e. p + q ∈ Qk for every p, q ∈ Qk ,
k = 1, 2? Can it be done when addition is exchanged for multiplication, i.e. p · q ∈ Qk for every
p, q ∈ Qk , k = 1, 2?

Solution. (a) No. If pq , rs ∈ Qk then of course ps+qr p

qs ∈ Qk . Adding n times q and m times s
r
nps+mqr
gives qs ∈ Qk for all positive integers n, m, hence ñp+m̃r ∈ Qk for all positive integers ñ, m̃.
So if pqkk , srkk ∈ Qk we get that p1 p2 + r1 r2 ∈ Q1 ∩ Q2 .
(b) Yes, for instance
nm o
Q1 = ∈ Q+ : (m, n) = 1 and 2 | n and Q2 = Q+ \ Q1 .
n

1
Problem 2. Alice has got a circular key ring with n keys, n ≥ 3. When she takes it out
of her pocket, she does not know whether it got rotated and/or flipped. The only way she
can distinguish the keys is by colouring them (a colour is assigned to each key). What is the
minimum number of colors needed?

Solution. Clearly at least two colors are needed in any case to distinguish between at least
two keys. For three, four or five keys on the ring, we will show that three colors are necessary.
For six or more keys on the ring, we will show that two colors suffice. Choose one key and
denote it with k1 . Order all other keys in natural order as they follow each other going from k1
around the ring in one direction. For 1 ≤ i ≤ n denote with c(ki ) color of the key ki . Without
loss of generality let c(k1 ) = 1.
Suppose that two colors suffice for n = 3. Then there are two similar possibilities for coloring
the keys. Either c(k2 ) = c(k3 ) = 2 or c(k2 ) = 1. In the first case one can not distinguish between
keys k2 and k3 . In the second case one can not distinguish between keys k1 and k2 . Hence for
n = 3 we need three colors.
Suppose that two colors suffice for n = 4. Then there are four possibilities for coloring the
keys. If c(k2 ) = c(k3 ) = c(k4 ) = 2, then k2 and k4 can not be distinguished (rotation of the key
ring through the line across k1 and k3 interchanges k2 and k4 ). If c(k2 ) = 1 and c(k3 ) = c(k4 ) = 2
then there is a rotation that interchanges k1 and k2 and also interchanges k3 and k4 (similar is
the case when c(k4 ) = 1 and c(k2 ) = c(k3 ) = 2). If c(k3 ) = 1 and c(k2 ) = c(k4 ) = 2 then there is
a rotation that interchanges k1 and k3 and there is also other rotation that interchanges k2 and
k4 . Hence for n = 4 at least three colors are needed. Consider the following coloring: c(k1 ) = 1,
c(k2 ) = 2, c(k3 ) = 3 and c(k4 ) = 1 (one possibility). Keys k1 and k4 have the same color, but
one can distinguish between them since k1 has a neighbor colored with color 1 and a neighbor
colored with color 2, while k4 has also one neighbor colored with color 1, but the other neighbor
is colored with color 3. Hence three colors suffice for n = 4.
Suppose that two colors suffice for n = 5. Then there are two possibilities for coloring the
keys: all other keys than k1 are colored with color 2 (the similar is the case when one key gets
color 1, only the roles of the colors are interchanged) or one of them gets color 1 and other
three get color 2 (the same is the case when two keys get color 2, only the roles of the colors are
interchanged). In first case one can not distinguish between keys k2 and k5 and also between
keys k3 and k4 (there is a rotation of the key ring where keys in both pairs interchange, while
k1 is fixed). When there is a key other than k1 with color 1 we need to consider two subcases.
If c(k2 ) = 1 (similar is the case when c(k5 ) = 1) we can not distinguish between k1 and k2 (also
between k3 and k5 ). If c(k3 ) = 1 (similar is the case when c(k4 ) = 1) we can not distinguish
between k1 and k3 (also between k4 and k5 ). Hence for n = 5 at least three colors are needed.
Consider the following coloring: c(k1 ) = 1, c(k2 ) = 2, c(k3 ) = 3 and c(k4 ) = c(k5 ) = 2 (one
possibility). Keys k2 , k4 and k5 have the same color, but one can distinguish between them
since k2 is the only one between them that has a neighbor colored with color 1 and a neighbor
colored with color 3, while only k4 has a neighbor colored with color 3 and a neighbor colored
with color 2. Hence three colors suffice for n = 5.
For n ≥ 6 consider the following coloring: c(k1 ) = 1, c(kn ) = 2, c(kn−1 ) = c(kn−2 ) = 1 and
c(ki ) = 2 for 2 ≤ i ≤ n − 3. Then k1 is the only key of color 1 with both neighbors colored with
color 2. Keys kn−1 and kn−2 both have neighbors of two different colors, but the distance (the
smallest of the two numbers: number of the keys lying between the two keys in one and other
direction) between kn−1 and k1 is one while the distance between kn−2 and k1 is two. Hence
one can distinguish between all three keys colored with color 1. Among keys colored with color
2 only kn has both neighbors colored with color 1. All other keys: ki for 2 ≤ i ≤ n − 3 have
either one or two neighbors colored with color 2. But any ki , where 2 ≤ i ≤ n − 3, has a pair of
distances: distance between ki and k1 and distance between ki and kn−2 that is different from
any other pair of distances of some key kj 6= ki for 2 ≤ j ≤ n − 3 . Hence we can distinguish
also between keys colored with color 2.

2
Problem 3. A function f : [0, ∞) → R \ {0} is called slowly changing if for any t > 1 the
limit lim ff(tx)
(x) exists and is equal to 1. Is it true that every slowly changing function has for
x→∞
sufficiently large x a constant sign (that is — it is true that for every slowly changing f there
exists N such that for every x, y > N we have f (x)f (y) > 0?)

Remark. The assumption f (x) 6= 0 is only technical, to avoid explaining what does the limit
mean in the other case, and in reality changes nothing.

Remark. The reader is encouraged to try and solve the problem himself before reading the
solution. The author’s and the proposer’s opinion is that although the solution is simple, it is
not so easy to find it (both tried, both succeeded, but both spent some time on it before getting
the correct idea).

Solution. Take t = 2. Take such a N > 0 that for x > N we have ff(2x)(x) > 0. This means
f (2x) and f (x) are of the same sign for x > N . Suppose that for any x > N we have that
f (x) and f (N ) are of a different sign. Let t = Nx . Then ff(tN )
(N ) < 0, and by easy induction
f (t2k N ) f (tx)
< 0 for any k ∈ N, which contradicts the assumption
f (2k N ) f (x) → 1 when x tends to ∞. The
contradiction proves the thesis.

3
Problem 4. Let f : [0, 1] → [0, ∞) be an arbitrary function satisfying
f (x) + f (y) x + y 
≤f +1 (1)
2 2
for all pairs x, y ∈ [0, 1]. Prove that for all 1 ≤ u < v < w ≤ 1,
w−v v−u
f (u) + f (w) ≤ f (v) + 2 .
w−u w−u

Solution. Let
 
w−v v−u
M (u, w) = sup f (u) + f (w) − f (v) ;
v∈(u,w) w−u w−u

we have to prove M (u, w) ≤ 2. Note that M (u, w) is finite, because

w−v v−u
f (u) + f (w) − f (v) ≤ 1 · f (u) + 1 · f (w) − 0 = f (u) + f (w) .
w−u w−u
Let ε > 0 be an arbitrary positive real number. Choose v such that
w−v v−u
f (u) + f (w) − f (v) > M (u, w) − ε .
w−u w−u
u+w
If v ≤ 2 , then apply (1) for x = u and y = u + 2(v − u) = 2v − u:
f (u) + f (2v − u)
≤ f (v) + 1 ;
2

w−v v−u
M (u, w) − ε < f (u) + f (w) − f (v)
w−u w−u
w−v v−u f (u) + f (2v − u)
≤ f (u) + f (w) − +1
w− u w−u 2 
1 w − (2v − u) (2v − u) − u
= f (u) + f (w) − f (2v − u) + 1
2 w−u w−u
1
≤ M (u, w) + 1 ;
2
M (u, w) ≤ 2 + 2ε .
u+w
Otherwise, if 2 < v, apply x = w − 2(w − v) = 2v − w and y = v in (1):
f (2v − w) + f (w)
≤ f (v) + 1 ;
2

w−v v−u
M (u, w) − ε < f (u) + f (w) − f (v)
w−u w−u
w−v v−u f (2v − w) + f (w)
≤ f (u) + f (w) − +1
w− u w−u 2 
1 w − (2v − w) (2v − w) − u
= f (u) + f (w) − f (2v − w) + 1
2 w−u w−u
1
≤ M (u, w) + 1 ;
2
M (u, w) ≤ 2 + 2ε .
In both cases we obtained M (u, w) ≤ 2 + 2ε. This holds for all ε, therefore M (u, w) ≤ 2.

4
 The 17th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 28th March 2007
Category II

Problem 1. Construct a set A ⊂ [0, 1] × [0, 1] such that A is dense in [0, 1] × [0, 1] and every
vertical and every horizontal line intersects A in at most one point. [10 points]

Problem 2. Let A be a real n × n matrix satisfying

A + At = I ,

where At denotes the transpose of A and I the n × n identity matrix. Show that det A > 0.
[10 points]

Problem 3. Let f : [0, 1] → R be a continuous function such that f (0) = f (1) = 0. Prove
that the set 
A := h ∈ [0, 1] : f (x + h) = f (x) for some x ∈ [0, 1]
is Lebesgue measureable and has Lebesgue measure at least 21 . [10 points]

Problem 4. Let S be a finite set with n elements and F a family of subsets of S with the
following property:
A ∈ F, A ⊆ B ⊆ S =⇒ B ∈ F.
Prove that the function f : [0, 1] → R given by
X
f (t) := t|A| (1 − t)|S\A|
A∈F

is nondecreasing (|A| denotes the number of elements of A). [10 points]

28-Mar-2007 11:33
Category II
Problem 1. Construct a set A ⊂ [0, 1] × [0, 1] such that A is dense in [0, 1] × [0, 1] and every
vertical and every horizontal line intersects A in at most one point.
α
Solution. Take α, β 6∈ Q such that β 6∈ Q. Then

A := ({nα}, {nβ}) : n ∈ N ,

where {x} denotes the fractional part of x, fulfills the assumptions.

1
Problem 2. Let A be a real n × n matrix satisfying

A + At = I ,

where At denotes the transpose of A and I the n × n identity matrix. Show that det A > 0.

Solution. The assumption A + At = I is equivalent to saying A = S + 12 I where S denotes

an arbitrary real skew symmetric matrix. In particular, there exists some orthogonal matrix T
that diagonalizes S and for which D := T t ST contains the eigenvalues of S. They are either
zero or purely imaginary and pairwise conjugated, i.e. of the form

r1 i, −r1 i, . . . , rs i, −rs i, 0, . . . , 0

with rk ∈ R for all k = 1, . . . , s. The determinant of A is evaluated as follows:

 1   1 
det A = det S + I = det D + I
2 2
since det(T t T ) = 1 and with the notations from above this expression is
s 
 1 n−2s Y s 
1  1   1 n−2s Y 1 
+ rk i − rk i = + rk2 .
2 2 2 2 4
i=1 i=1

As all factors are strictly positive the result follows.

2
Problem 3. Let f : [0, 1] → R be a continuous function such that f (0) = f (1) = 0. Prove
that the set 
A := h ∈ [0, 1] : f (x + h) = f (x) for some x ∈ [0, 1]
has Lebesgue measure at least 12 .

Solution. Let us observe, that if f is continuous then A is closed, thus A is Lebesgue

measurable. Moreover the set

B := h ∈ [0, 1] : 1 − h ∈ A

has the same Lebesgue measure as the set A. We show that A ∪ B = [0, 1].
For any h ∈ [0, 1] we define a function g : [0, 1] → R by

g(x) = f (x + h) − f (x) if x + h ≤ 1

and
g(x) = f (x + h − 1) − f (x) if x + h > 1 .
From the assumption we have that g is continuous. If f has its minimum and maximum,
respectively, in x0 and x1 , then g(x0 ) ≥ 0 and g(x1 ) ≤ 0. From Darboux property we have that,
there exists x2 such that g(x2 ) = 0, therefore h ∈ A or h ∈ B. This completes the proof.

3
Problem 4. Let S be a finite set with n elements and F a family of subsets of S with the
following property:
A ∈ F, A ⊆ B ⊆ S =⇒ B ∈ F
Prove that the function f : [0, 1] → R given by
X
f (t) := t|A| (1 − t)|S\A|
A∈F

is nondecreasing (|A| denotes the number of elements of A).

Solution. Without loss of generality assume S = {1, 2, . . . , n}. For each subset A and every
n
(j)
It,A in Rn , where
Q
t ∈ [0, 1] construct a set It,A :=
j=1
(
(j) [0, t) if j ∈ A
It,A :=
[t, 1] if j ∈
/ A.

It’s clear that for any two different subsets A and B the sets It,A and It,B are disjoint.SSince the
c
volume of It,A is equal to t|A| (1 − t)|A | we have that f (t) is equal to the volume of A∈F It,A .
So the claim will be proved if we prove that
[ [
It1 ,A ⊆ It2 ,A for all 0 < t1 < t2 < 1 . (1)
A∈F A∈F

Take an arbitrary x = (x1 , x2 , . . . , xn ) ∈ It1 ,A for some A ∈ F. Construct a set B ⊆ S such that
j ∈ B if and only if xj ≤ t2 . If j ∈ / B then xj > t2 > t1 which implies j ∈ / A. So A ⊆ B and thus
B ∈ F. Moreover, from the definition of B, we have x ∈ It2 ,B . This proves (1) and the problem
is solved.

4
 The 18th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 2nd April 2008
Category I

Problem 1. Find all complex roots (with multiplicities) of the polynomial

2008
X
1004 − |1004 − n| xn .


p(x) =
n=1

[10 points]

Problem 2. Find all functions f : (0, ∞) → (0, ∞) such that

f (f (f (x))) + 4f (f (x)) + f (x) = 6x .

[10 points]

Problem 3. Find all c ∈ R for which there exists an infinitely differentiable function
f : R → R such that for all n ∈ N and x ∈ R we have

f (n+1) (x) > f (n) (x) + c .

[10 points]

Problem 4. The numbers of the set {1, 2, . . . , n} are colored with 6 colors. Let

S := (x, y, z) ∈ {1, 2, . . . , n}3 : x + y + z ≡ 0 (mod n)



and x, y, z have the same color

and
D := (x, y, z) ∈ {1, 2, . . . , n}3 : x + y + z ≡ 0 (mod n)


and x, y, z have three different colors .
Prove that
n2
|D| ≤ 2|S| + .
2
(For a set A, |A| denotes the number of elements in A.) [10 points]

2-Apr-2008 12:08
Problem j18-I-1. Find all complex roots (with multiplicities) of the polynomial

2008
X
1004 − |1004 − n| xn .


p(x) =
n=1

Solution. Observe, by comparison of coefficients, that

1003
X 2
p(x) = x xn .
n=0

1003
x1004 −1
xn = we conclude that p has the simple root 0 and the roots exp πin
P
Since x−1 , 502 ,
n=0
n = 1, 2, . . . , 1003, with multiplicity 2. 

2-Apr-2008 j18-I-1 1  12:56

Problem j18-I-2. Find all functions f : (0, ∞) → (0, ∞) such that

f (f (f (x))) + 4f (f (x)) + f (x) = 6x .

Solution. Let a ∈ R+ be arbitrary. Set a0 = a, an = f (an−1 ) for n > 0. Then we obtain

recurrence relation
an+3 + 4an+2 + an+1 − 6an = 0 .
Characteristic equation is
y 3 − 4y 2 + y − 6 = 0
with roots −2, −3 and 1. The general solution of recurrence relation is

an = A(−3)n + B(−2)n + C .

If A or B are not equal to 0, we have a contradiction because in range of f we could find

negative values. So the only possible solution is an = C. Because of a0 = a we have an = a
for all n ∈ N0 . Substituting n = 1 we obtain

f (a) = f (a0 ) = a1 = a ,

so for all a ∈ R+ we have f (a) = a.

The only solution of the equation is f (x) = x, what can be easily checked. 

2-Apr-2008 j18-I-2 1  12:56

Problem j18-I-3. Find all c ∈ R for which there exists an infinitely differentiable function
f : R → R such that for all n ∈ N and x ∈ R we have

f (n+1) (x) > f (n) (x) + c . (1)

Solution. For c ≤ 0 we can take f (x) = e2x . Then f (n+1) (x) = 2n+1 e2x > 2n e2x =
(n)
f (x).
For positive c no function satisfies (1). We begin with two simple lemmas.
Lemma 1. If f satisfies (1), then for any x ∈ R there exists an y ≤ x such that f (y) ≤ − 2c .
Proof. If f (t) > − 2c on (−∞, x], then f 0 (t) > 2c for any t < x, thus
Z x
c
f (y) = f (x) − f 0 (t) dt ≤ f (x) − (x − y)
y 2

for any y < x, thus for sufficiently small y we have f (y) < 0, a contradiction.
Lemma 2. If f satisfies (1), then for any x ∈ R such that f (x) < 2c we have f (y) < 2c for
any y ≤ x.
Proof. Suppose that there exists a y ≤ x such that f (y) ≥ − 2c . Let z := sup t ≤ x :


f (t) ≥ − 2c . By the continuity of f (f is differentiable, thus continuous) we have f (z) ≥ − 2c .

By the assumption upon x we have z 6= x. However by (1) we have f 0 (z) ≥ 2c , thus f 0 is
positive on [z, z + ε] for some ε > 0, f is increasing, thus f (t) ≥ f (z) ≥ − 2c for t ∈ [z, z + ε],
a contradiction with the definition of z. Thus by contradiction the thesis is proved.
Now if f satisfies (1), then obviously f 0 also satisfies (1). Thus by Lemmas 1 and 2, there
exists an x0 such that f 0 (t) < − 2c on (−∞, x0 ]. This, however, means f (t) > f (x0 )+(x0 −t) 2c
0
for t < x0 , so for sufficiently small t0 < x0 we have f (t0 ) > − 3c 2 > f (t0 ) − c, which is a
contradiction with (1). Thus no such f exists. 

2-Apr-2008 j18-I-3 1  12:56

Problem j18-I-4. The numbers of the set {1, 2, . . . , n} are colored with 6 colors. Let
S := (x, y, z) ∈ {1, 2, . . . , n}3 : x + y + z ≡ 0 (mod n)


and x, y, z have the same color
and
D := (x, y, z) ∈ {1, 2, . . . , n}3 : x + y + z ≡ 0 (mod n)


and x, y, z have three different colors .
Prove that
n2
|D| ≤ 2|S| +
.
2
(For a set A, |A| denotes the number of elements in A.)
Solution. Denote by n1 , n2 , n3 , n4 , n5 , n6 the number of occurences of the colors. Clearly
n1 + . . . + n6 = n. We prove that
6
1 X X
|S| − |D| = n2u − nu nv . (1)
2 u=1 1≤u<v≤6

For arbitrary u, v, w ∈ {1, 2, . . . , 6}, denote by Nuvw the number of triples (x, y, z),
satisfying x + y + z ≡ 0 (mod n) and having colors u, v and w, respectively. For any u, v we
P6
obviously have Nuvw = nu nv and therefore
w=1

6
1 X X X
|S| − |D| = Nuuu − Nuvw
2 u=1 1≤u<v≤6 w6=u,v
6 
X X  X
n2u −


= Nuuv − nu nv − Nuuv − Nuvv
u=1 v6=u 1≤u<v≤6
6
X X
= n2u − nu nv .
u=1 1≤u<v≤6

Now, applying the AM-QM inequality,

6 6 6
1 X X 3 X 2 1  X 2
|S| − |D| = n2u − nu nv = nu − nu
2 u=1
2 u=1 2 u=1
1≤u<v≤6
6
1 1 X 2 n2
≥ − nu =− .
4 2 u=1
4


Second solution. We present a different proof for the relation (1). We use the nota-
tion Nuvw as well.
For every u = 1, 2, . . . , 6, let CuPbe the set of those numbers from {1, 2, . . . , n} which
have the uth color and let fu (t) := tx .
x∈Cu
Let ε := e2πi/n . We will use that for every integer s,
n−1 
1 X js 1 if s ≡ 0 (mod n)
ε =
n 0 if s 6≡ 0 (mod n) .
j=0

Then, for arbitrary colors u, v, w,

X X X 1 n−1
X
Nuvw = εj(x+y+z)
n j=0
x∈Cu y∈Cv z∈Cw
n−1 n−1
1 X  X jx  X jy  X jz  1X
= ε ε ε = fu (εj )fv (εj )fw (εj )
n j=0 n j=0
x∈Cu y∈Cv z∈Cw

2-Apr-2008 j18-I-4 1 12:56

and
n−1 6
1 1 X X 3 j X 
|S| − |D| = fu (ε ) − 3 fu (εj )fv (εj )fw (εj )
2 n j=0 u=1 u<v<w
n−1 6 6
1 X X j
X X 
= fu (ε ) fu2 (εj ) − fu (εj )fv (εj )
n j=0 u=1 u=1 u<v
n−1 n 6
X 1 X jx X 2 j X 
= ε fu (ε ) − fu (εj )fv (εj ) .
j=0
n x=1 u=1 u<v

The first factor is 0 except if j = 0. Hence,

6 6
1 X X X X
|S| − |D| = fu2 (1) − fu (1)fv (1) = n2u − nu n v .
2 u=1 u<v u=1 u<v

2-Apr-2008 j18-I-4 2  12:56

 The 18th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 2nd April 2008
Category II

Problem 1. Find all functions f : Z → Z such that

19f (x) − 17f (f (x)) = 2x

for all x ∈ Z. [10 points]

Problem 2. Find all continuously differentiable functions f : [0, 1] → (0, ∞) such that
f (1)
f (0) = e and
Z 1 Z 1
dx
2
+ f 0 (x)2 dx ≤ 2 .
0 f (x) 0

[10 points]

Problem 3.√ Find all

√ pairs of natural numbers (n, m) with 1 < n < m such that the
numbers 1, n n and m m are linearly dependent over the field of rational numbers Q.
[10 points]

Problem 4. We consider the following game for one person. The aim of the player is to
reach a fixed capital C > 2. The player begins with capital 0 < x0 < C. In each turn
let x be the player’s current capital. Define s(x) as follows:

x if x < 1
(
s(x) := C − x if C − x < 1
1 otherwise.

Then a fair coin is tossed and the player’s capital either increases or decreases by s(x), each
with probability 21 . Find the probability that in a finite number of turns the player wins by
reaching the capital C. [10 points]

2-Apr-2008 12:09
Problem j18-II-1. Find all functions f : Z → Z such that

19f (x) − 17f (f (x)) = 2x (1)

for all x ∈ Z.
Solution. Suppose that there exists a function f : Z → Z satisfying the above equation.
Then define a function g: Z → Z by

g(x) = x − f (x) . (2)

Taking into account (1) and (2), we get

17g(f (x)) = 2g(x) . (3)

Let us fix y ∈ Z and let a := g(y). Define a sequence (xn )n≥0 as follows

x0 := y , x1 := f (x0 ) , ..., xn := f (xn−1 ) , ...

for any n ∈ N. Now substituting xn into (3) in turn, we get

17 17n
a = g(x0 ) = g(x1 ) = . . . = n g(xn )
2 2
for any n > 0. Consequently, we infer that

2n a = 17n g(xn )

for any n > 0. Since 2 and 17 are relatively prime, we deduce that 17n | a for any n > 0
and therefore a = 0. Moreover, since y was arbitrary, it follows that g(y) = 0 for any y ∈ Z.
Thus y − f (y) = 0 for any y ∈ Z and hence f (y) = y for any y ∈ Z. This implies that only
one function satisfies the equation (1). So, this completes the solution. 

2-Apr-2008 j18-II-1 1  13:39

Problem j18-II-2. Find all continuously differentiable functions f : [0, 1] → (0, ∞) such
that ff (1)
(0) = e and
Z 1 Z 1
dx
2
+ f 0 (x)2 dx ≤ 2 .
0 f (x) 0

Solution. First, we note that if f is such function, then

Z 1 Z 1 Z 1 0 Z 1
1 2 f (x) dx
0≤ f 0 (x) − dx = f 0 (x)2 dx − 2 dx + 2
0 f (x) 0 0 f (x) 0 f (x)
Z 1 Z 1 Z 1
dx
= f 0 (x)2 dx − 2 (ln f (x))0 dx + 2
0 0 0 f (x)
Z 1 Z 1
f (1) dx
= f 0 (x)2 dx − 2 ln + 2
dx ≤ 0 ,
0 f (0) 0 f (x)

f (1) R1 R1
since f (0) = e and dx
0 f (x)2
+ 0
f 0 (x)2 dx ≤ 2. Therefore

Z 1
1 2
f 0 (x) − dx = 0 . (1)
0 f (x)

Since f is continuously differentiable function on [0, 1], the equality (1) is equivalent to

f 0 (x)f (x) = 1 ∀x ∈ [0, 1] . (2)

√
All positive solutions of the differential equation (2) are in the form f (x) = 2x + C for
some C > 0. Since ff (0)
(1)
= e, we have C = e22−1 , and thus
r
2
f (x) = 2x +
e2 − 1

is the unique function satisfying the conditions from the statement. 

2-Apr-2008 j18-II-2 1  13:39

Problem j18 - -
√ II 3. Find
√ all pairs of natural numbers (n, m) with 1 < n < m such that the
numbers 1, n n and m m are linearly dependent over the field of rational numbers Q.
Solution. The answer is n = 2, m = 4.
We begin with the following
√ √
Lemma. The minimal √ (over Q) polynomial f (X) for n n equals X k − ( n n)k , where k is the
minimal satisfying ( n n)k ∈ N.
√
Proof. n n is a root of X n −n = 0. So there is some nonempty subset A of {0, 1, ..., n−1}
such that Y
f (X) = (X − ζ l ) ,
l∈A

2π 2π



where ζ = cos n + i sin
n .
√ √
The free term of f (X) has an absolute value equal to ( n n)|A| . Hence√( n n)deg f (X) is
integer,√and deg f (X) ≥ k follows (k is as in the lemma). But, clearly n n is a root of
X k − ( n n)k , which has integer coefficients.
√ √
Let us assume that 1, n n, m m√are linearly√ dependent over Q, i.e. there are rational
a, b, c not all equal 0 such that
√ a + b n
n + c m
m = 0. √ √
Case a 6= 0. Then, as n n is irrational, we √ have b, c 6=√0. But a + b n n = −c m m has
the same degree of a minimal √ polynomial√as n n, and as m m. Let k be the degree of the
minimal polynomial for m
m. Then y = n n satisfies
√
(a + by)k = ( m m)k ,
√
but y k and ( m m)k are rational, and√
as a, b 6= 0 we obtain that there is a nonzero polynomial
with rational coefficients vanishing
√
n
n of degree smaller than k, a contradiction.
n
n nm
Case a = 0. Hence m m is rational, and this is equivalent to m
√ n is a mn-th power of

a rational. Let p be any prime, and pa k n, pb k m. So we must have mn | am − bn. But

am−bn ≤ am < mn, in view of a ≤ log2 n < n. In a similar way one obtains am−bn > −mn.
So we must have am = bn, the relation independent of the choice of prime p. Thus

n = mm/n ,
√ √ √ √ √
and n n = m m follows. As the function x x has maximum at x = e, we see that n n = m m
holds only for n = 2, m = 4. 

2-Apr-2008 j18-II-3 1  13:39

Problem j18-II-4. We consider the following game for one person. The aim of the player
is to reach a fixed capital C > 2. The player begins with capital 0 < x0 < C. In each turn
let x be the player’s current capital. Define s(x) as follows:

x if x < 1
(
s(x) := C − x if C − x < 1
1 otherwise.

Then a fair coin is tossed and the player’s capital either increases or decreases by s(x), each
with probability 21 . Find the probability that in a finite number of turns the player wins by
reaching the capital C.
Solution. Let us denote by f (x) the probability that player wins with starting capital x.
If x ≤ 1, then he loses if loses the first turn, and if he wins the first turn, he has
capital 2x. Thus f (x) = 12 f (2x).
If x ≥ C − 1 the player wins if he wins the first turn, and has 2x − C in other case, thus
f (x) = 21 + 12 f (2x − C).
In all other cases there is f (x) = 12 f (x − 1) + f (x + 1) .

We will prove that this implies f (x) = Cx .

Let us define g(x) = f (x) − Cx . It is bounded on [0, C] (as f (x) ∈ [0, 1]), and we have
1 x 1 2x

1

 2 f (2x) − C = 2 f (2x) − C = 2 g(2x) for x ≤ 1,

 1 x


 2 f (x − 1) + f (x + 1) − C




 = 1 f (x − 1) − x−1 + f (x + 1) − x+1


2 C C
g(x) = 1




 = 2 g(x − 1) + g(x + 1) for x ∈ (1, C − 1),

1 1 x

2 + 2 f (2x − C) − C





= 12 f (2x − C) − 2x−C

1
= 2 g(2x − C) for x ≥ C − 1.

C

Obviously g(0) = g(C) = 0. Let K = sup f (t) ∈ [0, ∞). Denote n0 = [C] − 1 ≥ 1.
t∈[0,C]
n
We will prove for any natural 0 < n ≤ n0 and x ∈ (n − 1, n] there is g(x) ≤ 2 2−1
n K.

If x ∈ (0, 1] there is g(x) = 21 g(2x) ≤ K

2 .
n−1
Assume, that for x ≤ n − 1 and take x̄ ∈ (n − 1, n]. There is g(x̄ − 1) ≤ 2 2n−1−1 K as
x̄ − 1 ∈ (n − 2, n − 1], and g(x̄ + 1) ≤ K. Thus

1
1  2n−1 − 1  2n − 1
g(x̄) = g(x̄ − 1) + g(x̄ + 1) ≤ K + K = K
2 2 2n−1 2n
as required.
g(x) ≤= 12 g(2x − C) ≤ K 2 for x ≥ C − 1.
Now take x ∈ (n0 , C − 1) (it is empty set for integer C). We have proved that g(x − 1) ≤
2n0 −1 2n0 −1
2n0 K (as x − 1 ∈ (n0 − 1, n0 )) and g(x + 1) ≤ K 2 (x + 1 > C − 1). Thus g(x) ≤ 2n0 K.
n0
Thus we have proved, that g(x) ≤ 2 2n−1 0
K for every x ∈ [0, C], which means that K = 0.
Similarly one can prove, that inf f (t) = 0. Thus g(x) ≡ 0, so f (x) = Cx . 
t∈[0,C]

2-Apr-2008 j18-II-4 1  13:39

 The 19th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 1st April 2009
Category I

Problem 1 Let ABC be a non-degenerate triangle in the euclidean plane. Define a sequence (Cn )∞ n=0 of points
as follows: C0 := C, and Cn+1 is the center of the incircle of the triangle ABCn . Find lim Cn .
n→∞
[10 points]
Solution If α is the angle at A, β the angle at B, then the limit is the point on the side AB dividing it in the
ratio α : β. Let αi and βi be the angles at A and B in ABCi , respectively. Since the center of the incircle is
the intersection of the angle bisectors, we have αi+1 = α2i and βi+1 = β2i ; so the limit point will obviously lie
on AB; furthermore, αβii = αβ =: q for all i. Thus, if Ki is the circumcircle of ABCi , S1,i and S2,i the arcs over
|S |
ACi and BCi , respectively, then |S1,i
2,i |
= q for all i. Now, as the Ci approache AB, the arcs converge to the
corresponding sides of the triangle. Hence, the result follows. 
 The 19th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 1st April 2009
Category I

Problem 2 Prove that the number

k
22 − 1 − 2k − 1
is composite (not prime) for all positive integers k > 2. [10 points]
Solution Denote k
−1
M = 22 − 2k − 1 .
If k is even then 3 | M , and M is composite, since M > 3 for k > 2.
Suppose k is odd. Then
k
 k−1   k−2   1 
2M = 22 − 1 − 2k+1 + 1 = 22 + 1 22 + 1 . . . 22 + 1 − 2k+1 + 1 .

a
 a


Let k + 1 = 2a q with positive odd integer q and a ≥ 1. Then 22 + 1  2M . Indeed, 22 + 1  2k+1 + 1
and  a    k−1   k−2   1 
22 + 1  22 + 1 22 + 1 . . . 22 + 1 ,


since a ≤ k − 1 for k > 2. 

 The 19th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 1st April 2009
Category I

Problem 3 Let k and n be positive integers such that k ≤ n − 1. Let S := {1, 2, . . . , n} and let A1 , A2 , . . . , Ak
be nonempty subsets of S. Prove that it is possible to color some elements of S using two colors, red and blue,
such that the following conditions are satisfied:

(i) Each element of S is either left uncolored or is colored red or blue.

(ii) At least one element of S is colored.
(iii) Each set Ai (i = 1, 2, . . . , k) is either completely uncolored or it contains at least one red and at least one
blue element.
[10 points]
Solution Consider the following system of k linear equations in n real variables x1 , x2 , . . . , xn :
X
xj = 0 , i = 1, 2, . . . , k .
j∈Ai

Since k < n, this system has a nontrivial solution (x1 , x2 , . . . , xn ), i.e. a solution with at least one nonzero xj .
Now color red all elements of the set {j ∈ S : xj > 0}, color blue all elements of the set {j ∈ S : xj < 0}, and
leave uncolored all elements of {j ∈ S : xj = 0}.
P is nontrivial, at least one element is colored. If Ai contains some red element j ∈ S then
Since the solution
xj > 0, and from j∈Ai xj = 0 we see that there exists some j 0 ∈ Ai such that xj 0 < 0, i.e. j 0 is colored blue.
Thus Ai must have elements of both colors. Analogously we argue when Ai contains a blue element. Therefore
we see that the above coloring satisfies all requirements. 
 The 19th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 1st April 2009
Category I

Problem 4 Let (an )∞ ∞

n=1 be a sequence of real numbers. We say that the sequence (an )n=1 covers the set of
∞
P k
positive integers if for any positive integer m there exists a positive integer k such that an = m.
n=1

a) Does there exist a sequence of real positive numbers which covers the set of positive integers?
b) Does there exist a sequence of real numbers which covers the set of positive integers?
[10 points]
Solution The answer to the second question is positive.
First we shall prove that for any n there exists a finite sequence (xi )ki=1
n
of real numbers such that
kn
X
x2m+1
i =0 for 0 ≤ m < n
i=1

and
kn
X
x2m+1
i 6= 0 for m ≥ n .
i=1
Pkn 2m+1
For the simplicity of notation we shall write Sm (xi ) for i=1 xi . We shall prove the thesis by induction
upon n. For n = 0 the appropriate sequence is x1 = 1.
Assume the thesis for n. For n + 1 consider the sequence

(yi )3k
i=1 = (−x1 , −x2 , . . . , −xkn , αx1 , αx2 , . . . , αxkn , αx1 , αx2 , . . . , αxkn ) ,
n

where α = 2−1/(2n+1) . As Sm (xi ) = 0 for m < n, we also have Sm (yi ) = 0. We also have

Sn (yi ) = −Sn (xi ) + 2−1 Sn (xi ) + 2−1 Sn (xi ) = 0 .

For m > n we have

Sm (yi ) = (1 − 2 · 2−(2m+1)/(2n+1) )Sm (xi ) 6= 0 .
Thus the induction step is finished, and the thesis is proved. Moreover it is easy to notice that |xi | ≤ 1 and the
length of the sequence is 3n . Denote by x(n) the sequence of length 3n with Sm (x(n)) = 0 for m < n.
Now to give the required sequence (ai ). Our sequence will be a concatenation of multiples of the finite
sequences x(n) given above. We begin with a1 = 1 (that is we begin by taking x(0)). In the n-th step we
assume that we have some finite sequence ai , with Sm (ai ) = m + 1 for m ≤ n. We also assume that the
elements added in the n-th step will be no larger than n1 .
To pass to the (n + 1)-st step let c = n + 2 − Sn+1 (ai ), and let d = Sn+1 (x(n + 1)). Take an integer
N >  (n+1)c
 
, let α = c , |α| < 1 . We add N copies of the sequence αx(n + 1) to the end of ai . This does
d dN n+1
not change Sm (ai ) for m < n + 1 (as Sm (x(n + 1)) = 0, and after the addition we have Sn+1 (ai ) = n + 2. Also
1
all the added elements are of absolute value no larger than n+1 .
Now to prove that for this series we have G2k+1 = k + 1. As Sm (ai ) = m + 1 after every step, no other limit
is possible, we only have to check convergence. Note, however, that after the n–th step we only add sequences
x(m) for m > n, which in Pturn are concatenations of sequences x(n), with some coefficients. Thus every 3n –th
2n+1
partial sum in the series ai is going to be exactly equal to n + 1. The partial sums “in the middle” cannot
n
differ from this value by more than 32 times the value of the maximal element |ai | in the appropriate interval,
and this converges to zero. Thus for any n we do, in fact, have convergence.
For the first question, obviously the same series suffices.
For the last question, the answer is negative. As ai are positive, we may rearrange them in decreasing order.
Take k0 to be the first k for which Gk is finite. For Gk0 to be finite, we have to have ai convergent to zero, thus
only a finite number of terms is larger than 1, assume these P are the first n terms. Note that as for i > n we
∞
have ai ≤ 1, we also have that aki decreases with k, and thus i=n+1 aki decreases with k, and thus is bounded
P∞
by C := i=n+1 aki 0 . As Gk are assumed to attain unbounded values, we have to have terms larger than 1,
thus n > 0.
Assume the first m terms of ai are equal, 1 ≤ m ≤ n. Then for k ≥ k0 and l ≤ k we have

Gl ≤ mak1 + nakm+1 + C .
On the other hand Gl ≥ mak+1
1 for l ≥ k. For sufficiently large k, however, we have

mak+1
1 > mak1 + nakm+1 + C + 2 ,

which means that there is an integer number between mak+1

1 and mak1 + nakm+1 + C which is not the value of
Gk for any k. 
 The 19th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 1st April 2009
Category II

Problem 1 A positive integer m is called self-descriptive in base b, where b ≥ 2 is an integer, if:

i) The representation of m in base b is of the form (a0 a1 . . . ab−1 )b
(that is m = a0 bb−1 + a1 bb−2 + · · · + ab−2 b + ab−1 , where 0 ≤ ai ≤ b − 1 are integers).

ii) ai is equal to the number of occurences of the number i in the sequence (a0 a1 . . . ab−1 ).
For example, (1210)4 is self-descriptive in base 4, because it has four digits and contains one 0, two 1s, one 2
and no 3s.
a) Find all bases b ≥ 2 such that no number is self-descriptive in base b.

b) Prove that if x is a self-descriptive number in base b then the last (least significant) digit of x is 0.
[10 points]
Solution
1. For b ≥ 7 it is easy to verify that the number of the form (b − 4)bb−1 + 2bb−2 + bb−3 + b4 is a self descriptive
number (it contains b − 4 instances of digit 0, two instances of digit 1, one instance of digit 2 and one
instance of digit b − 4), and numbers 21200(5) and 2020(4) are self-descriptive numbers in bases 5 and 4,
respectively.
It remains to show that for bases 2, 3 and 6 no self descriptive numbers exist. First note, that a self-
descriptive number (in any admissible base) contains at least one instance of the digit 0. If it does not,
then the first digit is 0, which is a contradiction.
It is easy to prove the claim for b = 2, 3.
Let us prove it for b = 6. Assume there exists x = (b0 b1 b2 b3 b4 b5 )(6) , where x is a self-descriptive number.
We observe the following about x:
P5
(a) i=0 bi = 6
(b) b0 6= 0
P5
(c) i=1 bi = |{bi , bi 6= 0, i ≥ 1}| + 1
(d) Other than the first digit, the set of all other non-zero digits consists of several 1’s and one 2.
Observation 1d implies that all but one of the digits b3 , b4 and b5 are 0, now it is easy to check, that no
such number is self-descriptive, which is a contradiction. Therefore base b = 6 contains no self-descriptive
numbers.

2. Assume that there is in fact a self-descriptive number x in base b that it is b-digits long but not a multiple
of b. The digit at position b − 1 must be at least 1, meaning that there is at least one instance of the digit
b − 1 in x. At whatever position a that digit b − 1 falls, there must be at least b − 1 instances of digit a in
x. Therefore, we have at least one instance of the digit 1, and b − 1 instances of a. If a > 1, then x has
more than b digits, leading to a contradiction of our initial statement. And if a = 0 or a = 1, that also
leads to a contradiction.
3. These numbers are: 1210, 2020, 21200, 3211000, 42101000, 521001000, 6210001000. That these are the
only such numbers, follows from previous observations.

 The 19th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 1st April 2009
Category II

Problem 2 Let E be the set of all continuously differentiable real valued functions f on [0, 1] such that f (0) = 0
and f (1) = 1. Define
Z 1
J(f ) = (1 + x2 )(f 0 (x))2 dx .
0

a) Show that J achieves its minimum value at some element of E.

b) Calculate min J(f ).
f ∈E
[10 points]
Solution By the fundamental theorem of Calculus, we have
Z 1 
0 0
1 = |f (1) − f (0)| =  f 00 (x) dx .
 
0

Next, by using the Cauchy-Schwartz inequality, we obtain

Z 1   Z 1 √1 + x 2 
00
f (x) dx = √ f 00 (x) dx
   
 2
0 0 1 + x
Z 1 1/2 Z 1 1
1/2
2 00 2
≤ (1 + x )(f (x)) dx 2
dx
0 0 1+x
Z 1 1/2  1 1/2
= (1 + x2 )(f 00 (x))2 dx arctan x

0 0
Z 1 1/2 √π
= (1 + x2 )(f 00 (x))2 dx .
0 2

Hence Z 1
4
inf (1 + x2 )(f 00 (x))2 dx ≥ .
f ∈E 0 π
Finally, let Z x
4
f (x) := arctan t dt
π 0

for x ∈ [0, 1]. Then f 0 (x) = π4 arctan x (by the fundamental theorem of Calculus) and f 00 (x) = 4 1
π 1+x2 , for
x ∈ [0, 1]. Consequently, we deduce that f ∈ E and
1
16 1 1
Z  4 1 2 Z
16 π 4
J(f ) = (1 + x2 ) dx = dx = 2 · = ,
0 π 1 + x2 π 2 0 1 + x2 π 4 π

which proves that J attains its minimum on E. This completes the solution. 
 The 19th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 1st April 2009
Category II

2 2
Problem 3 Let A be an n × n square matrix with integer entries. Suppose that p2 Ap = q 2 Aq + r2 In for
some positive integers p, q, r where r is odd and p2 = q 2 + r2 . Prove that |det A| = 1.
(Here In means the n × n identity matrix.) [10 points]
Solution Consider the function f : R → R.
2 2
f (x) = p2 xp − q 2 xq − r2 . (1)

Observe that   q 4 
2 2
f 0 (x) = p4 xq −1
xr − .
p

4
The roots of equation f 0 (x) = 0 are x1 = 0 and x2 = pq r2 (r 6= 0 and q 6= 1). From f (0) = −r2 < 0 and

4

f pq r2 < 0 we obtain

−1 if x < 1 ,

sgn f (x) = 0 if x = 1 , (2)

1 if x > 1 .


So x = 1 is the only real root of equation f (x) = 0.

Since the matrix A verifies f (A) = On , some eigenvalue λ ∈ σP (A) satisfies the equation f (λ) = 0. Let
λ1 , λ2 , . . . , λn be eigenvalues of the matrix A. We show that |λk | ≤ 1 for all k. The fact f (λ) = 0 can be written
as 2 2
p2 λp = q 2 λq + r2 . (3)
2 2
Passing the relation (3) at modulus we obtain p2 |λ|p ≤ q 2 |λ|q + r2 or

f (|λ|) ≤ 0 . (4)

From (2) and (4) we obtain 0 ≤ |λ| ≤ 1 or 0 ≤ |λk | ≤ 1 for all k = 1, . . . , n. Because f (0) = −r2 6= 0, it results
that λk 6= 0 for all k.
Hence
0 < |λk | ≤ 1 for all k = 1, . . . , n . (5)
From det A = λ1 λ2 · · · λn we obtain

|det A| = |λ1 λ2 · · · λn | = |λ1 ||λ2 | · · · |λn | ≤ 1 . (6)

From (5) and (6) we obtain

0 < |det A| ≤ 1 . (7)
Since A ∈ Mn (Z), it follows that |det A| ∈ N. From (7) we obtain the conclusion that |det A| = 1. 
 The 19th Annual Vojtěch Jarnı́k
International Mathematical Competition
Ostrava, 1st April 2009
Category II

Problem 4 Let k, m, n be positive integers such that 1 ≤ m ≤ n and denote S = {1, 2, . . . , n}. Suppose that
A1 , A2 , . . . , Ak are m-element subsets of S with the following property: for every i = 1, 2, . . . , k there exists a
partition S = S1,i ∪ S2,i ∪ · · · ∪ Sm,i (into pairwise disjoint subsets) such that
(i) Ai has precisely one element in common with each member of the above partition.

(ii) Every Aj , j 6= i is disjoint from at least one member of the above partition.
n−1


Show that k ≤ m−1 . [10 points]
(i)
Solution Without loss of generality assume that 1 ∈ S1 for all i = 1, 2, . . . , k, because otherwise we simply
rename members of each partition.
For every i = 1, 2, . . . , k define the polynomial
m  X
Y 
Pi (x2 , x3 , . . . , xn ) = xs
l=2 s∈Sl
(i)

and regard it as a polynomial over R in variables x2 , x3 , . . . , xn .

Observe that Pi is a homogenous polynomial of degree m − 1 in n − 1 variables. Also observe that all
monomials in Pi are products of different x’s, i.e. there are no monomials with squares or higher powers. The
(i) (i)
last statement follows simply from the
fact that S2 , . . . , Sm are mutually disjoint. Such polynomials form a
n−1
linear space over R of dimension m−1 and polynomials Pi belong to that space. If we prove that polynomials
n−1


Pi , i = 1, 2, . . . , k are linearly independent, the inequality k ≤ m−1 will follow from the dimension argument.
For any i = 1, 2, . . . , k let χi be the characteristic vector of A ∩ {2, 3, . . . , n}. In other words, χi ∈ {0, 1}n−1
where the j-th coordinate of χi equals 1 if j + 1 ∈ A, and 0 otherwise.
(i)
For every i we know that each Ai ∩ Sl has exactly one element and therefore
m m
(i)
Y Y
Pi (χi ) = |Ai ∩ Sl | = 1 = 1.
l=2 l=2

(i) (i)
On the other hand, if j 6= i then either some Aj ∩ Sl , l ≥ 2 is empty, or all Aj ∩ Sl , l ≥ 2 are nonempty but
(i) (i)
Aj ∩ S1 = ∅. In the latter case we must have |Aj ∩ Sl | = 2 for some l ≥ 2. In any case we have at least one
even factor in the following product, and so
m
(i)
Y
Pi (χj ) = |Aj ∩ Sl | ≡ 0 (mod 2) .
l=2

Therefore all diagonal entries in the matrix [Pi (χj )]i,j=1,2,...,k are odd, while all non-diagonal entries are
even. Consequently, its determinant is an odd integer, in particular it is not 0, and thus the matrix is regular.
If polynomials Pi were linearly dependent, we would conclude that rows of [Pi (χj )]i,j=1,2,...,k are also linearly
dependent, but this is not the case. Therefore Pi , i = 1, 2, . . . , k must be linearly independent and this completes
the proof. 
 The 20th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th March 2010
Category I

Problem 1

a) Is it true that for every bijection f : N → N the series

∞
X 1
n=1
nf (n)

is convergent?

b) Prove that there exists a bijection f : N → N such that the series

∞
X 1
n=1
n + f (n)

is convergent.

(N is the set of all positive integers.) [10 points]

Solution a) Yes. Applying the inequality, if 0 ≤ a1 ≤ · · · ≤ an and 0 ≤ b1 ≤ · · · ≤ bn and σ : {1, . . . , n} →
{1, . . . , n} is a permutation, then
Xn Xn
aj bσ(j) ≤ aj bj ,
j=1 j=1

for every n we get

n n ∞
X 1 X 1 X 1
≤ ≤ .
j=1
jf (j) j=1 j 2 j=1
j2
Pn 1


Since the sequence is increasing and bounded, it converges.
j=1 jf (j)
b) No. We will construct a permutation f : N → N such that the series
∞
X 1
n=1
n + f (n)

is convergent. Let f : N → N be given in the following way: f (1) = 4 and for [(n!)2 + 1, ((n + 1)!)2 ] ∩ N we put
n−1
X
f ((n!)2 + k) = [(n + 2)!]2 − (k − 1) if 1 ≤ k < [(n + 1)!]2 − 1 − (−1)j [(n − j)!]2 .
j=0

and
n−1
X
f ([(n + 1)!]2 − k) = [(n − 1)!]2 + k + 1 if 0 ≤ k ≤ 1 + (−1)j [(n − j)!]2 .
j=0

Then
[(n+1)!]2
X 1 ((n + 1)!)2 − (n!)2 (n!)2 − [(n − 1)!]2
≤ +
n + f (n) (n!)2 + [(n + 2)!]2 + 1 [(n + 1)!]2 + [(n − 1)!]2 + 1
j=(n!)2 +1
1 1
<+ .
(n + 2)2 (n + 1)2
Pn 1


Thus we show that the sequence j=1 j+f (j) is bounded. Since it is increasing, it converges. 
 The 20th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th March 2010
Category I

Problem 2 Let A and B be two complex 2 × 2 matrices such that AB − BA = B 2 . Prove that AB = BA.
[10 points]
Solution We may concludethat  AB = BA  only if 2 6= 0 in F (that is, char F 6= 2).
 if and
1 1 0 0
If char F = 2, take B = ,A= .
0 1 1 0
   2 
a b 2 a + bc b(a + d)
Assume that char F 6= 2. Let B = , then B = . We have a2 + d2 + 2bc =
c d c(a + d) d2 + bc
trace B 2 = trace AB − trace BA = 0. If B is invertible, then A = B(A + B)B −1 , hence

trace A = trace(B(A + B)B −1 ) = trace(A + B) = trace A + trace B,

so trace B = 0, d = −a, trace B 2 = 2(a2 + bc) = 0. Since char F 6= 2, it implies a2 + bc = 0, hence B 2 = 0 and
AB = BA. If B is not invertible, then det B = ad − bc = 0, so (a + d)2 = a2 + d2 + 2bc = 0, a + d = 0, a = −d,
a2 + bc = −ad + bc = 0, so B 2 = 0. 
 The 20th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th March 2010
Category I

Problem 3 Prove that there exist positive constants c1 and c2 with the following properties:
a) For all real k > 1,
Z 1 p  c1
1 − x2 cos(kx) dx < 3/2 .
 

0 k
b) For all real k > 1,
Z 1 p  c
2
1 − x2 sin(kx) dx > .
 
k

0
[10 points]
√
Solution Put f (x) = 1 − x2 .
1. Integrating by parts, we have
Z 1 i1 Z 1
h 1 1
f (x) · cos kx dx = f (x) · sin kx − f 0 (x) · sin kx dx .
0 k 0 0 k
The first term is 0 − 0 = 0. The second term is (−1/k) times
Z √1−1/k Z 1
0
f (x) · sin kx dx + √ f 0 (x) · sin kx dx . (1)
0 1−1/k

Here the first term equals

√
h 1 i√1−1/k Z 1−1/k
1
0
−f (x) · cos kx + f 00 (x) · cos kx dx ,
k 0 0 k
whose absolute value is p
2 p
2 1 − 1/k 2
≤ − f0 1 − 1/k = p <√ .
k k 1/k k
The absolute value of the second term in (1) is
Z 1
1
≤ √ |f 0 (x)| dx = −[f (x)]1√ =√ .
1−1/k k
1−1/k

Thus, we may choose c1 = 2 + 1 = 3.

2. Integrating by parts, we have
Z 1 i1 Z 1
h 1 1
f (x) · sin kx dx = − f (x) · cos kx + f 0 (x) · cos kx dx .
0 k 0 0 k
The first term is 1/k. The second term is (1/k) times
Z √ 1−1/k Z 1
f 0 (x) · cos kx dx + √ f 0 (x) · cos kx dx . (2)
0 1−1/k

Here the first term equals

√
h 1 i√1−1/k Z 1−1/k
1
0
f (x) · sin kx − f 00 (x) · sin kx dx ,
k 0 0 k
whose absolute value is p
2 p
2 1 − 1/k 2
≤ − f0 1 − 1/k = p <√ .
k k 1/k k
The absolute value of the second term in (2) is
Z 1
1
≤ √ |f 0 (x)| dx = −[f (x)]1√ =√ .
1−1/k k
1−1/k
Thus, Z 1  
1 3
f (x) · sin kx dx > 1− √ .
0 k k
This proves the desired claim for k ≥ 3π.
The integral has a positive lower bound for k < 3π as well, since
Z 1 Z 1

1 − cos kx
f (x) · sin kx dx = −f 0 (x) · dx > 0 .
0 0 k


 The 20th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th March 2010
Category I

Problem 4 For every positive integer n let σ(n) denote the sum of all its positive divisors. A number n is
called weird if σ(n) ≥ 2n and there exists no representation

n = d1 + d2 + · · · + dr ,

where r > 1 and d1 , . . . , dr are pairwise distinct positive divisors of n.

Prove that there are infinitely many weird numbers. [10 points]
Solution The idea is to show that given a weird number, one can construct a sequence of weird numbers
tending to infinity.
We claim that for weird n and p a prime greater than σ(n) and coprime to n, the number pn is also weird.
In fact, if 1 = d1 , d2 , . . . , dk = n are the positive divisors of n, the ones of pn are d1 , d2 , . . . , dk , pd1 , . . . , pdk and
they are pairwise distinct as (p, n) = 1. Suppose now that we have

pn = di1 + · · · + dir + p(dj1 + · · · + djs )

with ik , jl ∈ {1, . . . , k}. Then we have

di1 + · · · + dir = p(n − dj1 − · · · − djs ) .

Note that n ∈/ {dj1 , . . . , djs } as the representation must have more than only one summand and the assumption
that n is weird implies n − dj1 − . . . − djs 6= 0. Hence as the right hand expression is divisible by p and non
zero, so must be di1 + · · · + dir which is impossible as p > σ(n).
It remains to find a weird number. A possible reasoning could be: look for a number n with σ(n) = 2n + 4
that is not divisible by 3 and 4. Then the smallest possible divisors are 1, 2, 5 so that it will be impossible to
represent 4, and hence n, as a sum of pairwise distinct divisors of n. Checking for numbers with three distinct
prime factors 2, p, q yields
σ(2pq) = 3(p + 1)(q + 1) = 3pq + 3p + 3q + 3
and hence we need
3pq + 3p + 3q + 3 = 4pq + 4 ⇐⇒ (p − 3)(q − 3) = 8 .
This equality is solved by p = 5 and q = 7 which yields the weird number n = 70. 
 The 20th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th March 2010
Category II

Problem 1 Let a and b be given positive coprime integers. Then for every integer n there exist integers x, y
such that
n = ax + by .
Prove that n = ab is the greatest integer for which xy ≤ 0 in all such representations of n. [10 points]
Solution The greatest such integer is a · b.
If ab = ax + by, then a | y and b | x. Thus if x > 0, then x ≥ b and by = ab − ax ≤ ab − ab = 0, so y ≤ 0.
Now let n > ab. Let n = ax + by be the representation such that x is positive and as small as possible. Then
since n = a(x − b) + b(y + a) is another representation of n, x − b must not be positive and therefore x ≤ b.
Hence by = n − ax ≥ n − ab > 0, so y > 0. 
 The 20th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th March 2010
Category II

Problem 2 Prove or disprove that if a real sequence (an ) satisfies an+1 − an → 0 and a2n − 2an → 0 as n → ∞,
then an → 0. [10 points]
Solution The proposition is true.
From the condition an+1 −an → 0 we conclude by Cesaro’s lemma that ann → 0. Since the sequence a2n −2an
must be bounded, we know that
C := sup{|a2n − 2an | : n ∈ N} < ∞ .
Considering the identity
m
an a m+1 X an·2k an·2k+1 
− n·2m+1 = −
n n·2 n · 2k n · 2k+1
k=0

we conclude by letting m → ∞ and n fixed that

∞
an X an·2k a k+1 
= k
− n·2k+1 .
n n·2 n·2
k=0

Now from
∞ 
a  X ∞
 n  an·2k an·2k+1  X C C
 ≤ − ≤ =
n n · 2k n · 2k+1 n · 2k+1 n
 
k=0 k=0

we infer that |an | ≤ C, i. e. the sequence (an ) must be bounded.

Now suppose that (an ) does not converge to 0. Then, by Bolzano’s theorem, there must exist a subsequence
(ank ) converging to some number a 6= 0. From the hypothesis we conclude in turn that

a2nk → 2a ,
a4nk → 4a ,
..
.

which would result in an unbounded set of accumulation points a, 2a, 4a, . . . of (an ) in contradiction to (an )
being bounded. 
 The 20th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th March 2010
Category II

Problem 3 Let A and B be two n × n matrices with integer entries such that all of the matrices

A, A+B, A + 2B , A + 3B , ... , A + (2n)B

are invertible and their inverses have integer entries, too. Show that A + (2n + 1)B is also invertible and that
its inverse has integer entries. [10 points]

Solution Suppose that the n × n matrix M has integer entries and M has inverse matrix M −1 with integer
entries. Then M · M −1 = I implies det M · det M −1 = 1. Thus det M = 1 or det M = −1. Set M (t) = A + tB.
The determinant of the matrix M (t)

det M (t) = det (A + tB) = det A + · · · + tn det B

is the polynomial of degree n in t. The polynomial det M (t) takes values 1 or −1 at points t = 0, 1, 2, . . . , 2n.
Hence det M (t) takes the value 1 or the value −1 at least n + 1 times. This implies that det M (t) is a
constant polynomial: M (t) = 1 or M (t) = −1 for all t. Consequently, det M (2n + 1) = ±1. Hence the matrix
A + (2n + 1)B is invertible. By Cramer’s formula, the inverse matrix has integer entries, since the determinant
is equal to 1 or −1. 
 The 20th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 25th March 2010
Category II

Problem 4 Let f : [0, 1] → R be a function satisfying

|f (x) − f (y)| ≤ |x − y|
for every x, y ∈ [0, 1]. Show that for every ε > 0 there exists a countable family of rectangles (Ri ) of dimensions
ai × bi , ai ≤ bi , in the plane such that
 [ X
(x, f (x)) : x ∈ [0, 1] ⊂ Ri and ai < ε .
i i

(The edges of the rectangles are not necessarily parallel to the coordinate axes.) [10 points]
Solution Assume without loss of generality that f (0) = 0, thus |f (x)| ≤ 1 for x ∈ [0, 1].
First notice that if C ⊂ [0, 1] is a set of Lebesgue measure no larger than ε/3, then it can be covered by a
countable family of intervals Ii of total measure at most ε/2, and thus {(x, f (x) : x ∈ C} is covered by rectangles
Ii × [−1, 1], and their total width is at most ε/2.
Notice that as we are interested in only one dimension of the rectangle, and the graph we are to covered is
bounded, we may as well think in terms of covering with strips instead of rectangles.
For now on fix ε > 0. We shall introduce a few definitions. Let x, y ∈ [0, 1]. We say that the interval [x, y]
is covered, if |f (z) − α(z)| < ε|x − y| for all z ∈ [x, y], where α is the linear function meeting f at x and y. The
inclination of an interval [x, y], denoted i(x, y), is the number |f (x) − f (y)|/|x − y|. Notice the inclination of
any interval cannot be larger than 1 as f is 1-Lipschitz.
Now we prove the following lemma.
Lemma There exists a constant δ > 0 such that the following holds. Consider any interval [x, y] ⊂ [0, 1]. Then
either [x, y] is covered, or there exists a subinterval [x0 , y 0 ] ⊂ [x, y] of length |y 0 − x0 | > δ|x − y| and inclination
at least i(x, y) + ε.
Proof The proof is pretty simple. If [x, y] is not covered, then there exists a point z ∈ [x, y] with |f (z)−α(z)| >
ε|x − y|. Without loss of generality assume f (x) < f (y) and f (z) − α(z) > ε|x − y|. The interval [x, z] in this
case has inclination
f (y)−f (x)
f (z) − f (x) α(z) + ε(y − x) − f (x) y−x (z − x) + ε(y − x)
i(x, z) = |f (x) − f (z)|/|x − z| = ≥ =
z−x z−x z−x
f (y) − f (x) y−x
= +ε ≥ i(x, y) + ε.
x−y z−x
The cases of f (x) > f (y) and (or) f (z) − α(z) < −ε|x − y| are similar. Moreover we have
f (z) > α(z) + ε|x − y| = f (x) ± i(x, y)(z − x) + ε(y − x) .
Thus
2|z − x| ≥ |f (z) − f (x)| + i(x, y)|z − x| ≥ f (z) − f (x) ± i(x, y)(z − x) ≥ ε|x − y| ,
thus |z − x| ≥ ε|x−y|
2 , which finishes the proof of the lemma with δ = ε/2. 

Take a constant n > 1/ε. If begin with an interval [x, y] and apply the lemma n times, we end up with
an interval of length at least |x − y|δ n , which is either covered, or has inclination at least nε — the second
is impossible, however, as the inclination of any interval is at most 1. Thus for any interval we can find its
subinterval of length at least δ n times the length of the original, which is covered. Thus we have the following
corollary: for any interval [x, y] ⊂ [0, 1] there exists a covered subinterval [x0 , y 0 ] of [x, y] of length at least
c|x − y| for some fixed constant c.
Now we are ready to solve the S problem. We shall construct a family of disjoint intervals Ci ⊂ [0, 1], with
the Lebesgue measure of [0, 1] \ Ci no larger than ε. Each of these intervals will be covered, and thus we
shall be able to cover the whole graph of f by rectangles — each interval is covered, and thus the appropriate
piece of the graph is contained in a rectangle of width at most 2ε, while the remaining part can be covered by
a countable family of vertical rectangles of total width at most 2ε. As ε was arbitrary, this will end the proof.
The construction of Ci s follows directly from the corollary — we choose C0 = [x0 , y0 ] to be the interval given
by the corollary for [0, 1], then C1 and C2 the intervals for [0, x0 ] and [y0 , 1] respectively, then (in the third
step), C3 , C4 , C5 and C6 are given for [0, x1 ], [y1 , x0 ], [y0 , x2 ] and [y2 , 1] respectively, and so on. In each step a
constant fraction of measure is removed, thus after sufficiently many steps no more than ε measure remains. 
 The 21st Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2011
Category I

Problem 1
(a) Is there a polynomial P (x) with real coefficients such that
1 k+2
P = ,
k k
for all positive integers k?

(b) Is there a polynomial P (x) with real coefficients such that

1 1
P = ,
k 2k + 1
for all positive integers k?
Solution (a) YES. It suffices to define a polynomial W (x) as follows

W (x) = 2x + 1.

(b) NO. Suppose that such a polynomial W (x) exists. Define a polynomial F (x) as follows

F (x) = (x + 2)W (x) − x.

Then 1  1 1

1
F +2 W
= − = 0,
k k k k
for all k ∈ N. Hence, the polynomial F (x) admits infinitely many zeros. Consequently,

(x + 2)W (x) − x = 0,

for all x ∈ R. But this implies that

x
W (x) = ,
x+2
for all x ∈ R – a contradiction. 
 The 21st Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2011
Category I

Problem 2 Let (an )∞n=1 be unbounded and strictly increasing sequence of positive reals such that the arithmetic
mean of any four consecutive terms an , an+1 , an+2 , an+3 belongs to the same sequence. Prove that the sequence
an+1 /an converges and find all possible values of its limit.
Solution Since an < an+1 < an+2 < an+3 , one has
1
an < (an + an+1 + an+2 + an+3 ) < an+3 ,
4
thus (an + an+1 + an+2 + an+3 )/4 ∈ {an+1 , an+2 }. Hence for any n ∈ N precisely one of the two identities

an + an+1 + an+2 + an+3 = 4an+1 (1)

or
an + an+1 + an+2 + an+3 = 4an+2 (2)
holds. Let A be the set of indices n ∈ N for which (1) holds and let B be the set of indices n ∈ N for which (2)
holds. Clearly, A ∪ B = N, A ∩ B = ∅. We shall prove that one of A or B is finite. Indeed, suppose the contrary,
that both A and B are infinite. Since A and B partition N, there exists a positive integer k, such that k ∈ B,
k + 1 ∈ A. From (1) and (2), it follows that

ak + ak+1 + ak+2 + ak+3 = 4ak+2 and ak+1 + ak+2 + ak+3 + ak+4 = 4ak+2 .

Hence ak = ak+4 , which contradicts the fact that an is strictly increasing. We now consider two cases.
Case 1) The set A is infinite, the set B is finite. By (1), the sequence an satisfies a linear recurrence
an − 3an+1 + an+2 + an+3 = 0 for all n > n0 . The characteristic polynomial of the linear recurrence

φ(λ) = λ3 + λ2 − 3λ + 1 = (λ − 1)(λ2 + 2λ − 1)
√ √
has roots λ1 = 1, λ2 = −1 − 2, λ3 = −1 + 2. Hence
√ √
an = C1 + C2 (−1 − 2)n + C3 (−1 + 2)n , C1 , C2 , C3 ∈ R, n > n0 .

Observe that λ2 < −1, 0 < λ3 < 1. If C2 6= 0, then limn→∞ |an | = ∞ and an alternates in sign for n sufficiently
large which contradicts the monotonicity property. If C2 = 0, then the sequence an is bounded, which leads to
the contradiction again. Thus we reject the case one.
Case 2) The set A is finite, the set B is infinite. By (1), the sequence an satisfies a linear recurrence
an + an+1 − 3an+2 + an+3 = 0 for all n > n0 . The characteristic polynomial of the linear recurrence

φ(λ) = λ3 − 3λ2 + λ + 1 = (λ − 1)(λ2 − 2λ − 1)

√ √
has roots λ1 = 1, λ2 = 1 − 2, λ3 = 1 + 2. Hence
√ √
an = C1 + C2 (1 − 2)n + C3 (1 + 2)n , C1 , C2 , C3 ∈ R, n > n0 .

Note that −1 < λ2 < 0, λ3 > 1. If C3 ≤ 0, then the sequence an is bounded from above. Hence C3 > 0 so
an ∼ C3 λn3 as n → ∞ . The standard limit calculation now shows that bn converges and has limit value
an+1 √
lim bn = lim = λ3 = 1 + 2.
n→∞ n→∞ an


 The 21st Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2011
Category I

Problem 3 Prove that

∞ ∞
X
k 1 + x2k+2 X
k xk
x = (−1)
(1 − x2k+2 )2 (1 − xk+1 )2
k=0 k=0

for all x ∈ (−1, 1).

Solution We use the binomial series
∞
1 X
2
= (j + 1)uj , |u| < 1
(1 − u) j=0

to get
∞ ∞ ∞ ∞ X ∞
X 1 + x2k+2 X X X
xk = x k
(1 + x 2k+2
) (j + 1)x j(2k+2)
= xk (1 + x2k+2 )(j + 1)xj(2k+2) =
(1 − x2k+2 )2 j=0 j=0 k=0
k=0 k=0
∞ ∞ ∞
x2
 
X
2j
X
k 2k+2 j2k
X
2j 1
= (j + 1)x x (1 + x )x = (j + 1)x + =
j=0 j=0
1 − x2j+1 1 − x2j+3
k=0
∞ ∞ ∞ ∞
X (j + 1)x2j X jx2j X (2j + 1)x2j d X
= + = =− log(1 − x2j+1 )
j=0
1 − x2j+1 j=1
1−x 2j+1
j=0
1 − x 2j+1 dx j=0

and
∞ ∞ ∞ ∞ ∞ ∞
X (−x)k X
k
X
(k+1)j
X
j
X
k kj
X (j + 1)xj
= (−x) (j + 1)x = (j + 1)x (−x) x = =
(1 − xk+1 )2 j=0 j=0 j=0
1 + xj+1
k=0 k=0 k=0
∞
d X
= log(1 + xj+1 ).
dx j=0

The proposition now follows by logarithmic differentiation of the classical identity

∞ ∞
Y 1 Y
2n+1
= (1 + xn ),
n=0
1 − x n=1

which can be proved as follows:

∞ ∞ Q∞ Q∞ ∞
Y Y
n 1 − x2n n=1 (1 − x2n ) n=1 (1 − x2n ) Y 1
(1 + x ) = n
= Q∞ n)
= Q∞ 2n )
Q∞ 2n−1 )
= 2n−1
.
n=1 n=1
1 − x n=1 (1 − x n=1 (1 − x n=1 (1 − x n=1
1 − x


 The 21st Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2011
Category I

Problem 4 Let a, b, c be elements of finite order in some group. Prove that if a−1 ba = b2 , b−2 cb2 = c2 and
c−3 ac3 = a2 , then a = b = c = e, where e is the unit element.
Solution Let r(g) denote the rank of g ∈ G. Assume that the assertion does not hold. Let p be the smallest
prime number dividing r(a)r(b)r(c). Without loss of generality we can assume that p | r(b) (if p | r(a) or
p | r(c), then the reasoning is the same). Then there exists k such that r(b) = pk. Let d := bk . Then r(d) = p.
m
Lemma For any m ∈ N, a−m dam = d2 .
Proof First we prove that
a−1 da = d2 .
Indeed, multiplying the equation a−1 ba = b2 k-times with itself we get

(a−1 ba)(a−1 ba) · · · (a−1 ba) = b2 b2 · · · b2 ;

and hence
a−1 bk a = (b2 )k = (bk )2 .
Now, the assertion of the above lemma follows from the following calculations:
2 2 3 m
d = ad2 a−1 = a(ad2 a−1 )2 a−1 = a2 d2 a−2 = a2 (ad2 a−1 )2 a−2 = a3 d2 a−3 = · · · = am d2 a−m . (1)

Observe that Fermat’s little theorem implies that 2p ≡ 2 (mod p). Consequently,
p
a−p dap = d2 = d2 = a−1 da. (2)

Since gcd(r(a), p − 1) = 1, there exist integers r and s such that

r · r(a) + s · (p − 1) = 1. (3)

From (2) we get

a−l(p−1) dal(p−1) = d,
for all l ∈ Z (see the calculations in (1)). Finally, putting l := s, we obtain
(3)
d = a−s(p−1) das(p−1) = arr(a)−1 da−rr(a)+1 = a−1 da = d2 ,

which implies that d = e, a contradiction. 

 The 21st Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2011
Category II

Problem 1 Let n > k and let A1 , . . . , Ak be real n × n matrices of rank n − 1. Prove that

A1 · . . . · Ak 6= 0 .
g f
Solution Consider two linear operators V → V → V of an n-dimensional vector space V. If Ker(f ) ⊂ Im(g),
then dim (Im (f g)) = dim (Im(g)) − dim (Ker(f )) . But we have the inequality

dim (Im(f g)) ≥ dim (Im(g)) − dim (Ker(f ))

in the general case. Applying the correspondence between linear operators and matrices, we obtain the inequality
rank (AB) ≥ rank B − (n − rank A) for every two matrices A and B. The inequality rank (A1 · . . . · Ak ) ≥
(rank (A1 ) + . . . + rank (Ak )) − (k − 1) n can be deduced from the inequality rank (AB) ≥ rank A + rank B − n
by the simple induction. We obtain the inequality rank (A1 · . . . · Ak ) ≥ k (n − 1) − (k − 1) n = n − k in our
case. Thus, if k < n then rank (A1 · . . . · Ak ) ≥ 1 and the product A1 · . . . · Ak can not be equal to zero. 
 The 21st Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2011
Category II

Problem 2 Let k be a positive integer. Compute

∞ X
∞ ∞
X X 1
··· .
n1 =1 n2 =1 nk
n n
=1 1 2
. . . nk (n 1 + . . . + nk + 1)

Solution
∞ X
∞ ∞ ∞ X ∞ ∞ Z 1
X X 1 X X 1
··· = ··· xn1 +...+nk dx =
n1 =1 n2 =1 nk =1
n 1 n 2 . . . nk (n1 + . . . + n k + 1) n1 =1 n2 =1 nk =1
n 1 n 2 . . . n k 0
Z 1 X ∞ X ∞ ∞ n1 +...+nk Z 1 Z ∞
X x −u
= ··· dx = k
(− log(1−x)) dx = [1−x = e ] = uk e−u du = Γ(k+1) = k!
0 n =1 n =1 n n . . . nk
n =1 1 2 0 0
1 2 k


 The 21st Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2011
Category II

p(z)
Problem 3 Let p and q be complex polynomials with deg p > deg q and let f (z) = . Suppose that all
q(z)
roots of p lie inside the unit circle |z| = 1 and that all roots of q lie outside the unit circle. Prove that

deg p − deg q
max |f 0 (z)| > max |f (z)|.
|z|=1 2 |z|=1

Solution Without loss of generality we can assume that the maximum of |f | is attained at 1.
n1
Q n2
Q
Let p(z) = a (z − ck ) and q(z) = b (z − d` ) where n1 = deg p and n2 = deg q. Then
k=1 `=1

n n
f 0 (z) X1
1 X2
1
= − .
f (z) z − ck z − d`
k=1 `=1

Since |ck | < 1 and |d` | > 1 for all k and `, we have
1 1
Re >
1 − ck 2
and
1 1
Re < .
1 − dk 2
Therefore,
|f 0 (1)| f 0 (1) 1 1 deg p − deg q
≥ Re > n1 · − n2 · =
|f (1)| f (1) 2 2 2
and
|f 0 (1)| deg p − deg q
max |f 0 (z)| ≥ |f 0 (1)| = · |f (1)| ≥ max |f (z)| .
|z|=1 |f (1)| 2 |z|=1


 The 21st Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 31st March 2011
Category II

Problem 4 Let Q[x] denote the vector space over Q of polynomials with rational coefficients in one variable
x. Find all Q-linear maps Φ : Q[x] → Q[x] such that for any irreducible polynomial p ∈ Q[x] the polynomial
Φ(p) is also irreducible.
(A polynomial p ∈ Q[x] is called irreducible if it is non-constant and the equality p = q1 q2 is impossible for
non-constant polynomials q1 , q2 ∈ Q[x].)
Solution
The answer is Φ(p(x)) = ap(bx + c) for some non-zero rationals a, b and some rational c. It is clear that
such operators preserve irreducibility. Let’s prove that any irreducibility-preserving operator is of such form.
We start with the following
Lemma 1 Assume that f, g ∈ Π are two polynomials such that for all rational numbers c the polynomial f + cg
is irreducible. Then either g ≡ 0, or f is non-constant linear polynomial and g is non-zero constant.
Proof Let g(x0 ) 6= 0 for some rational x0 . Then for c = −f (x0 )/g(x0 ) we have (f + cg)(x0 ) = 0, so the
polynomial f + cg is divisible by x − x0 . Hence f + cg = C(x − x0 ) for some non-zero rational C. Choose
x1 6= x0 such that g(x1 ) 6= 0. Then for c1 = −f (x1 )/g(x1 ) 6= c (since f (x1 ) + cg(x1 ) = C(x1 − x0 ) 6= 0) we have
f + c1 g = C1 (x − x1 ). Subtracting we get that (c1 − c)g is linear, hence g is linear, hence f too. If f (x) = ax + b,
g(x) = a1 x + b1 , then a 6= 0 (since f is irreducible) and if a1 6= 0, then for c = −a/a1 the polynomial f + cg is
constant, hence not irreducible. So a1 = 0 and we are done. 

Now denote gk = Φ(xk ).

Lemma 2 g0 is non-zero constant and g1 is non-constant linear function.
Proof Since x+c is irreducible for any rational c, we get that g1 +cg0 is irreducible for any rational c. By Lemma
1 we have that either g0 = 0 or g0 is constant and g1 is linear non-constant. Assume that g0 = 0. Note that for
any rational α one may find rational β such that x2 + αx + β is irreducible, hence g2 + αg1 = Φ(x2 + αx + β) is
irreducible for any rational α. It follows by Lemma 1 that g1 is constant, hence not irreducible. A contradiction,
hence g0 6= 0 and we are done. 

Denote g0 = C, g1 (x) = Ax + B. Consider the new operator p(x) → C −1 Φ(p(A−1 Cx − A−1 B)). This
operator of course preserves irreducibility, consider it instead Φ.
Now g0 = 1, g1 (x) = x and our goal is to prove that gn = xn for all positive integers n. We use induction by
n. Assume that n ≥ 2 and gk (x) = xk is already proved for k = 0, 1, . . . , n − 1. Denote h(x) = gn (x) − xn and
assume that h is not identical 0. For arbitrary monic irreducible polynomial f of degree n we have Φ(f ) = f +h,
hence f + h is irreducible aswell. Choose rational x0 such that h(x0 ) 6= 0, our goal is to find irreducible f such
that f (x0 ) = −h(x0 ) and hence f + h has a root in x0 .
There are many ways to do it, consider one of them, via Eisenstein’s criterion. Recall it.
Eisenstein’s criterion Assume that f (x) = an xn + · · · + a0 is a polynomial with rational coefficients and p is a
prime number so that ak = bk /ck with coprime integers bk , ck such that bk is divisible by p for k = 0, 1, . . . , n−1,
both bn and cn are not divisible by p and b0 is not divisible by p2 . Then f is irreducible.
Without loss of generality, x0 = 0 (else denote x − x0 by new variable). Then we want to find an irreducible
polynomial f (x) = xn + an−1 xn−1 + · · · + a1 x − h(0). Denote −h(0) = u/v for coprime positive integer v and
non-zero integer u. Take L = 6uv and consider the prime divisor p of the number vLn /u − 1. Clearly, p does
not divide 6uvL. Then consider the polynomial (x + L)n − Ln + u/v. If vLn /u − 1 is not divisible by p2 , then
we are done by Eisenstein’s criterion (with new variable y = x + L). If vLn /u − 1 is divisible by p2 , then add
px to our polynomial and now Eisenstein’s criterion works.
Unless h(x) = −xn + . . . , the polynomial f + h is not linear and so is not irreducible. If n ≥ 3, then we may
add px2 or 2px2 to our polynomial f and get non-linear f + h (but still irreducible f ). Finally, if n = 2, and
h(x) = −x2 + ax + b, then choose irreducible polynomial of the form f (x) = x2 − ax + c and get f + h being
constant (hence not irreducible).
The induction step and the whole proof are finished. 
 The 22nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 30th March 2012
Category I

Problem 1 Let f : [0, 1] → [0, 1] be a differentiable function such that |f 0 (x)| 6= 1 for all x ∈ [0, 1]. Prove that
there exist unique points α, β ∈ [0, 1] such that f (α) = α and f (β) = 1 − β.
Solution Existence: Since f is derivable in [0, 1], then f is continuous in [0, 1]. Considering the functions
g(x) = f (x) − x and h(x) = f (x) − (1 − x) that are continuous in [0, 1] and applying Bolzano’s theorem we get
that exists α ∈ [0, 1] such that g(α) = 0 and β ∈ [0, 1] with h(β) = 0. That is, there exist α, β ∈ [0, 1] for which
f (α) = α and f (β) = 1 − β.
Uniqueness: Suppose that there exist α, α0 ∈ [0, 1], α < α0 ) such that f (α) = α and f (α0 ) = α0 . On account
of Lagrange’s theorem, there exists θ ∈ (α, α0 ) ⊂ [0, 1] such that

f (α0 ) − f (α) α0 − α
f 0 (θ) = = =1
α0 − α α0 − α
contradiction. Likewise, if we assume that there exist β, β 0 ∈ [0, 1], ( β < β 0 ) such that f (β) = 1 − β and
f (β 0 ) = 1 − β 0 . On account of Lagrange’s theorem, there exists θ0 ∈ (β, β 0 ) ⊂ [0, 1] such that

f (β 0 ) − f (β) (1 − β 0 ) − (1 − β)
f 0 (θ0 ) = 0
= = −1
β −β β0 − β
contradiction. This completes the proof. 
 The 22nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 30th March 2012
Category I

Problem 2 Determine all 2 × 2 integer matrices A having the following properties:

1. the entries of A are (positive) prime numbers,
2. there exists a 2 × 2 integer matrix B such that A = B 2 and the determinant of B is the square of a prime
number.
Solution Let  
p1 p2
A= = B2,
p3 p4
and d = det(B) = q 2 with p1 , p2 , p3 , p4 , q ∈ P; here P denotes the set of positive prime numbers.
By Cayley-Hamilton Theorem,
B 2 = tr(B)B − det(B)E,
where E is the 2 × 2 identity matrix. Without loss of generality, we assume that tr(B) ≥ 0, otherwise, replace
B by −B. The equality  
2 p1 + d p2
tr(B)B = B + dE = A + dE =
p3 p4 + d
implies that tr(B) divides the numbers p2 and p3 . Moreover,

(tr(B))2 = tr(tr(B)B) = p1 + p4 + 2d ≥ 2 + 2 + 8 = 12 =⇒ tr(B) > 3.

It follows that    
1 p1 + d p2 a 1
tr(B) = p2 = p3 , and B = =
tr(B) p3 p4 + d 1 b
for some positive integers a and b. Hence,

a2 + 1
 
a+b
A = B2 = .
a+b b2 + 1

The numbers a2 + 1, b2 + 1, a + b cannot all be odd, thus, one of them equals 2. Since ab = d + 1 = q 2 + 1 ≥ 5
we have max(a, b) ≥ 3. Hence, a + b ≥ 3 + 1 > 2.
Now we assume that a2 + 1 ≤ b2 + 1. Then a2 + 1 = 2 and a = 1. Note that d = ab − 1 = b − 1 and
0 < p2 = a+b = b+1 = d+2 = q 2 +2. If q 6= 3 then q 2 ≡ 1 mod 3 =⇒ p2 ≡ 0 mod 3 =⇒ p2 = 3 =⇒ q 2 = 1,
which is impossible. Hence, q = 3, b = p2 − a = 32 + 2 − 1 = 10,
   
1 1 2 11
B= , and A = B 2 = .
1 10 11 101

Similarly, if a2 + 1 > b2 + 1 we obtain the matrix

 
101 11
A= .
11 2

Answer:    
2 11 101 11
A= , and A = .
11 101 11 2

 The 22nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 30th March 2012
Category I

Problem 3 Determine the smallest real number C such that the inequality
x 1 y 1 z 1
√ · +√ · +√ · ≤C
yz x + 1 zx y + 1 xy z + 1

holds for all positive real numbers x, y and z with

1 1 1
+ + = 1.
x+1 y+1 z+1
Solution In what follows we shall deal with the harder version of the problem only.
1. We consider the case x = y = t. Then
2 1
+ =1
t+1 z+1
that is
2
z= .
t−1
Thus the inequality under consideration becomes
2
2t 1 1
r · + √t−1 · 2 ≤C
2 t+1 t·t t−1+1
t·
t−1
that is √ √ √
2· t· t−1 2
+ ≤ C.
t+1 t(t + 1)
√
Letting here t → ∞ leads to C ≥ 2.

2. We are now going to prove that always

x 1 y 1 z 1 √
√ · +√ · +√ · < 2.
yz x + 1 zx y + 1 xy z + 1

In order to achieve this goal we make use of the following transformation

1 1 1
a= , b= , c= .
x+1 y+1 z+1

Then the three new variables satisfy a, b, c ∈ (0; 1) and are subject to the condition a + b + c = 1.
Furthermore
1−a 1−b 1−c
x= , y= , z=
a b c
that is (due to 1 − a = b + c, etc.)

b+c c+a a+b

x= , y= , z=
a b c
yield for the claimed inequality
√ √ √
(a + b) ab (b + c) bc
(c + a) ca √
p +p +p < 2.
(b + c)(c + a) (c + a)(a + b) (a + b)(b + c)

Upon clearing fractions this inequality becomes

p p p p
(a + b) ab(a + b) + (b + c) bc(b + c) + (c + a) ca(a + c) < 2(a + b)(b + c)(c + a).
We smuggle the condition 1 = a + b + c into the inequality and get
p p p p
(a + b) ab(a + b) + (b + c) bc(b + c) + (c + a) ca(a + c) < 2(a + b)(b + c)(c + a)(a + b + c).

Next, we deal with the right-hand expressions. For them we have

p p
(a + b)(b + c)(c + a) = ab(a + b) + bc(b + c) + ca(c + a) + 2abc

and √ p p
2(a + b + c) = 2(a + b + c)2 = (a + b)2 + (a + c)2 + (b + c)2 + 2(ab + bc + ca)
But, employing the Cauchy-Schwarz inequality yields for our inequality
p p p
(a + b) ab(a + b) + (b + c) bc(b + c) + (c + a) ca(a + c) ≤
p p
(a + b)2 + (b + c)2 + (c + a)2 · ab(a + b) + ac(a + c) + bc(b + c).

This together with the two previously stated equations completes the proof. It is also evident that there
cannot exist any triples (a, b, c), and thus also (x, y, z), yielding equality.

 The 22nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 30th March 2012
Category I

Problem 4 Find all positive integers n for which there exists a positive integer k such that the decimal
representation of nk starts and ends with the same digit.
Solution The number nk ends with zero whenever n is divisible by 10 and starts with nonzero digit. We show
that the claim is true for all other n’s.
It can be easily shown that all the numbers

n, n5 , n9 , . . . , n4m+1 , . . . (1)

ends with the same digit. In fact, n5 − n = n(n − 1)(n + 1)(n2 + 1) is even and for each possible reminder of n
modulo 5 there is a factor divisible by 5 in this product. Thus n5 − n is divisible by 10 and in the same fashion
we can show this for n9 − n5 , n13 − n9 , . . .
Now it suffices to show that for any nonzero digit c there is a number in the sequence (1) which starts with
c. For any nonnegative integer m put dm = n4m+1 /10l , where l is the greatest integer for which 10l ≤ n4m+1 .
Thus 1 ≤ dm < 10 and bdm c is the first digit of n4m+1 . Clearly all the dm ’s are different, since for m0 > m we
have 0 0 0
dm0 n4m +1 /10l n4(m −m)
= 4m+1 = 6= 1
dm n /10l 10l0 −l
(the numerator is not a power of 10 for n not divisible by 10).
The sequence (dm )∞ 2 ε
m=1 has the following property: If dm+i = dm · q, then dm+2i = dm · q · 10 , where
ε ∈ {−1, 0, 1}. This is true since when
0 00
dm = n4m+1 /10l , dm+i = n4(m+i)+1 /10l and dm+2i = n4(m+2i)+1 /10l ,
0
we have q = dm+i /dm = n4i /10l −l and so
00 0 00
−l
dm+2i /dm = n8i /10l = q 2 · 102l −l−l = q 2 · 10ε

for some integer ε. But dm , dm · q, dm · q 2 · 10ε ∈ [1, 10), i. e. ε ∈ {−1, 0, 1}.

Since all the terms of the sequence (dm )∞ m=1 are different and all lie in the interval [1, 10), there have to
be two terms dm and dm0 such that |dm0 − dm | < 10 1
. Without loss of generality let m0 > m. There are two
possibilities.
1
Let dm0 > dm . Then we have dm0 < dm + 10 . Thus
1
dm + 10 1 1
1 < q = dm0 /dm < =1+ ≤1+ 10 .
dm 10dm

By previous remark dm ·q 2 lies in the studied sequence, whenever it lies in the interval [1, 10). Repeating this idea
we have the numbers dm , dm · q, dm · q 2 , dm · q 3 , . . . , dm · q i all lying in the studied sequence and after overrunning
the value 10 we have the numbers dm · q i+1 /10, dm · q i+2 /10, . . . in the sequence, and so on. Computing the
difference of two consecutive terms in this recurrence we get

dm · q j+1 − dm · q j = dm · q j (q − 1) < dm · q j · 1
10 <1 for j < i,
j+1 j j j 1
dm · q /10 − dm · q /10 = dm · q (q − 1)/10 < dm · q /10 · 10 <1 for j > i

and for the first term after overrunning 10 we obtain

dm · q i+1 /10 = dm · q i /10 · q < 10/10 · (1 + 1

10 ) = 11
10 < 2.

Since the difference is less then 1 and after overrunning we jump into the interval [1, 2), we must get at least
one dm+j(m0 −m) in the interval [c, c + 1) for every nonzero digit c.
1
Let dm0 < dm . Then we have dm < dm0 + 10 . Thus
1
dm0 + 10 1 1
1 < q = dm /dm0 < =1+ ≤1+ 10 .
dm0 10dm0
In the very similar way as in the previous case (new terms are generated by dividing instead of multiplying by
q) we obtain the new sequence of terms with consecutive differences less then 1 and after underrunning 1 we
jump to
10
dm /q i+1 · 10 = dm /q i · 10/q > 1 · 100
1 = 11 > 9.
1 + 10
Thus also in this case we must obtain some dm+j(m0 −m) in the interval [c, c + 1) for every nonzero digit c.
This ends the proof.
Answer. Integers satisfying the given conditions are all integers not divisible by 10. 
 The 22nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 30th March 2012
Category II

Problem 1 Let f : [1, ∞) → (0, ∞) be a non-increasing function such that

f (2n+1 ) 1
lim sup < .
n→∞ f (2n ) 2

Prove that Z ∞
f (x) dx < ∞ .
1
Solution Since
2n+1 f (2n+1 )
lim sup < 1,
n→∞ 2n f (2n )
then by ratio test we obtain that the series
∞
X
2n f (2n )
n=1

converges. Using Cauchy condensation test we obtain that

∞
X
f (n)
n=1

converges. Now, by integral test for convergence we have

Z ∞
f (x) dx < ∞.
1


 The 22nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 30th March 2012
Category II

Problem 2 Let M be the (tridiagonal) 10 × 10 matrix

 
−1 3 0 ··· ··· ··· 0
.. 
 

 3 2 −1 0 .
.. 
 
 ..
 0
 −1 2 −1 . . 
M =  ... .. ..  .

 0 −1 2 . 0 .
 . .. .. ..
 ..

 . . . −1 0 
 .
 ..

 0 −1 2 −1  
0 ··· ··· ··· 0 −1 2

Show that M has exactly nine positive real eigenvalues (counted with multiplicities).
Solution Let xT = (0, x1 , ..., x9 ). Then the direct calculation shows that

xT M x = x21 + (x2 − x1 )2 + · · · + (x9 − x8 )2 + x29 . (1)

Let λmin := min{λ | λ ∈ σ(M )} (recall that if a matrix M is symmetric then σ(M ) ⊂ R). Moreover, since M is
symmetric, there exists an orthogonal matrix C such that C T M C = diag{λmin , λ1 ..., λ9 }. Hence we infer that
y T (λmin I − M )y ≤ 0 for y ∈ R10 . Let y T = (1, −1, 0, ..., 0). Then 2λmin ≤ y T M y = −5. Thus λmin < 0.
Let V1 = {(0, x1 , ..., x9 ) | xi ∈ R} ⊂ R10 . Then, in view of (1), we have

yT M y ≥ 0 (2)

for any y ∈ V1 and y T M y = 0 if and only if y = 0.

Suppose on the contrary that M admits at least two nonpositive eigenvalues λ1 , λ2 ∈ σ(M ). Consequently,
there exist y1 , y2 ∈ R10 such that y1 ⊥ y2 , y1T y1 = y2T y2 = 1 and M yi = λi yi (i = 1, 2). Put V2 := span{y1 , y2 }.
Then for any y = α1 y1 + α2 y2 ∈ V2 one has

y T M y = α12 · λ1 + α22 · λ2 ≤ 0. (3)

Finally, we obtain that

dim V1 + dim V2 = 9 + 2 = 11 > 10.
Therefore V1 ∩V2 6= {0}. Take 0 6= y ∈ V1 ∩V2 . Then, in view of (2), y T M y > 0. But (3) implies that y T M y ≤ 0
– a contradiction. 
 The 22nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 30th March 2012
Category II

Problem 3 Let (A, +, ·) be a ring with unity, having the following property: for all x ∈ A either x2 = 1 or
xn = 0 for some n ∈ N. Show that A is a commutative ring.
Solution Denote by U (A) the multiplicative group of units of the ring A (U (A) = {x | x is invertible}). Note
first that (U (A), ·) is commutative, because if x, y ∈ U (A), (xy)2 = 1 ⇒ xy · xy = 1, and multiplying by x to
the left and by y to the right and using also the fact that x2 = 1 = y 2 , we get that

xy = yx. (1)

We now show that if

x∈
/ U (A) then 1 − x ∈ U (A).
Assume, by contradiction, that
∃x ∈
/ U (A) so y = 1 − x ∈
/ U (A). (2)
By hypothesis,
∃n and m ∈ N so xn = 0; y m = 0
and as
xy = x(1 − x) = x − x2 = (1 − x)x = yx
we get that X
(x + y)n+m = i
Cn+m xi y j = 0,
i+j=n+m

Note that whenever i + j = n + m we have

i ≥ n or j ≥ m and so xi = 0 or y j = 0;

So
1=x+y ∈
/ U (A),
which is a contradiction; thus (2) is proved.
Commutativity in A follows now from (1) and (2) with a case by case analysis: x, y ∈ A,

1. if x ∈ U (A), y ∈ U (A) then (1) ⇒ xy = yx ;

2. if x ∈ U (A), y ∈
/ U (A) then (2) ⇒ 1 − y ∈ U (A) and from (1) we have x(1 − y) = (1 − y)x ⇒ xy = yx ;
3. if x ∈
/ U (A), y ∈ U (A) analogous to the case 2 and
4. if x ∈
/ U (A), y ∈
/ U (A) then (2) ⇒ 1 − x, 1 − y ∈ U (A)

and using
(1 − x)(1 − y) = (1 − y)(1 − x) ⇔ 1 − x − y + xy = 1 − y − x + xy ⇔ xy = yx.
Now cases 1 → 4 above show that A is a commutative ring. 
 The 22nd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 30th March 2012
Category II

Problem 4 Let a, b, c, x, y, z, t be positive real numbers with 1 ≤ x, y, z ≤ 4. Prove that

x y z y+z−x z+x−y x+y−z
+ + ≥ + + .
(2a)t (2b)t (2c)t (b + c)t (c + a)t (a + b)t

Solution We will use the following variant of Schur’s inequality.

Lemma 1 For arbitrary A, B, C > 0,

x(A − B)(A − C) + y(B − A)(B − C) + z(C − A)(C − B) ≥ 0.

Proof Without loss of generality we can assume A ≤ B ≤ C. Let U = B − A and V = C − B. Then

LHS = xU (U + V ) − yU V + z(U + V )V ≥ U (U + V ) − 4U V + (U + V )V = (U − V )2 ≥ 0.

Lemma 2 For every p > 0, Z ∞

1 1
k
= tk−1 e−pt dt.
p Γ(k) 0

Proof Substituting u = pt,

Z ∞ Z ∞
k−1 −pt 1 Γ(k)
t e dt = k uk−1 e−u du = .
0 p 0 pk


Now, applying Lemma 1 to A = e−at , B = e−bt , and C = e−ct , the statement can be proved as
Z ∞  
k−1 −at −bt −at −ct −bt −at −bt −ct −ct −at −ct −bt
0≤ t x(e − e )(e − e ) + y(e − e )(e − e ) + z(e − e )(e − e ) dt
0
Z ∞  
k−1 −2at −2bt −2ct −(b+c)t −(c+a)t −(a+b)t
= t xe + ye + ze − (y + z − x)e − (z + x − y)e − (x + y − x)e dt
0
 
x y z y+z−x z+x−y x+y−z
= Γ(k) + + − − − .
(2a)k (2b)k (2c)k (b + c)k (c + a)k (a + b)k


 The 23rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 12th April 2013
Category I

Problem 1 Let f : [0, ∞) → R be a differentiable function with |f (x)| ≤ M and f (x)f 0 (x) ≥ cos x for
x ∈ [0, ∞), where M > 0. Prove that f (x) does not have a limit as x → ∞.
Solution Consider a function F : [0, ∞) → R given by

F (x) := f 2 (x) − 2 sin x.

Then:
• |F (x)| 6 f 2 (x) + 2|sin x| 6 M + 2.
• F 0 (x) = 2f (x)f 0 (x) − 2 cos x > 0.

Hence we infer that F is increasing and bounded. Let


nπ if n = 2k − 1,
xn =
nπ + π2 if n = 2k.

Then (F (xn )) is increasing and bounded and hence convergent. Assume on the contrary that lim f (x) exists.
x→∞
In turn, this implies that lim f 2 (xn ) exists. Thus the sequence F (xn ) − f 2 (xn ) is convergent. But
n→∞

F (xn ) − f 2 (xn ) = −2 sin(xn ) .

Consequently we get that the sequence (sin(xn )) is convergent. This contradicts the fact that (sin(xn )) is not
convergent since 
0 if n = 2k − 1,
sin(xn ) =
1 if n = 2k.

 The 23rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 12th April 2013
Category I

Problem 2 Let A = (aij ) and B = (bij ) be two real 10 × 10 matrices such that aij = bij + 1 for all i, j and
A3 = 0. Prove that det B = 0.
Solution Let H be the matrix 10 × 10 consisting of units. Then A = B + H. As A3 = 0 then

B 3 = (A − H)3 = A3 + a sum of 7 matrices of the rank ≤ 1.

Therefore rank B 3 ≤ 7. Since B is of size 10 × 10, B is degenerate. 

 The 23rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 12th April 2013
Category I

Problem 3 Let S be a finite set of integers. Prove that there exists a number c depending on S such that for
each non-constant polynomial f with integer coefficients the number of integers k satisfying f (k) ∈ S does not
exceed max(deg f, c).
Solution For each set T ⊆ Z let N (f, T ) denote the number of distinct integers k for which f (k) ∈ T . Suppose
that the cardinality of S is at least 2 and suppose for some two elements s1 6= s2 of S the equations f (x) = s1
and f (x) = s2 both have integer solutions, say, x = k1 and x = k2 , respectively. (Otherwise, we immediately
obtain N (f, S) ≤ deg f .) Put d = d(S) for the difference between the largest and the smallest elements of S.
We claim that then N (f, S) ≤ 4d(S).
Indeed, if for some k ∈ Z we have f (k) = s ∈ S, where s 6= s1 (and so k 6= k1 ), then k − k1 divides the
integer f (k) − f (k1 ) = s − s1 . Thus |k − k1 | ≤ |s − s1 | ≤ d. Clearly, there are at most 2d of such integers k
(since k 6= k1 ), so N (f, S \ {s1 }) ≤ 2d. By the same argument, we must have N (f, S \ {s2 }) ≤ 2d. Since S is
contained in the union of the sets S \ {s1 } and S \ {s2 }, we deduce that

N (f, S) ≤ N (f, S \ {s1 }) + N (f, S \ {s2 }) ≤ 2d + 2d = 4d.

Therefore, N (f, S) ≤ max(deg f, 4d(S)). 

 The 23rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 12th April 2013
Category I

Problem 4 Let n and k be positive integers. Evaluate the following sum

k  2  
X k n + 2k − j
j=0
j 2k

n n!


where k = k! (n−k)! .
Solution We show that
k  2    2
X k n + 2k − j n+k
= . (1)
j=0
j 2k k

(2k)!n!
Multiplying equation (1) by we get
(n + k)!k!
k   k  
X k k! (n + 2k − j)! (2k)!n! X k n! (n + 2k − j)!
=
j=0
j j!(k − j)! (2k)!(n − j)! (n + k)!k! j=0 j j!(n − j)! (n + k)!(k − j)!
k    
X k n n + 2k − j
= . (2)
j=0
j j k−j

On the right side in the formula (1) after multiplying we obtain

    
n + k (n + k)! (2k)!n! n+k 2k
= .
k k!n! (n + k)!k! k k

Applying Cauchy identity

  k   
m+n X n m
= ,
k r=0
r k−r

to formula (2) we have

k−j 
k    X  
X k n n−j 2k
. (3)
j=0
j j r=0
r k−j−r

By changing the order of summation in formula (3) putting s = r + j we get

k    X
k   
X k n n−j 2k
=
j=0
j j s=j s − j k−s
k    X
k   
X k n n−j 2k
. (4)
j=0
j j s=0 s − j k−s

Once again by changing the order of summation in formula (4) it follows

k  Xs    
X 2k k n n−j
.
s=0
k − s j=0 j j s−j

On account of the Cauchy identity we have

  k   
2k X n k
.
k s=0 s k−s
Finally we show that
 Xs        
2k k n n−j 2k n k
= .
k − s j=0 j j s−j k s k−s

By applying well-known formula

     
n m n n−k
= .
m k k m−k

it follows
 Xs      Xs        X s   
2k k n n−j 2k k n s 2k n k s
= =
k − s j=0 j j s−j k + s j=0 j s j k+s s j=0 j s−j
           
2k n k+s 2k n k+s n 2k 2k − k
= = =
k+s s s k+s s k s k k+s−k
       
n 2k k n 2k k
= = .
s k s s k k−s

This completes the proof of Li-en-Szua formula. 

 The 23rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 12th April 2013
Category II

Problem 1 Let Sn denote the sum of the first n prime numbers. Prove that for any n there exists the square
of an integer between Sn and Sn+1 .
Solution We have √ √
x<m< y ⇒ x < m2 < y,
√ √
so if y − x > 1, there is certainly a square between x and y.
We have √ √
√
y − x > 1 ⇒ y − x > 1 + 2 x,
hence it suffices to prove p
Sn+1 − Sn > 1 + 2 Sn .
For n = 1, 2, 3, 4 the assertion can be seen directly. For n ≥ 5, we use

Sn < 1 + 3 + 5 + . . . + pn ,
2
√
√ sum equals 1/4(1+pn ) , so it follows that 2 Sn < 1+pn .
where the sum contains all odd integers up to pn . Their
As pn+2 is at least pn + 2, we get Sn+1 − Sn > 1 + 2 Sn as desired. 
 The 23rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 12th April 2013
Category II

Problem 2 An n-dimensional cube is given. Consider all the segments connecting any two different vertices
of the cube. How many distinct intersection points do these segments have (excluding the vertices)?
Solution We may think that every vertex of the cube has a view (ε1 , . . . , εn ) where εi ∈ {0, 1} for i = 1, 2, . . . , n.
A cross-point of two segments has a view (α1 , . . . , αn ) where αi ∈ {0, 12 , 1}. For example, if A = (0, 0, 0, 1, 1),
B = (1, 0, 0, 0, 1), C = (1, 0, 0, 1, 1), D = (0, 0, 0, 0, 1) then AB ∩ CD = ( 12 , 0, 0, 12 , 1). However a row containing
less than 2 of 12 may be not a cross-point. Therefore, there are exactly 3n − 2n − n2n−1 of cross-points. 
 The 23rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 12th April 2013
Category II

√ √ √
Problem 3 Prove that there is no polynomial P with integer coefficients such that P ( 3 5 + 3 25) = 5 + 3 5.
Solution First we prove two lemmas.
√ √ √
Lemma 1. There is no polynomial w(x) = ax + b with integer coefficients such that w( 3 5 + 3 25) = 5 + 3 5;
√ √
Proof Assume on the contrary that such a polynomial w(x) = ax + b exists. Since 3 5 and 3 25 are irrational,
it follows that a 6= 0 and a 6= 1. Furthermore, one has
√3
√3
√3
√
3
√
3
a( 5 + 25) + b = 5 + 5 =⇒ (a − 1) 5 + a 25 ∈ Q
 √ √  2 √ √
=⇒ (a − 1) 5 + a 25 ∈ Q =⇒ (a − 1)2 25 + 5a2 5 ∈ Q
3 3 3 3

5a2  √ √   √ √ 
(a − 1) 5 + a 25 + (a − 1)2 25 + 5a2 5 ∈ Q
3 3 3 3
=⇒
(1 − a)
(a − 1)3 − 5a3 √ √
 
3 3
=⇒ 25 ∈ Q =⇒ 25 ∈ Q,
(a − 1)
√
which contradicts the fact that 3 25 ∈ nQ, where Q and nQ denote the set of rational and irrational numbers,
respectively. This completes the proof of the lemma. 
√
3
Lemma
√ 2. √
There exists exactly one polynomial w(x) of degree two and rational coefficients such that w( 5 +
3
25) = 5 + 3 5;
Proof Consider a polynomial w(x) = ax2 + bx + c, where a, b, c ∈ Q. Then
√ √ √ √ √ √ √ √
w( 5 + 25) = 5 + 5 ⇐⇒ a( 5 + 25)2 + b( 5 + 25) + c = 5 + 5
3 3 3 3 3 3 3 3

 
 a+b = 0  a = 1/4
⇐⇒ 5a + b = 1 ⇐⇒ b = −1/4
10a + c = 5 c = 10/4
 

This implies that there exists only one polynomial w(x) with the required properties, i.e.,
1 2 1 10 √
3
√
3
√
3
w(x) = x − x+ and w( 5 + 25) = 5 + 5,
4 4 4
which completes the proof of the second lemma.

√
3
√
3
Now we are ready to solve the problem. Let x0 := 5 + 25. Then
√ √ √3
√
3
√ √
x30 = ( 5 + 25)3 = 5 + 3 54 + 3 55 + 25 = 30 + 15 5 + 15 5 = 15x0 + 30.
3 3 3 3

We put Q(x) := x3 − 15x − 30. Then Q(x0 ) = 0. Assume on the contrary that such a polynomial P (x) exists.
Then there exist two polynomials R(x) and w(x) with integer coefficients such that

P (x) = Q(x)R(x) + w(x),

where the degree deg w(x) of w(x) is less than or equal 2. Consequently we obtain
√3
√
3
√
3
√
3
√
3
√3
√3
√3
√
3
√
3
√
3
5 + 5 = P ( 5 + 25) = Q( 5 + 25)R( 5 + 25) + w( 5 + 25) = w( 5 + 25).

From this it follows that there exists a polynomial w(x) of degree less than or equal 2 with integer coefficients
such that
√
3
√3
√3
w( 5 + 25) = 5 + 5,

a contradiction with Lemma 1 and Lemma 2. This completes the solution. 

 The 23rd Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 12th April 2013
Category II

Problem 4 Let F be the set of all continuous functions f : [0, 1] → R with the property
Z x 
 f (t) 
 √ dt  ≤ 1 for all x ∈ (0, 1] .

0 x−t 
Z 1 
 
Compute sup  f (x) dx.
f ∈F 0
Solution We will use the following lemma.
Lemma For every functions f ∈ L1 [0, 1],
Z 1 Z x  Z 1
f (t)dt dx
√ √ =π f.
0 0 x−t 1−x 0

Proof Changing the order of integration then substituting t = −1 + 2 x−t

1−t ,
!
Z 1 Z x  Z 1 Z 1
f (t)dt dx dx
√ √ = f (t) p dt
0 0 x−t 1−x 0 t (x − t)(1 − x)
!
Z 1 Z 1 Z 1
dt
= f (t) p dt = π f.
0 −1 (1 + t)(1 − t) 0

Now, by Lemma, for all f ∈ F ⊂ L1 [0, 1] we have

Z 1  Z 1 Z x
1 1 dx
 Z
≤ 1 f (t)dt  dx 2
  
 f  √ √ ≤ √ =

0
 π
0

0 x − t  1−x π 0 1 − x π
R 
 1 
so supf ∈F  0 f  ≤ π2 .
1
For the function g(x) = √ we have
π x
Z x
1 x
Z
g(t)dt dt
√ = p = 1.
0 x−t π 0 t(x − t)

1
Define a sequence f1 , f2 , . . . of [0, 1] → R functions as fn (x) = q . Then fn ∈ C[0, 1] and 0 < f ≤ g, so
1
π x+ n
R1 R1 2
fn ∈ F. As fn (x) →
 R g(x)
 pointwise, we have 0
fn → 0
g= π.
 1 
Hence, supf ∈F  0 f  = π2 . 
 The 24th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2014
Category I

√
Problem 1 Find all complex numbers z such that |z 3 + 2 − 2i| + z z̄|z| = 2 2 . (z̄ is the conjugate of z.)
Solution p √
(−2)2 + 22 = 2 2 = |z 3 + 2 − 2i| + z z̄|z| = |z 3 − (−2 + 2i)| + |z 3 | .
By the triangle inequality number z 3 must be a point of the straight line segment with ends 0 and −2 + 2i =
(1 + i)3 , so z must be a point of the union of the three straight line segments with the common end 0 and the
remaining end equal to either 1 + i or
√ !
1 3 1 √ √ 
(1 + i) − + i = −1 − 3 + i( 3 − 1)
2 2 2

or √ !
1 3 1 √ √ 
(1 + i) − − i = −1 + 3 − i( 3 + 1) .
2 2 2

Second solution √
Since z z̄ = |z|2 = |z 2 | the equation may be rewritten as |z 3 + 2 − 2i| + |z 3 | = 2 2. Let z 3 = x + yi where
x, y ∈ R. The equation is equivalent to
p p √
(x + 2)2 + (y − 2)2 + x2 + y 2 = 2 2. (1)

Therefore √ p
(x + 2)2 + (y − 2)2 = 8 − 4 2 x2 + y 2 + x2 + y 2 ,
√ p
so x − y = 2 x2 + y 2 and (x + y)2 = 0, i.e. y = −x. Therefore the equation 1 takes the form
p p √
(x + 2)2 + (−x − 2)2 + x2 + (−x)2 = 2 2 (2)

which is equivalent to
|x + 2| + |x| = 2. (3)
Therefore −2 ≤
 x√≤ 0. This means that z 3 = x − xi for some x ∈ [−2, 0], i.e. z 3 = r(cos 135◦ + i sin 135◦ )
for some r ∈ 0, 2 2 . Therefore z = % (cos(45◦ + n · 120◦ ) + i sin(45◦ + n · 120◦ )) with n ∈ {0, 1, 2} and
√
0 ≤ % ≤ 2.

 The 24th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2014
Category I

Problem 2 We have a deck of 2n cards. Each shuffling changes the order from a1 , a2 , . . . , an , b1 , b2 , . . . , bn to
a1 , b1 , a2 , b2 , . . . , an , bn . Determine all even numbers 2n such that after shuffling the deck 8 times the original
order is restored.
Solution Note that the cards a1 and bn always stay on the top/bottom of the deck respectively. From now on
we will ignore the card bn . Let us number the positions of the cards: f (ai ) = i − 1, f (bi ) = n + i − 1. Note that
the shuffle will put the card with position i to position 2i for every i < n, or to 2i − (2n − 1) for every n ≤ i.
This shows that the shuffling works like the mapping

ϕ : Z2n−1 → Z2n−1 , k 7→ 2k.

Shuffling 8 times will map each k to 256k. So we can reformulate the question:
For what numbers 2n will the following congruence hold for every k ∈ Z2n−1 :

k ≡ 256k (mod 2n − 1)

It is easy to see that this congruence holds iff it is true for k = 1:

1 ≡ 256 (mod 2n − 1)
Which holds iff 2n − 1 | 255. So the set we are looking for is: {2, 4, 6, 16, 18, 52, 86, 256}.

 The 24th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2014
Category I

Problem 3 Let n ≥ 2 be an integer and let x > 0 be a real number. Prove that
 p n p
1 − tanh x + tanh (nx) < 1 .

e2t − 1
Recall that tanh t = .
e2t + 1
Solution We will prove that for all real numbers x, y > 0
 √  p  p
1 − tanh x 1 − tanh y < 1 − tanh(x + y). (1)

Since
ex − e−x e2x − 1 sinh x cosh y + sinh y cosh x tanh x + tanh y
tanh x = = and tanh(x + y) = =
ex + e−x e2x + 1 cosh x cosh y + sinh x sinh y 1 + tanh x tanh y

the inequality (1) is equivalent to

r
u2 + v 2
(1 − u) (1 − v) < 1 − for 0 < u, v < 1 (2)
1 + u2 v 2
√ √
via the substitutions u := tanh x and v := tanh y. The inequality (2) can be shown as follows:

(1 − u)(1 − v) > 0
=⇒ 2(1 + uv) > (1 + u)(1 + v)
(1 + u)(1 + v)uv (1 + u)(1 + v)
=⇒ 2uv > = (1 + u)(1 + v) −
1 + uv 1 + uv
(1 + u)(1 + v) (1 + u)(1 + v)
=⇒ > > (1 + u)(1 + v) − 2uv = 2 − (1 − u)(1 − v)
1 + u2 v 2 1 + uv
(1 − u2 )(1 − v 2 )
=⇒ > 2(1 − u)(1 − v) − (1 − u)2 (1 − v)2 = 1 − (1 − (1 − u)(1 − v))2
1 + u2 v 2
r r
u2 + v 2 (1 − u2 )(1 − v 2 )
=⇒ = 1 − < |1 − (1 − u)(1 − v)| = 1 − (1 − u)(1 − v)
1 + u2 v 2 1 + u2 v 2
r
u2 + v 2
=⇒ (1 − u)(1 − v) < 1 − .
1 + u2 v 2
Thus we have shown (1) and from this the assertion follows by induction: Take y = x for the case n = 2 and
y = nx for the inductive step n → n + 1. 
 The 24th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2014
Category I

Problem 4 Let P1 , P2 , P3 , P4 be the graphs of four quadratic polynomials drawn in the coordinate plane.
Suppose that P1 is tangent to P2 at the point q2 , P2 is tangent to P3 at the point q3 , P3 is tangent to P4 at
the point q4 , and P4 is tangent to P1 at the point q1 . Assume that all the points q1 , q2 , q3 , q4 have distinct
x-coordinates. Prove that q1 , q2 , q3 , q4 lie on a graph of an at most quadratic polynomial.
Solution We may subtract a quadratic trinomial from all the given trinomials so that the points q1 , q2 , q3
get to the 0x axis. After this the trinomials remain trinomials, possibly degenerate, and the tangency is not
affected. Let q40 be the point of intersection of P3 and 0x distinct from q3 ; and let q400 be the intersection of P4
and 0x distinct from q1 . If q40 and q400 coincide then they also coincide with q4 and the assertion follows.
Assume not. Every parabola (graph of a quadratic trinomial) that intersects the 0x axis twice, intersects
it at the same angles. Having applied this to all the four parabolas in the circular order, we obtain that P3
in q40 and P4 in q400 6= q40 have the same slopes. But they also touch each other in the point q4 ; therefore they
are homothetic with respect to the point q4 . This homothety takes q40 into q400 , because the slope is preserved
under the homothety and the point on a parabola is defined uniquely by its slope. Hence q4 must also lie on
the 0x axis, this is what we need to prove. The argument with homothety fails in one of the parabolas P3 or
P4 degenerate to a straight line; but in this case the point q4 also must be on 0x evidently. 
 The 24th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2014
Category II

Problem 1 Let f : (0, ∞) → R be a differentiable function. Assume that

f 0 (x)
 
lim f (x) + = 0.
x→∞ x

Prove that
lim f (x) = 0 .
x→∞

Solution Assume that

f 0 (x)
 
lim f (x) + = 0.
x→∞ x
Fix ε > 0. Then there exists x0 > a such that
0
 
f (x) + f (x)  < ε
 
 x  2

for all x > x0 . Take x > x0 and define two functions g, h : [x0 , x] → R by
x2 x2
g(x) = e 2 f (x) and h(x) = e 2 .

Now by applying Cauchy’s mean value theorem we get

g(x) − g(x0 ) g 0 (η) f 0 (η)

= 0 = f (η) + ,
h(x) − h(x0 ) h (η) η

where η ∈ (x0 , x). Hence

  f 0 (η)  
  
1 2 2  1 2 2 
f (x) − f (x0 )e 2 (x0 −x )  = f (η) + · − e 2 (x0 −x )  .

η 
1

Consequently, for all x > x0 one has

 1 2 2 
 ε 1 2 2 
 1 2 2 ε
|f (x)| < f (x0 )e 2 (x0 −x )  + 1 − e 2 (x0 −x )  < |f (x0 )|e 2 (x0 −x ) + .
 
2 2
1 2 2
Since lim e 2 (x0 −x )
= 0, there exists x
e > x0 such that
x→∞

1 2 2 ε
|f (x0 )|e 2 (x0 −x )
<
2
for all x > x
e. Finally, taking into account the above considerations, we infer that

|f (x)| < ε

for all x > x

e, which implies the desired conclusion. 
 The 24th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2014
Category II

Problem 2 Let p be a prime number and let A be a subgroup of the multiplicative group F∗p of the finite
field Fp with p elements. Prove that if the order of A is a multiple of 6, then there exist x, y, z ∈ A satisfying
x + y = z.
Solution Obviously the existence of x, y, z ∈ A with x + y = z is equivalent to the existence of x0 , y 0 ∈ A with
x0 + y 0 = 1 (just divide by z). First observe that A is cyclic, so 6 | d implies the existence of some a ∈ A of
order 6, hence a6 = 1. This yields a3 = −1, so that we obtain

1 − a6
a + a3 + a5 = a(1 + a2 + a4 ) = a = 0,
1 − a2
from which we conclude
a + a5 = −a3 = 1.
Setting x0 := a, y 0 := a5 gives the desired equality. 
 The 24th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2014
Category II

Problem 3 Let k be a positive even integer. Show that

k/2   
X k+2 2(k − n) + 1 (k + 1)(k + 2)
(−1)n = .
n=0
n k+1 2

Solution
Lemma For any positive integer k and −1 < x < 1 the following formula holds
 k X ∞  
1 k+n−1 n
= x .
1−x n=0
k−1

Proof For k = 1 we have

∞ ∞  
1 X
n
X n n
= x = x .
1 − x n=0 n=0
0
Assume inductively that
 k ∞  
1 X k+n−1 n
= x .
1−x n=0
k−1
Then for k + 1, by using the Cauchy product for two infinite series, we obtain
∞   ! X ∞ ∞
 k+1  k   ! n   !
1 1 1 X k+n−1 n n
X X k+m−1 n
= = x x = x .
1−x 1−x 1−x n=0
k−1 n=0 n=0 m=0
k−1

But, by using once again the induction principle, one can prove that
n    
X k+m−1 k+n
= .
m=0
k−1 k

Finally, we obtain that

∞ n   ! X ∞  
X X k+m−1 n k+n n
x = x ,
n=0 m=0
k−1 n=0
k
which completes the proof of the lemma. 

Observe that
(1 − x2 )k+2 1
(1 − x)k+2 = = (1 − x2 )k+2 ,
(1 + x)k+2 (1 + x)k+2
for −1 < x < 1. On the other hand,
k+2  
X k+2 n
(1 − x)k+2 = (−1)n x ,
n=0
n
k+2
! ∞  !
1 X k + 2  X k + n + 1
(1 − x2 )k+2 = n 2n
(−1) x n n
(−1) x .
(1 + x)k+2 n=0
n n=0
k+1

Consequently, by comparing the corresponding coefficients of xk in the above equalities we obtain the following
equation:
  k/2    
k+2 X k+2 k + (k − 2n) + 1
(−1)k = (−1)n (−1)k−2n ,
k n=0
n k+1

which implies the desired equality:

k/2     
X k+2 2(k − n) + 1 k+2
(−1)n = .
n=0
n k+1 k


 The 24th Annual Vojtěch Jarník
International Mathematical Competition
Ostrava, 4th April 2014
Category II

Rb
Problem 4 Let 0 < a < b and let f : [a, b] → R be a continuous function with a
f (t) dt = 0. Show that
Z b Z b
f (x)f (y) ln(x + y) dx dy ≤ 0 .
a a

Solution
Lemma (well-known)
∞
e−t − e−Kt
Z
dt = log K
0 t
for every K > 0.
Proof Let
∞
e−t − e−Kt
Z
h(K) = dt.
0 t
Z ∞
1
From h(1) = 0 and h0 (K) = e−Kt dt = we get
0 K
Z K
dk
h(K) = = log K.
1 k


To prove the problem statement, take the sum of the identities

Z bZ b Z ∞ −t b b
e − e−(x+y)t
 Z Z
f (x)f (y) dt dx dy = f (x)f (y) log(x + y) dx dy
a a 0 t a a
!
Z bZ b Z ∞ −t −(x+ 21 )t b ∞ 1 b
−e + e −e−t + e−(x+ 2 )t
Z Z Z
f (x)f (y) dt dx dy = f (x) dt dx · f (y) dy = 0 and
a a 0 t a 0 t a
Z ∞ 1
! 1
Z bZ b b ∞ b
−e−t + e−(y+ 2 )t −e−t + e−(y+ 2 )t
Z Z Z
f (x)f (y) dt dx dy = f (y) dt dx · f (x) dx = 0.
a a 0 t a 0 t a

As can be seen,
1 1
!
b b b b ∞
−e−t + e−(x+ 2 )t + e−(y+ 2 )t − e−(x+y)t
Z Z Z Z Z
f (x)f (y) log(x + y) dx dy = f (x)f (y) dt dx dy
a a a a 0 t
∞ b b 1 1
(e− 2 t − e−xt )(e− 2 t − e−yt )
Z Z Z
=− f (x)f (y) dx dy dt
0 a a t
!2
Z ∞ Z b
− 21 t −xt

dt
=− f (x) e −e dx ≤ 0.
0 a t