The University of Western Australia

SCHOOL OF MATHEMATICS AND STATISTICS

BLAKERS MATHEMATICS COMPETITION

2016 Problems
Note. Our convention is that N = {1, 2, . . . } (the positive integers).
1. Digitally prodigious

A number N is formed by concatenating the digits of the numbers 2, 22 , . . . , 22016 and the
numbers 5, 52 , . . . , 52016 in their usual decimal representation.
How many digits does N have?
If the second sequence of numbers were powers of 8, instead of 5, and each of the numbers
were written as hexadecimals, how many digits would N have then?
Solution. Let d(n) be the number of digits of a natural number n in its decimal represen-
tation. Suppose an integer n has k digits. Then

10k−1 ≤ n < 10k

∴ k − 1 ≤ log10 n < k
∴ d(n) = blog10 nc + 1.

Observe that
log10 (2m ) + log10 (5m ) = log10 (10m ) = m.
Suppose blog10 (2m )c = k − 1. Then, since log10 (2m ) and log10 (5m ) are irrational,

k − 1 < log10 (2m ) <k

m
∴ −k < − log10 (2 ) < −k + 1
m
∴ m − k < m − log10 (2 )
= log10 (5m ) < m − k + 1.

Therefore,

d(2m ) + d(5m ) = (k − 1) + 1 + (m − k) + 1
=m+1
2016
X 2016
X
m
∴ d(N ) = d(2 ) + d(5m )
m=1 m=1
2016
X
= (m + 1)
m=1
= 2 + 3 + · · · + 2017
2016(2 + 2017)
=
2
= 2 035 152.

Hence, N has 2 035 152 digits.

 Is there a difference if the second sequence of numbers are powers of 8 instead of powers of 5,
and the numbers are now written as hexadecimals? Indeed, there is a difference, essentially
due to log16 (2m ) and log16 (8m ) being rational. Let δ(n) be the number of digits of a natural
number n when written as a hexadecimal. Then, analogously, we have

δ(n) = blog16 nc + 1.

Suppose log16 (2m ) = k ∈ Z, which is the case whenever m is a multiple of 4. Then log16 (8m ) =
m − k, and so, in this case,

δ(2m ) + δ(8m ) = k + 1 + (m − k) + 1 = m + 2.

So we get an extra digit whenever m is a multiple of 4, and hence the number of hexadecimal
digits in our new N is
2 035 152 + 504 = 2 035 656.

2. An 8-dollar problem

There are just two essentially different ways of arranging three $2 coins and two $1 coins in
a ring so that each coin is tangent to two others and all five coins are externally tangent to
a circle inside the ring.
For which arrangement is the radius of the inner circle larger? See diagram below.

$2 $2
$1 $1
$2 $2
$1 $2
$2 $1

Note. The official diameters of the $1 and $2 coins is 25.0 mm and 20.5 mm, respectively,
though one may solve this problem knowing only that a $1 coin is larger than a $2 coin.
Solution.
Using the law of cosines for the triangle formed by joining the centres of an adjacent $1 and
$2 coin and the central circle, with θ the angle at the centre of the central circle, we have

(r1 + r2 )2 = (r + r1 )2 + (r + r2 )2 − 2(r + r1 )(r + r2 ) cos θ,

where r1 , r2 , r are the radii of the $1 and $2 coins, and central circle, respectively. Thus,
 2r1 r2 
θ = arccos 1 − .
(r + r1 )(r + r2 )
For the combination with the two $1 coins together, we have
 2r12   2r1 r2   2r22 
2π = arccos 1 − + 2 arccos 1 − + 2 arccos 1 − ,
(r + r1 )2 (r + r1 )(r + r2 ) (r + r2 )2

2
 while for the combination with the $1 coins separated,
 2r1 r2   2r22 
2π = 4 arccos 1 − + arccos 1 − .
(r + r1 )(r + r2 ) (r + r2 )2

With r1 = 12.5 and r2 = 10.25, we find for the first case r ≈ 7.7740, and for the second case
r ≈ 7.7644. So the arrangement with the $1 coins together is the one where the radius of the
inner circle is larger.

3. Attacking chess

A rook and a bishop are randomly placed on two different squares of an 8 × 8 chessboard
devoid of any other pieces.
What is the probability that one of the pieces can take the other, according to the rules of
chess?

Solution. If the rook is on one of the 4 central squares, it threatens 14 squares (left-right
and up-down) and is threatened from 13 squares along diagonals, a total of 27 squares (of
the 63 squares not occupied by the rook). This total is reduced to 25 if the rook is on one
of the 12 squares surrounding the 4 central squares, 23 for the 20 squares surrounding the
preceding 12, and 21 for the 28 squares along the sides of the chessboard. The probability
required is therefore
4 27 12 25 20 23 28 21 13
· + · + · + · = = 0.36111 . . .
64 63 64 63 64 63 64 63 36
The problem may be generalised to an n × n chessboard, in which case the probability is
10n − 2
, n > 1.
3n(n + 1)

4. Fortune chests

Before you are 4 chests, made respectively of iron, copper, teak and ebony. You are told
that one chest contains treasure, while the other 3 chests contain poisonous scorpions. You
are allowed to choose one chest, lift its lid and plunge your hand inside to take its contents.
On each of the 4 lids is inscribed a statement that is either true or false, but you do not
know which of them are true.
Here are four statements:

On the iron chest: “If the treasure is in the teak chest, then the statement on the ebony
chest is false”.
On the copper chest: “The treasure is in the teak chest or the ebony chest.”
On the teak chest: “One and only one of the four statements is true”.
On the ebony chest: “The statements given in the two metal chests are either both
true or both false”.

3
 In which chest can you safely plunge your hand to retrieve the treasure?
Solution. The treasure is in the ebony chest. In order to prove this, let us first denote by
Fe, Cu, T, and E, the statement inscribed on the lids of the iron, copper, teak, and ebony
chest, respectively.
First suppose Fe is false. Then the treasure would be in the teak chest and E would be true.
Hence Cu is false (as we already have Fe false), so that the treasure is not in the teak chest,
contradicting what we had previously deduced.
Therefore, Fe is true. But then T is false (otherwise there would be at least two true
statements, namely Fe and T, a contradiction). So at least one of the two statements Cu
and E is true. If E is false, then Cu would be also (since we already know that Fe is true),
a contradiction. Therefore, E is true and thus so is Cu. Hence, the treasure is in the teak or
ebony chest. If the treasure is in the teak, then the true statement Fe implies that E is false,
a contradiction. And so, the treasure is in the ebony chest.

5. Functionally sound?
What are all pairs (f, g) of functions f, g : R → R such that f and g never cancel∗ and that
 f 0 f 0
= 0 and f 0 g 0 = f g?
g g
∗
:the correct word here for cancel is vanish, i.e. f, g are never zero.
Solution. Since f and g never vanish, we have f (x) 6= 0 and g(x) 6= 0 for all x ∈ R and
since f 0 g 0 = f g we have f 0 (x) 6= 0 and g 0 (x) 6= 0 for all x ∈ R. Rearranging f 0 g 0 = f g we have
f 0g0
f= .
g
Also,
 f 0 f 0g − f g0 f0
= = .
g g2 g0
Combining these we have
f 0 g 2 − f 0 (g 0 )2 f0
3
= 0
g g
0
 0  0 
f g g 3
∴ 0 − − 1 = 0.
g g g
Since we deduced that f 0 and g 0 never vanish, we have
g 0  g 0 3
− − 1 = 0,
g g
which has solution
g0 1 √ √
= C = −18− 3 (9 − 69)1/3 + (9 + 69)1/3 ≈ −1.3247.


g
Hence g = AeCx and f = Bex/C for some constants A and B.

4
6. Box of balls

A box contains 1000 blue balls. Outside the box there is an unlimited supply of blue, red
and green balls. Every 10 seconds, we choose two balls from the box and replace them by
one or two outside balls, using the following replacement rules:

• Two blue with a green;

• Two red with a green;
• Two green with a blue and red;
• One blue and one green with a red;
• One red and one green with a blue;
• One blue and one red with a blue and a red.

Is it possible that the box contains only one ball after a finite time?

Solution.
For each state S of the box, containing b blue balls, g green balls and r red balls, associate
the function f (S) = b + 2g + 3r. Let S0 denote the initial state and St the state after 10t
seconds. So we have f (S0 ) = 1000 and we easily check that, whatever the chosen operation
to go from St to St+1 , we have f (St+1 ) = f (St ) or f (St ) − 4. Since f (S0 ) is a multiple of 4, we
deduce that f (St ) is a multiple of 4 for all t ∈ N ∪ {0}. Thus, a state St in which there would
be a single ball in the box is impossible, since we would have f (St ) = 1, 2 or 3 depending on
whether the remaining ball is blue, green or red.

7. Integer pairs

What are all pairs (m, n) of positive integers such that 22016 + 22012 + 22008 + 2m = n2 ?

Solution.
The given equation can be rewritten 3 · 7 · 13 · 22008 + 2m = n2 . Since 2m is not divisible by 3,
nor is n, i.e. n ≡ ±1 (mod 3), and so n2 ≡ 1 (mod 3). It follows that 2m ≡ 1 (mod 3), and
hence m is even since 2 ≡ −1 (mod 3). Let m = 2k with integer k > 0. Then the equation
becomes 3 · 7 · 13 · 22008 = (n − 2k )(n + 2k ). Therefore,

n − 2k = a2b (1)
k d
and n + 2 = c2 (2)

where a, b, c, d ∈ Z such that, ac = 3 · 7 · 13 and b + d = 2008. By subtracting (1) and (2),

we obtain 2k+1 = c2d − a2b . Since a and c are odd, it follows that b = d (since with b 6= d,
on division of both sides by 2min b,d , the highest power of 2 dividing c2d − a2b , one would find
that 2k+1 has a odd divisor greater than 1, a contradiction).
Thus we have n − 2k = a21004 and n + 2k = c21004 , where a, c are such that ac = 3 · 7 · 13 and
a − c is a power of 2. Enumerating the possibilities and checking each in turn, we find that
the only solutions are (a, c) = (39, 7) or (21, 13), with 39 − 7 = 25 and 21 − 13 = 23 .
Thus we conclude that (m, n) = (2016, 23 · 21004 ) or (2012, 17 · 21004 ).

5
8. Tangential

What is the value of the product

√
√
√ √
√
3 + tan(1◦ ) 3 + tan(2◦ ) 3 + tan(3◦ ) · · · 3 + tan(28◦ ) 3 + tan(29◦ ) ?

Solution.
For n = 1, 2, . . . , 29 we have
√ sin(60◦ ) sin(n◦ ) sin(60◦ + n◦ ) cos(30◦ − n◦ )
3 + tan(n◦ ) = + = = 2 .
cos(60◦ ) cos(n◦ ) cos(60◦ ) cos(n◦ ) cos(n◦ )

Hence
29 29
Y √
Y cos(30◦ − n◦ ) cos(29◦ ) · · · cos(1◦ )
3 + tan(n◦ ) = 2 ◦
= 229 = 229 .
cos(n ) cos(1◦ ) · · · cos(29◦ )
n=1 n=1

9. Continuously decreasing

Is it true, that if f : R → R is a continuous decreasing function, then the system of equations

x = f (y),
y = f (z),
z = f (x),

has a unique solution in R3 ?

Solution.
Let us first prove that a continuous, decreasing function f : R → R has a unique fixed point.
Indeed, the continuous function g : R → R defined by g(x) = f (x) − x is decreasing (since
x < y implies f (x) > f (y), and consequently, g(x) = f (x) − x > f (y) − y = g(y)). Moreover,
for all x ≥ 0, we have g(x) = f (x)−x ≤ f (0)−x, and for all x ≤ 0, g(x) = f (x)−x ≥ f (0)−x.
It follows that g(x) → ∓∞, as x → ±∞. Thus, by the Intermediate Value Theorem, there
exists x0 ∈ R such that g(x0 ) = a, that is to say, f (x0 ) = x0 . The function f cannot have
another fixed point, because if to the contrary x and x0 are two fixed points such that x < x0 ,
then x = f (x) > f (x0 ) = x0 , a contradiction. The triple (x0 , x0 , x0 ), where x0 is the only fixed
point of f , is clearly a solution of the system. On the other hand, if (x, y, z) is a solution
of the system then x = f (y) = f (f (z)) = f (f (f (x))), and similarly, y = f (f (f (y))) and
z = f (f (f (z))). In other words, x, y and z are necessarily fixed points of the continuous
function f ◦3 = f ◦ f ◦ f , which is decreasing as a < b implies successively f (a) > f (b),
f (f (a)) < f (f (b)) and f (f (f (a))) > f (f (f (b))). The f ◦3 function therefore has a unique
fixed point, and the fixed point x0 of f is clearly a fixed point of f ◦3 . Hence the triple
(x0 , x0 , x0 ) is the unique solution of the system. So, indeed, it is true that the system has a
unique solution in R3 .

6
10. Disjoint circles

In the Euclidean plane, draw three disjoint circles, each exterior to the others, and with
their centres noncollinear.
If for each of the three circles, one joins the centre of that circle to the intersection of the
interior common tangents to the two other circles, are the three lines obtained necessarily
concurrent?

Solution.
Denote the three circles by C, C 0 , C 00 , and let their centres and radii be O, O0 , O00 , and r,
r0 , r00 , respectively. The common internal tangents to C and C 0 (resp. C 0 and C 00 , C 00 and
C) intersect at a point P 00 (resp. P , P 0 ), that is located symmetrically on the line joining the
two circle centres. Let T and T 0 be the points of contact of one of these tangents with C and
C 0 , respectively. Since the triangles OT P 00 and O0 T 0 P 00 are similar,
|OP 00 | r
= 0.
|O0 P 00 | r
Similarly,
|O0 P | r0 |O00 P 0 | r
00
= 00
and 0
= 0.
|O P | r |OP | r
Hence,
|OP 00 | |O0 P | |O00 P 0 | r r0 r00
· · = · · = 1,
|O0 P 00 | |O00 P | |OP 0 | r0 r00 r
and therefore, by Ceva’s Theorem applied to triangle OO0 O00 , the cevians OP , O0 P 0 and O00 P 00
are concurrent. So, yes, the three lines defined in the problem are necessarily concurrent.

O0
P 00

P
O

P0
O00