The CENTRE for EDUCATION

in MATHEMATICS and COMPUTING

cemc.uwaterloo.ca

2023 Euclid Contest

Tuesday, April 4, 2023

(in North America and South America)

Wednesday, April 5, 2023

(outside of North America and South America)

Solutions

©2023 University of Waterloo

2023 Euclid Contest Solutions Page 2

1. (a) Since the average of the 5 numbers n, 2n, 3n, 4n, and 5n is 18, we obtain the equation
n + 2n + 3n + 4n + 5n
= 18.
5
15n
Therefore, = 18 and so 3n = 18 or n = 6.
5
(b) Solution 1
Adding the equations 2x + y = 5 and x + 2y = 7, we obtain (2x + y) + (x + 2y) = 5 + 7
and so 3x + 3y = 12.
x+y 3x + 3y 12
Therefore, the average of x and y is = = = 2.
2 6 6
Solution 2
Since 2x + y = 5, then 4x + 2y = 10.
Subtracting the second equation, we obtain (4x + 2y) − (x + 2y) = 10 − 7 which gives
3x = 3 and so x = 1.
Thus, y = 5 − 2x = 3.
1+3
The average of x and y is thus = 2.
2
t2 + 2t + 3
(c) Since the average of the three numbers t2 , 2t and 3 is 9, then = 9.
3
Therefore, t2 + 2t + 3 = 27 and so t2 + 2t − 24 = 0 which gives (t + 6)(t − 4) = 0.
Since t < 0, then t = −6.

1+r p+5
2. (a) Since Q(5, 3) is the midpoint of P (1, p) and R(r, 5), then = 5 and = 3.
2 2
Thus, 1 + r = 10 which gives r = 9, and p + 5 = 6 which gives p = 1.
Therefore, p = 1 and r = 9.
(b) Solution 1
The point with coordinates P (3, 6) is 6 units above the x-axis.
A line with slope 3 moves 2 units to the right as it moves 6 units up. Therefore, to move
from P (3, 6) to the x-axis along a line with slope 3 results in a move of 6 units down and
2 units left. Thus, its x-intercept is 3 − 2 = 1.
A line with slope −1 moves 6 units to the left as it moves 6 units up. Therefore, to move
from P (3, 6) to the x-axis along a line with slope −1 results in a move of 6 units down
and 6 units right. Thus, its x-intercept is 3 + 6 = 9.
The distance between these x-intercepts is 9 − 1 = 8.

Solution 2
The line with slope 3 that passes through P (3, 6) has equation y − 6 = 3(x − 3) or
y = 3x − 3.
The x-intercept of this line has y = 0 and so 0 = 3x − 3 or 3x = 3, which gives x = 1.
The line with slope −1 that passes through P (3, 6) has equation y − 6 = (−1)(x − 3) or
y = −x + 9.
The x-intercept of this line has y = 0 and so 0 = −x + 9 or x = 9.
The distance between these x-intercepts is 9 − 1 = 8.
2023 Euclid Contest Solutions Page 3

(c) The line with equation y = 2x + 7 has slope 2.

The line with equation y = tx + t has slope t.
Since these lines are perpendicular, the product of their slopes is −1 and so 2t = −1 which
gives t = − 12 .
We now need to find the point of intersection of the lines with equations y = 2x + 7 and
y = − 21 x − 12 .
Equating expressions for y, we obtain 2x + 7 = − 12 x − 12 or 52 x = − 15
2
, which gives x = −3.
Therefore, y = 2x + 7 = 2(−3) + 7 = 1, and so the point of intersection of these lines is
(−3, 1).

3. (a) Since 64 = 26 , its positive divisors are 1, 2, 4, 8, 16, 32, and 64.
The sum of these divisors is 1 + 2 + 4 + 8 + 16 + 32 + 64 = 127.
(b) Suppose that the four consecutive integers that Fionn originally wrote on the blackboard
were x, x + 1, x + 2, and x + 3.
When Lexi erases one of these integers, the sum of the remaining three integers is equal
to one of the following:

(x + 1) + (x + 2) + (x + 3) = 3x + 6
x + (x + 2) + (x + 3) = 3x + 5
x + (x + 1) + (x + 3) = 3x + 4
x + (x + 1) + (x + 2) = 3x + 3

We are told that the sum of these integers is 847.

We note that 847 = 3 · 282 + 1, which is one more than a multiple of 3. Since 3x + 3 and
3x + 6 are always multiples of 3 and 3x + 5 is 2 more than a multiple of 3, then we must
have 3x + 4 = 847 and so 3x = 843 or x = 281. (Alternatively, we could have set each
of the four sums above equal to 847 to determine in which case or cases we obtained an
integer solution for x.)
Therefore, the original integers were 281, 282, 283, 284 and Lexi erased x + 2 = 283.
(c) From the given information, the 7 terms in the arithmetic sequence are

d2 , d2 + d, d2 + 2d, d2 + 3d, d2 + 4d, d2 + 5d, d2 + 6d

Since the sum of these 7 terms is 756, we obtain the following equivalent equations:

d2 + (d2 + d) + (d2 + 2d) + (d2 + 3d) + (d2 + 4d) + (d2 + 5d) + (d2 + 6d) = 756
7d2 + 21d = 756
d2 + 3d = 108
d2 + 3d − 108 = 0
(d + 12)(d − 9) = 0

and so d = −12 or d = 9.
The corresponding arithmetic sequences are

144, 132, 120, 108, 96, 84, 72 and 81, 90, 99, 108, 117, 126, 135
2023 Euclid Contest Solutions Page 4

4. (a) In 1 hour, Liang paints 13 of the room.

Thus, in 2 hours, Liang paints 23 of the room.
Edmundo needs to paint 1 − 23 = 13 of the room.
In 1 hour, Edmundo paints 14 of the room.
3
Since 14 = 12 and 31 = 12
4
, this means that Edmundo paints for 1
3
÷ 1
4
= 4
12
÷ 3
12
= 4
3
of an
hour.
Therefore, Edmundo paints for 80 minutes.
A
(b) When converted to a fraction, A% is equal to .
100
A
When an amount is increased by A%, we can find its new value by multiplying by 1 + .
100
A
When an amount is decreased by A%, we can find its new value by multiplying by 1− .
  100
A
When $400 is increased by A%, the amount becomes $400 1 + .
100  
A A
When this value is decreased by A%, the amount becomes $400 1 + 1− .
100 100
Therefore,
  
A A
$400 1 + 1− = $391
100 100
  
A A 391
1+ 1− =
100 100 400
2
A 9
1− 2
=1−
100 400
A2 9
=
1002 400
2
A 32
=
1002 202
A 3
= (since A > 0)
100 20
3
A = 100 · = 15
20
Therefore, A = 15.
2023 Euclid Contest Solutions Page 5

5. (a) The quadratic function f (x) = x2 + (2n − 1)x + (n2 − 22) has no real roots exactly when
its discriminant, ∆, is negative.
The discriminant of this function is

∆ = (2n − 1)2 − 4(1)(n2 − 22)

= (4n2 − 4n + 1) − (4n2 − 88)
= −4n + 89

We have ∆ < 0 exactly when −4n + 89 < 0 or 4n > 89.

This final inequality is equivalent to n > 89
4
= 22 14 .
Therefore, the smallest positive integer that satisfies this inequality, and hence for which
f (x) has no real roots, is n = 23.
(b) Using the cosine law in 4P QR,

P R2 = P Q2 + QR2 − 2 · P Q · QR · cos(∠P QR)

212 = a2 + b2 − 2ab cos(60◦ )
441 = a2 + b2 − 2ab · 12
441 = a2 + b2 − ab

ST TU a b
Using the sine law in 4ST U , we obtain = and so = .
sin(∠T U S) sin(∠T SU ) 4/5 sin(30◦ )
a b
Therefore, = and so a = 54 · 2b = 85 b.
4/5 1/2
Substituting into the previous equation,

2
441 = 85 b + b2 − 85 b b

441 = 6425
b2 + b2 − 85 b2
441 = 6425
b2 + 2525
b2 − 40
25
b2
441 = 4925
b2
225 = b2

Since b > 0, then b = 15 and so a = 85 b = 8

5
· 15 = 24.
2023 Euclid Contest Solutions Page 6

6. (a) Solution 1
We make two copies of the given triangle, labelling them 4ABC and 4DEF , as shown:

A D

B C E F

The combined area of these two triangles is 2 · 770 cm2 = 1540 cm2 , and the shaded area
in each triangle is the same.
Next, we rotate 4DEF by 180◦ :

A
F E

B C D

and join the two triangles together:

A E

B C

We note that BC and AE (which was F E) are equal in length (since they were copies of
each other) and parallel (since they are 180◦ rotations of each other). The same is true
for AB and EC.
Therefore, ABCE is a parallelogram.
Further, ABCE is divided into 11 identical parallelograms (6 shaded and 5 unshaded)
by the horizontal lines. (Since the sections of the two triangles are equal in height, the
horizontal lines on both sides of AC align.)
The total area of parallelogram ABCE is 1540 cm2 .
6
Thus, the shaded area of ABCE is 11 · 1540 cm2 = 840 cm2 .
Since this shaded area is equally divided between the two halves of the parallelogram, then
the combined area of the shaded regions of 4ABC is 21 · 840 cm2 = 420 cm2 .
2023 Euclid Contest Solutions Page 7

Solution 2
We label the points where the horizontal lines touch AB and AC as shown:
A
B1 C1
B2 C2
B3 C3
B4 C4
B5 C5
B6 C6
B7 C7
B8 C8
B9 C9
B10 C10

B C

We use the notation |4ABC| to represent the area of 4ABC and use similar notation
for the area of other triangles and quadrilaterals.
Let A be equal to the total area of the shaded regions.
Thus,

A = |4AB1 C1 | + |B2 B3 C3 C2 | + |B4 B5 C5 C4 | + |B6 B7 C7 C6 | + |B8 B9 C9 C8 | + |B10 BCC10 |

The area of each of these quadrilaterals is equal to the difference of the area of two triangles.
For example,

|B2 B3 C3 C2 | = |4AB3 C3 | − |4AB2 C2 | = −|4AB2 C2 | + |4AB3 C3 |

Therefore,

A = |4AB1 C1 | − |4AB2 C2 | + |4AB3 C3 | − |4AB4 C4 | + |4AB5 C5 |

− |4AB6 C6 | + |4AB7 C7 | − |4AB8 C8 | + |4AB9 C9 | − |4AB10 C10 | + |4ABC|

Each of 4AB1 C1 , 4AB2 C2 , . . ., 4AB10 C10 is similar to 4ABC because their two base
angles are equal due.
Suppose that the height of 4ABC from A to BC is h.
Since the height of each of the 11 regions is equal in height, then the height of 4AB1 C1
1 2
is 11 h, the height of 4AB2 C2 is 11 h, and so on.
When two triangles are similar, their heights are in the same ratio as their side lengths:
To see this, suppose that 4P QR is similar to 4ST U and that altitudes are drawn
from P and S to V and W .
P
S

Q V R T W U
Since ∠P QR = ∠ST U , then 4P QV is similar to 4ST W (equal angle; right
PQ PV
angle), which means that = . In other words, the ratio of sides is equal
ST SW
to the ratio of heights.
1 1
Since the height of 4AB1 C1 is 11
h, then B1 C1 = 11 BC.
12 12
Therefore, |4AB1 C1 | = 21 · B1 C1 · 11
1
h = 12 · 11
1 1
BC · 11 h = 11 1
2 · 2 · BC ·h= 112
|4ABC|.
2 2
Similarly, since the height of 4AB2 C2 is 11 h, then B2 C2 = 11 BC.
2023 Euclid Contest Solutions Page 8
1 2 1 2 2 22 22
Therefore, |4AB2 C2 | = 2
· B2 C2 · 11
h = 2
· 11
BC · 11
h = 112
· 21 · BC · h = 112
|4ABC|.
This result continues for each of the triangles.
Therefore,
12 22 32 42 52
A= 112
|4ABC| − 11 2 |4ABC| + 112 |4ABC| − 112 |4ABC| + 112 |4ABC|

62 72 82 92 102 112
− 11 2 |4ABC| + 112 |4ABC| − 112 |4ABC| + 112 |4ABC| − 112 |4ABC| + 112 |4ABC|

= 1
112
|4ABC|(112 − 102 + 92 − 82 + 72 − 62 + 52 − 42 + 32 − 22 + 1)
= 1
112
(770 cm2 )((11 + 10)(11 − 10) + (9 + 8)(9 − 8) + · · · + (3 + 2)(3 − 2) + 1)
= 1
112
(770 cm2 )(11 + 10 + 9 + 8 + 7 + 6 + 5 + 4 + 3 + 2 + 1)
= 1
11
(70 cm2 ) · 66
2
= 420 cm

Therefore, the combined area of the shaded regions of 4ABC is 420 cm2 .
(b) Solution 1
We label five additional points in the diagram:

P Q R S
D A
T
C
B

Since P Q = QR = RS = 1, then P√S = 3 and P R =√2. √

Since ∠P ST = 90◦ , then P T = P S 2 + ST 2 = 32 + 12 = 10 by the Pythagorean
Theorem.
We are told that ABCD is a square.
Thus, P T is perpendicular to QC and to RB.
Thus, 4P DQ is right-angled at D and 4P AR is right-angled at A.
Since 4P DQ, 4P AR and 4P ST are all right-angled and all share an angle at P , then
these three triangles are similar.
PA PR 3·2 PD PQ 1·3
This tells us that = and so P A = √ . Also, = and so P D = √ .
PS PT 10 PS PT 10
Therefore,
6 3 3
DA = P A − P D = √ − √ = √
10 10 10
 2
3 9
This means that the area of square ABCD is equal to DA2 = √ = .
10 10
2023 Euclid Contest Solutions Page 9

Solution 2
We add coordinates to the diagram as shown:
(0, 3) (1, 3) (2, 3) (3, 3)

D A
(0, 2) (3, 2)
C
(0, 1) B (3, 1)

(0, 0) (1, 0) (2, 0) (3, 0)

We determine the side length of square ABCD by determining the coordinates of D and
A and then calculating the distance between these points.
3−2 1
The slope of the line through (0, 3) and (3, 2) is =− .
0−3 3
1
This equation of this line can be written as y = − x + 3.
3
The slope of the line through (0, 0) and (1, 3) is 3.
The equation of this line can be written as y = 3x.
The slope of the line through (1, 0) and (2, 3) is also 3.
The equation of this line can be written as y = 3(x − 1) = 3x − 3.
1
Point D is the intersection point of the lines with equations y = − x + 3 and y = 3x.
3
1 10 9
Equating expressions for y, we obtain − x + 3 = 3x and so x = 3 which gives x = .
3 3  10
27 9 27
Since y = 3x, we get y = and so the coordinates of D are , .
10 10 10
1
Point A is the intersection point of the lines with equations y = − x + 3 and y = 3x − 3.
3
1 10
Equating expressions for y, we obtain − x + 3 = 3x − 3 and so x = 6 which gives
3 3
18
x= .
10  
24 18 24
Since y = 3x − 3, we get y = and so the coordinates of A are , . (It is easier
10 10 10
to not reduce these fractions.)
Therefore,
s 2  2 s 2  2 r r
9 18 27 24 9 3 90 9
DA = − + − = − + = =
10 10 10 10 10 10 100 10
r !2
9 9
This means that the area of square ABCD is equal to DA2 = = .
10 10
2023 Euclid Contest Solutions Page 10

7. (a) Each possible order in which Akshan removes the marbles corresponds to a sequence of 9
colours, 3 of which are red and 6 of which are blue.
We write these as sequences of 3 R’s and 6 B’s.
Since are told that the first marble is red and the third is blue, we would like to consider
all sequences of the form

R B

The 7 blanks
 must be filled with the remaining 2 R’s and 5 B’s.
7 7·6
There are = = 21 ways of doing this, because 2 of the 7 blanks must be chosen in
2 2
which to place the R’s. (We could count these 21 ways directly by working systematically
through the possible pairs of blanks.)
Of these 21 ways, some have the last two marbles being blue.
These correspond to the sequences of the form

R B BB

In these sequences,
  the 5 blanks must be filled with the remaining 2 R’s and 3 B’s.
5 5·4
There are = = 10 ways of doing this, because 2 of the 5 blanks must be chosen
2 2
in which to place the R’s.
Therefore, 10 of the 21 possible sequences end in two B’s, and so the probability that the
10
last two marbles removed are blue is .
21
(b) Factoring the first equation, we obtain

ac + ad + bc + bd = a(c + d) + b(c + d) = (a + b)(c + d)

We now have the equations

(a + b)(c + d) = 2023
(a + b) + (c + d) = 296

If we let s = a + b and t = c + d, we obtain the equations

st = 2023
s + t = 296

Noting that s and t are integers since a, b, c, and d are integers, we look for divisor pairs
of 2023 whose sum is 296.
To find the divisors of 2023, we first find its prime factorization:

2023 = 7 · 289 = 7 · 172

Therefore, the divisors of 2023 are 1, 7, 17, 119, 289, 2023.

This means that the divisor pairs of 2023 are

2023 = 1 · 2023 = 7 · 289 = 17 · 119

The one divisor pair with a sum of 296 is 7 and 289. (Alternatively, we could have found
these by substituting t = 206 − s into st = 2023 and using the quadratic formula.)
2023 Euclid Contest Solutions Page 11

Since a < b < c < d, then a + b < c + d and so s = a + b = 7 and t = c + d = 289.

Since a and b are positive integers with a < b and a + b = 7, then the possible pairs (a, b)
are
(a, b) = (1, 6), (2, 5), (3, 4)
We know that c and d are positive integers with c < d and c + d = 289, but also with
b < c < d.
When (a, b) = (1, 6), this means that the possibilities are

(c, d) = (7, 282), (8, 281), (9, 280), . . . , (143, 146), (144, 145)

There are 144 − 7 + 1 = 138 such pairs.

When (a, b) = (2, 5), the possibilities are

(c, d) = (6, 283), (7, 282), (8, 281), (9, 280), . . . , (143, 146), (144, 145)

There are 138 + 1 = 139 such pairs.

When (a, b) = (3, 4), the possibilities are

(c, d) = (5, 284), (6, 283), (7, 282), (8, 281), (9, 280), . . . , (143, 146), (144, 145)

There are 139 + 1 = 140 such pairs.

In total, there are 138 + 139 + 140 = 417 possible quadruples (a, b, c, d).

8. (a) Since 4ABC is right-angled at B, then

BC 2 = AC 2 − AB 2
= ((n + 1)(n + 4))2 − (n(n + 1))2
= (n + 1)2 (n + 4)2 − n2 (n + 1)2
= (n + 1)2 (n + 4)2 − n2

= (n + 1)2 n2 + 8n + 16 − n2

= (n + 1)2 (8n + 16)

= 4(n + 1)2 (2n + 4)

The length of BC is an integer exactly when 4(n + 1)2 (2n + 4) is a perfect square.
Since 4(n + 1)2 is a perfect square, then BC is an integer exactly when 2n + 4 is a perfect
square.
We note that 2n + 4 ≥ 6 (since n ≥ 1) and that 2n + 4 is even.
Since n < 100 000, then 6 ≤ 2n + 4 < 200 004, and so we need to count the number of
even perfect squares between 6 and 200 004.
The smallest
√ even perfect square in this range is 42 = 16.
Since 200 004 ≈ 447.2, the largest even perfect square in this range is 4462 .
446
Therefore, the number of even perfect squares in this range is − 1 = 222.
2
Thus, there are 222 positive integers n for which the length of BC is an integer.
2023 Euclid Contest Solutions Page 12
r x
p
(b) Let f (x) = log2 x · log2 (4x) + 1 + log2 x · log2 + 9.
64
Using logarithm laws,

log2 x · log2 (4x) + 1 = log2 x(log2 4 + log2 x) + 1

= log2 x(2 + log2 x) + 1 (since 22 = 4)
= (log2 x)2 + 2 · log2 x + 1
= (log2 x + 1)2

and
x
log2 x · log2 + 9 = log2 x(log2 x − log2 64) + 9
64
= log2 x(log2 x − 6) + 9 (since 26 = 64)
= (log2 x)2 − 6 log2 x + 9
= (log2 x − 3)2

Therefore,
r x
p p p
f (x) = log2 x · log2 (4x) + 1+ log2 x · log2 + 9 = (log2 x + 1)2 + (log2 x − 3)2
64
√ √
Before proceeding, we recall that if a ≤ 0, then a2 = −a and if a > 0, then a2 = a.
When log2 x ≤ −1, we know that log2 x + 1 ≤ 0 and log2 x − 3 < 0, and so

f (x) = −(log2 x + 1) − (log2 x − 3) = 2 − 2 log2 x

When −1 < log2 x ≤ 3, we know that log2 x + 1 > 0 and log2 x − 3 ≤ 0, and so

f (x) = (log2 x + 1) − (log2 x − 3) = 4

When log2 x > 3, we know that log2 x + 1 ≥ 0 and log2 x − 3 > 0, and so

f (x) = (log2 x + 1) + (log2 x − 3) = 2 log2 x − 2

We want to find all values of x for which f (x) = 4.

When log2 x ≤ −1, f (x) = 2 − 2 log2 x = 4 exactly when log2 x = −1.
When −1 < log2 x ≤ 3, f (x) is always equal to 4.
When log2 x > 3, f (x) = 2 log2 x − 2 = 4 exactly when log2 x = 3.
Therefore, f (x) = 4 exactly when −1 ≤ log2 x ≤ 3, which is true exactly when 21 ≤ x ≤ 8.
(It seems surprising that the solution to this equation is actually an interval of values,
rather than a finite number of specific values.)
2023 Euclid Contest Solutions Page 13

9. (a) If there are 5 or more people seated around a table with 8 chairs, then there are at most
3 empty chairs. But there must be an empty chair between each pair of people, and this
is not possible with 5 people and 3 empty chairs.
Therefore, there are at most 4 people seated.
If there were only 2 people seated, then there would be 6 empty chairs which would mean
that at least one of the two “gaps” around the circular table had at least 3 empty chairs,
and so another person could be seated, meaning that the table wasn’t full.
Therefore, there are at least 3 people seated.
This means that a full table with 8 chairs has either 3 or 4 people.
If there are 4 people, there are 4 empty chairs, and so there is exactly 1 empty chair
between each pair of people.
Thus, people are seated in chairs {1, 3, 5, 7} or in chairs {2, 4, 6, 8}.
If there are 3 people, there are 5 empty chairs.
With 3 people, there are 3 gaps totalling 5 chairs, and each gap has at most 2 chairs in it.
Therefore, the gaps must be 1, 2, 2 in some order. This is the only list of three positive
integers, each equal to 1 or 2, that adds to 5.
The gap of 1 can be between any pair of seats. In other words, the gap of 1 could
be between {1, 3}, {2, 4}, and so on. In each case, the position of the third person is
completely determined because the remaining two gaps have 2 chairs each.
Thus, with 3 people, they are seated in chairs

{1, 3, 6}, {2, 4, 7}, {3, 5, 8}, {4, 6, 1}, {5, 7, 2}, {6, 8, 3}, {7, 1, 4}, {8, 2, 5}

In total, there are thus 10 ways to seat people at a table with 8 chairs:

{1, 3, 5, 7}, {2, 4, 6, 8}, {1, 3, 6}, {2, 4, 7}, {3, 5, 8}, {4, 6, 1}, {5, 7, 2}, {6, 8, 3}, {7, 1, 4}, {8, 2, 5}

(b) Suppose that k is a positive integer.

Suppose that t people are seated at a table with 6k + 5 chairs so that the table is full.
When t people are seated, there are t gaps. Each gap consists of either 1 or 2 chairs. (A
gap with 3 or more chairs can have an additional person seated in it, so the table is not
full.)
Therefore, there are between t and 2t empty chairs.
This means that the total number of chairs is between t + t and t + 2t.
In other words, 2t ≤ 6k + 5 ≤ 3t.
Since 2t ≤ 6k + 5, then t ≤ 3k + 25 . Since k and t are integers, then t ≤ 3k + 2.
We note that it is possible to seat 3k + 2 people around the table in seats

{2, 4, 6, . . . , 6k + 2, 6k + 4}

This table is full becase 3k + 1 of the gaps consist of 1 chair and 1 gap consists of 2 chairs.
Since 3t ≥ 6k + 5, then t ≥ 2k + 35 . Since k and t are integers, then t ≥ 2k + 2.
We note that it is possible to seat 2k + 2 people around the table in seats

{3, 6, 9, . . . , 6k, 6k + 3, 6k + 5}

This table is full becase 2k + 1 of the gaps consist of 2 chairs and 1 gap consists of 1 chair.
We now know that, if there are t people seated at a full table with 6k + 5 chairs, then
2k + 2 ≤ t ≤ 3k + 2.
To confirm that every such value of t is possible, consider a table with t people, 3t−(6k+5)
2023 Euclid Contest Solutions Page 14

gaps of 1 chair, and (6k + 5) − 2t gaps of 2 chairs.

From the work above, we know that 3t ≥ 6k + 5 and so 3t − (6k + 5) ≥ 0, and that
2t ≤ 6k + 5 and so (6k + 5) − 2t ≥ 0.
The total number of gaps is 3t − (6k + 5) + (6k + 5) − 2t = t, since there are t people
seated.
Finally, the total number of chairs is

t + 1 · (3t − (6k + 5)) + 2 · ((6k + 5) − 2t) = t + 3t − 4t − (6k + 5) + 2(6k + 5) = 6k + 5

as expected.
This shows that every t with 2k + 2 ≤ t ≤ 3k + 2 can produce a full table.
Therefore, the possible values of t are those integers that satisfy 2k + 2 ≤ t ≤ 3k + 2.
There are (3k + 2) − (2k + 2) + 1 = k + 1 possible values of t.
(c) Solution 1
For each integer n ≥ 3, we define f (n) to be the number of different full tables of size n.
We can check that
• f (3) = 3 because the full tables when n = 3 have people in chairs {1}, {2}, {3},
• f (4) = 2 because the full tables when n = 4 have people in chairs {1, 3}, {2, 4}, and
• f (5) = 5 because the full tables when n = 4 have people in chairs {1, 3}, {2, 4}, {3, 5},
{4, 1}, {5, 2}.
In the problem, we are told that f (6) = 5 and in part (a), we determined that f (8) = 10.
This gives us the following table:
n f (n)
3 3
4 2
5 5
6 5
7 ?
8 10
Based on this information, we make the guess that for every integer n ≥ 6, we have
f (n) = f (n − 2) + f (n − 3).
For example, this would mean that f (7) = f (5) + f (4) = 5 + 2 = 7 which we can verify is
true.
Based on this recurrence relation (which we have yet to prove), we deduce the values of
f (n) up to and including n = 19:

n f (n)
n f (n)
11 22
3 3
12 29
4 2
13 39
5 5
14 51
6 5
15 68
7 7
16 90
8 10
17 119
9 12
18 158
10 17
19 209

We now need to prove that the equation f (n) = f (n − 2) + f (n − 3) is true for all n ≥ 6.
2023 Euclid Contest Solutions Page 15

We think about each full table as a string of 0s and 1s, with 1 representing a chair that is
occupied and 0 representing an empty chair.
Let a(n) be the number of full tables with someone in seat 1 (and thus nobody in seat 2).
Let b(n) be the number of full tables with someone in seat 2 (and thus nobody in seat 1).
Let c(n) be the number of full tables with nobody in seat 1 or in seat 2.
Since every full table must be in one of these categories, then f (n) = a(n) + b(n) + c(n).
A full table with n seats n ≥ 4 must correspond to a string that starts with 10, 01 or 00.
Since there cannot be more than two consecutive 0s, we can further specify this, namely
to say that a full table with n seats must correspond to a string that starts with 1010 or
1001 or 0100 or 0101 or 0010. In each case, these are the first 4 characters of the string
and correspond to full (1) and empty (0) chairs.
Consider the full tables starting with 1010. Note that such strings end with 0 since the
table is circular. Removing the 10 from positions 1 and 2 creates strings of length n − 2
that begin 10. These strings will still correspond to a full table, and so there are a(n − 2)
such strings. (We note that all possible strings starting 1010 of length n will lead to all
possible strings starting with 1010 of length n − 2.)
Consider the full tables starting with 1001. Note that such a string ends with 0 since the
table is circular. Removing the 100 from positions 1, 2 and 3 creates strings of length
n − 3 that begin 10. (There must have been a 0 in position 5 after the 1 in position 4.)
These strings will still correspond to full tables, and so there are a(n − 3) such strings.
Consider the full tables starting with 0100. Removing the 100 from positions 2, 3 and 4
creates strings of length n − 3 that begin 01. (There must have been a 1 in position 5
after the 0 in position 4.) These strings will still correspond to full tables, and so there
are b(n − 3) such strings.
Consider the full tables starting with 0101. Removing the 01 from positions 3 and 4 creates
strings of length n − 2 that begin 01. (The 1 in position 4 must have been followed by one
or two 0s and so these strings maintains the desired properties.) These strings will still
correspond to full tables, and so there are b(n − 2) such strings.
Consider the full tables starting with 0010. These strings must begin with either 00100 or
00101.
If strings start 00100, then they start 001001 and so we remove the 001 in positions 4, 5
and 6 and obtain strings of length n − 3 that start 001 (and thus start 00). There are
c(n − 3) such strings.
If strings start 00101, we remove the 01 in positions 4 and 5 and obtain strings of length
n − 2 that start 001 (and thus start 00). There are c(n − 2) such strings.
These 6 cases and subcases count all strings counted by f (n).
Therefore,
f (n) = a(n − 2) + a(n − 3) + b(n − 3) + b(n − 2) + c(n − 3) + c(n − 2)
= a(n − 2) + b(n − 2) + c(n − 2) + a(n − 3) + b(n − 3) + c(n − 3)
= f (n − 2) + f (n − 3)
as required, which means that the number of different full tables when n = 19 is 209.

Solution 2
Extending our approach from (b), the number of people seated at a full table with 19
chairs is at least 19
3
= 6 13 and at most 19
2
= 9 12 .
Since the number of people is an integer, there must be 7, 8 or 9 people at the table, which
means that the number of empty chairs is 12, 11 or 10, respectively.
2023 Euclid Contest Solutions Page 16

Suppose that there are 9 people and 9 gaps with a total of 10 empty chairs.
In this case, there is 1 gap with 2 empty chairs and 8 gaps with 1 empty chair.
There are 19 pairs of chairs in which we can put 2 people with a gap of 2 in between:
{1, 4}, {2, 5}, . . ., {19, 3}.
Once we choose one of these pairs, the seat choice for the remaining 8 people is completely
determined by placing people in every other chair.
Therefore, there are 19 different full tables with 9 people.
Suppose that there are 8 people and 8 gaps with a total of 11 empty chairs.
In this case, there are 3 gaps with 2 empty chairs and 5 gaps with 1 empty chair.
There are 7 different circular orderings in which these 8 gaps can be arranged:

22211111 22121111 22112111 22111211 22111121 21212111 21211211

We note that “22211111” would be the same as, for example, “11222111” since these gaps
are arranged around a circle.
If the three gaps of length 2 are consecutive, there is only one configuration (22211111).
If there are exactly 2 consecutive gaps of length 2, there are 4 relative places in which the
third gap of length 2 can be placed.
If there are no consecutive gaps of length 2, these gaps can either be separated by 1 gap
each (21212111) with 3 gaps on the far side, or can be separated by 1 gap, 2 gaps, and 2
gaps (21211211). There is only one configuration for the gaps in this last situation.
There are 7 different circular orderings for these 8 gaps.
Each of these 7 different orderings can be placed around the circle of 19 chairs in 19 dif-
ferent ways, because each can be started in 19 different places. Because 19 is prime, none
of these orderings overlap.
Therefore, there are 7 · 19 = 133 different full tables with 8 people.
Suppose that there are 7 people and 7 gaps with a total of 12 empty chairs.
In this case, there are 2 gaps with 1 empty chair and 5 gaps with 2 empty chairs.
The 2 gaps with 1 empty chair can be separated by 0 gaps with 2 empty chairs, 1 gap with
2 empty chairs, or 2 gaps with 2 empty chairs. Because the chairs are around a circle, if
there were 3, 4 or 5 gaps with 2 empty chairs between them, there would be 2, 1 or 0 gaps
going the other way around the circle.
This means that there are 3 different configurations for the gaps.
Each of these configurations can be placed in 19 different ways around the circle of chairs.
Therefore, there are 3 · 19 = 57 full tables with 7 people.
In total, there are 19 + 133 + 57 = 209 full tables with 19 chairs.

Solution 3
As in Solution 2, there must be 7, 8 or 9 people in chairs, and so there are 7, 8 or 9 gaps.
If there are 7 gaps, there are 2 gaps of 1 chair and 5 gaps of 2 chairs.
If there are 8 gaps, there are 5 gaps of 1 chair and 3 gaps of 2 chairs.
If there are 9 gaps, there are 8 gaps of 1 chair and 1 gap of 2 chairs.
We consider three mutually exclusive cases: (i) there is a person in chair 1 and not in chair
2, (ii) there is a person in chair 2 and not in chair 1, and (iii) there is nobody in chair 1
or in chair 2. Every full table fits into exactly one of these three cases.

Case (i): there is a person in chair 1 and not in chair 2

We use the person in chair 1 to “anchor” the arrangement, by starting at chair 1 and
2023 Euclid Contest Solutions Page 17

arranging the gaps (and thus the full chairs) clockwise around the table from chair 1. 
7
If there are 7 gaps, we need to choose 2 of them to be of length 1, and so there are
2
ways of arranging the gaps starting at chair 1.  
8
If there are 8 gaps, we need to choose 3 of them to be of length 2, and so there are
3
ways of arranging the gaps starting at chair 1.  
9
If there are 9 gaps, we need to choose 1 of them to be of length 2, and so there are
1
ways of arranging the gaps starting at chair
      1.
7 8 9
In this case, there are a total of + + = 21 + 56 + 9 = 86 full tables.
2 3 1
Case (ii): there is a person in chair 2 and not in chair 1
We use the same reasoning starting with the person in chair 2 as the anchor.
Again, there are 86 full tables in this case.

Case (iii): there is nobody in chair 1 or chair 2

Since there is nobody in chair 1 or chair 2, there must be a person in chair 3 and also in
chair 19, which fixes one gap of 2 chairs.
Here, we use the person in chair 3 as the anchor.
If there
 are 7 gaps, there are 2 gaps of 1 chair and 4 gaps of 2 chairs left to place. There
6
are ways of doing this.
2
If there
 are 8 gaps, there are 5 gaps of 1 chair and 2 gaps of 2 chairs left to place. There
7
are ways of doing this.
2
If there are 9 gaps, there are 8 gaps of 1 chair and 0 gaps of 2 chairs left to place. There
is 1 way to do this.    
6 7
In this case, there are a total of + + 1 = 15 + 21 + 1 = 37 full tables.
2 2
In total, there are 86 + 86 + 37 = 209 full tables with 19 chairs.

   
1 2 1 2
10. (a) Since 0 < < < 1, then = = 0.
3 3 3 3
     
3 4 5 3 4 5
Since 1 ≤ < < < 2, then = = = 1.
3 3 3 3 3 3
These fractions can continue to be grouped in groups of 3with the last full group of 3
57 58 59 57 58 59
satisfying 19 ≤ < < < 20, which means that = = = 19.
3 3 3 3 3 3
60
The last term is = b20c = 20.
3
2023 Euclid Contest Solutions Page 18

If the given sum is S, we obtain

S = 2 · 0 + 3 · 1 + 3 · 2 + · · · + 3 · 19 + 1 · 20
= 0 + 3(1 + 2 + · + 19) + 20
= 3 · 12 · 19 · 20 + 20
= 570 + 20
= 590

(b) For every positive integer m > 4, let

         
1 2 3 m−2 m−1
q(m) = + + + ... + +
3 3 3 3 3
3k − 3 3k − 2 3k − 1
Extending our work from (a), we know that k − 1 ≤ < < < k for
    3 3 3
3k − 3 3k − 2 3k − 1
each positive integer k, and so = = = k − 1.
3 3 3
Every positive integer m > 4 can be written as m = 3s or m = 3s + 1 or m = 3s + 2, for
some positive integer s, depending on its remainder when divided by 3.
We can thus write
         
1 2 3 3s − 2 3s − 1
q(3s) = + + + ... + +
3 3 3 3 3
= 2 · 0 + 3(1 + 2 + 3 + · · · + (s − 1))
1
= 3 · · (s − 1)s
2
3s(s − 1)
=
2
3s(3s − 3)
=
6
           
1 2 3 3s − 2 3s − 1 3s
q(3s + 1) = + + + ... + + +
3 3 3 3 3 3
= q(3s) + s
3s(3s − 3) 3s · 2
= +
6 6
3s(3s − 1)
=
6
 
3s + 1
q(3s + 2) = q(3s + 1) +
3
3s(3s − 1)
= +s
6
3s(3s − 1) 3s · 2
= +
6 6
3s(3s + 1)
=
6
We want to find a polynomial p(x) for which q(m) = bp(m)c for every positive integer
m > 4.
2023 Euclid Contest Solutions Page 19

In other words, we want to find a polynomial p(x) for which

3s(3s − 3) 3s(3s − 1) 3s(3s + 1)

bp(3s)c = bp(3s + 1)c = bp(3s + 2)c =
6 6 6
for every positive integer s.
(x − 1)(x − 2)
We will show that the polynomial p(x) = satisfies the desired conditions.
6
If x = 3s + 1 for some positive integer s, then

(x − 1)(x − 2) (3s + 1 − 1)(3s + 1 − 2) 3s(3s − 1)

= =
6 6 6
We note that 3s is a multiple of 3. Since 3s and 3s − 1 are consecutive integers, then one
3s(3s − 1)
of these is a multiple of 2. Thus, 3s(3s − 1) is a multiple of 6 and so is an
6
integer.  
3s(3s − 1) 3s(3s − 1)
This means that = .
6 6
 
3s(3s − 1) 3s(3s − 1)
Therefore, q(3s + 1) = = = bp(3s + 1)c.
6 6
If x = 3s + 2 for some positive integer s, then

(x − 1)(x − 2) (3s + 2 − 1)(3s + 2 − 2) 3s(3s + 1)

= =
6 6 6
We note that 3s is a multiple of 3. Since 3s and 3s + 1 are consecutive integers, then one
3s(3s + 1)
of these is a multiple of 2. Thus, 3s(3s + 1) is a multiple of 6 and so is an
6
integer.  
3s(3s + 1) 3s(3s + 1)
This means that = .
6 6
 
3s(3s + 1) 3s(3s + 1)
Therefore, q(3s + 2) = = = bp(3s + 2)c.
6 6
If x = 3s for some positive integer s, then

(x − 1)(x − 2) (3s − 1)(3s − 2) 9s2 − 9s + 2

= =
6 6 6
9s2 − 9s 9s(s − 1)
Now, = is an integer because 9s is a multiple of 3 and one of s and
6 6
s − 1 is even.
9s2 − 9s + 2 9s2 − 9s 1 9s2 − 9s + 2 1
Since = + , then is more than an integer which
6 6 3 6 3
2 2

9s − 9s + 2 9s − 9s 3s(3s − 3)
means that = = = q(3s).
6 6 6
 
3s(3s − 3) (3s − 1)(3s − 2)
Therefore, q(3s) = = = bp(3s)c.
6 6
(x − 1)(x − 2)
This means that the polynomial p(x) = satisfies the required conditions.
6
2023 Euclid Contest Solutions Page 20

(c) Before working on the specific question we have been asked, we simplify the given expres-
sion for f (n) by noting that if k ≥ n, then kn ≤ k 2 < k 2 +1. 
kn kn
This means that if k ≥ n, we have 0 < 2 < 1 and so = 0.
n +1 k2 + 1
This means that, for a fixed positive integer n, the apparently infinite sum that represents
f (n) can be stopped when k = n − 1 because every subsequent term equals 0.
Thus,
         
n 2n 3n (n − 2)n (n − 1)n
f (n) = 2 + 2 + 2 + ··· + +
1 +1 2 +1 3 +1 (n − 2)2 + 1 (n − 1)2 + 1

We note that

f (1) = 0 (since no terms are non-zero)

 
1·2
f (2) = 2 =1
1 +1
       
1·3 2·3 3 6
f (3) = 2 + 2 = + =1+1=2
1 +1 2 +1 2 5
           
1·4 2·4 3·4 4 8 12
f (4) = 2 + 2 + 2 = + + =2+1+1=4
1 +1 2 +1 3 +1 2 5 10

Suppose that t is an odd positive integer for which f (t + 1) − f (t) = 2.

We will assume that t is not a prime number, and show that f (t + 1) − f (t) 6= 2. This will
show us that if f (t + 1) − f (t) = 2, it must be the case that t is prime. Since t is odd and
not prime, then t = 1 or t is composite.
We note that when t = 1, we obtain f (2) − f (1) = 1 − 0 = 1 6= 2.
Next, suppose that t is odd and composite.
Since t is odd and composite, then t can be written as t = rs for some odd positive integers
r ≥ s > 1. (t can be written in this form in at least one way, so we take one of these
possibilities.)
In this case, consider f (t + 1) − f (t).
We can write this as
       
t+1 2(t + 1) (t − 1)(t + 1) t(t + 1)
f (t + 1) − f (t) = + +··· + +
12 + 1 22 + 1 (t − 1)2 + 1 t2 + 1
     
t 2t (t − 1)t
− 2 − 2 −··· −
1 +1 2 +1 (t − 1)2 + 1

We re-write this as
       
t+1 t 2(t + 1) 2t
f (t + 1) − f (t) = − 2 + − 2 + ···
12 + 1 1 +1 22 + 1 2 +1
     
(t − 1)(t + 1) (t − 1)t t(t + 1)
+ − +
(t − 1)2 + 1 (t − 1)2 + 1 t2 + 1
   
k(t + 1) kt
In the t − 1 sets of parentheses, we have terms of the form − 2 for each
k2 + 1 k +1
integer k from 1 to t − 1.
k(t + 1) kt
We know that 2 > 2 because both k and t are positive, the denominators are
k +1 k +1
2023 Euclid Contest Solutions Page 21

equal and k(t + 1) > kt.

   
k(t + 1) kt
Thus, ≥ . (The greatest integer less than or equal to the first fraction
k2 + 1 k2 + 1
must be at least as large as the greatest integer less than or equal to the second fraction.)
This means that the t − 1 differences in parentheses, each of which is an integer, is at
least 0.
To show that f (t+1)−f (t) 6= 2, we show that there are at least 2 places where the difference
is at least 1, and that the final term is at least 1. This will tell us that f (t + 1) − f (t) ≥ 3
and so f (t + 1) − f (t) 6= 2, which will tell us that t cannot be composite, and so t must
be prime, as required.
 
t(t + 1)
Consider .
t2 + 1  
2 2 t(t + 1) t(t + 1)
Since t(t + 1) = t + t ≥ t + 1, then 2 ≥ 1, which means that ≥ 1.
t +1 t2 + 1
       
t+1 t t+1 t
Consider 2 − 2 = − .
1 +1 1 +1 2 2
Since t is odd, then we write t = 2u + 1 for some positive integer u, which gives
         
t+1 t 2u + 2 2u + 1 1
− = − = bu + 1c − u + = (u + 1) − u = 1
2 2 2 2 2
Recall that t = rs with r ≥ s > 1.
   
r(t + 1) rt
Consider the term − 2 .
r2 + 1 r +1
We have
         2   2 
r(t + 1) rt r(rs + 1) r · rs r s+r r s
− = − = −
r2 + 1 r2 + 1 r2 + 1 r2 + 1 r2 + 1 r2 + 1
r2 s + r r2 s + s r2 s r2 s + s
We note that 2 ≥ 2 = s and 2 < 2 = s.
r +1 r +1 r +1 r +1
 2 
r s+r
Thus, ≥ s.
r2 + 1
 2   2     
r s r s r(t + 1) rt
Also, 2 < s which means 2 ≤ s − 1 and so − 2 ≥ 1.
r +1 r +1 r2 + 1 r +1
Therefore, if t is odd and not prime, then f (t + 1) − f (t) 6= 2 because we have found three
terms that are equal to at least 1 meaning that f (t + 1) − f (t) ≥ 3, and so if f (t + 1) − f (t),
then t must be prime.
Here is an alternative approach so show that f (t + 1) − f (t) ≥ 3 when t is odd and
composite.    
k(t + 1) kt
As above, we look for at least 3 integers k for which − 2 ≥ 1. Here,
k2 + 1 k +1
we allow for the possibility that k = t knowing that the second term in this difference will
be 0 in this case.    
k(t + 1) kt
The positive integer k has the property that − ≥ 1 exactly when
k2 + 1 k2 + 1
k(t + 1) kt
there is an integer N for which 2 ≥N > 2 .
k +1 k +1
2023 Euclid Contest Solutions Page 22

k2 + 1
This pair of inequalities is equivalent to the pair of inequalities t + 1 ≥ N · >t
k
N
which is in turn equivalent to t + 1 ≥ N k + > t.
k
The following three pairs (N, k) of integers satisfy this equation:
t+1 N
• k = 1 and N = (noting that t is odd), which give N k + = t + 1;
2 k
N s s
• k = r and N = s, which give N k + = rs + (noting that < 1);
k r r
N 1
• k = t and N = 1, which give N k + =t+ .
k t
This shows that f (t + 1) − f (t) ≥ 3 when t is odd and composite, as required.