Hypatia Combined Solutions

The CENTRE for EDUCATION
in MATHEMATICS and COMPUTING

cemc.uwaterloo.ca
2023 Hypatia Contest
Wednesday, April 5, 2023

(in North America and South America)
Thursday, April 6, 2023

(outside of North America and South America)
Solutions
©2023 University of Waterloo

2023 Hypatia Contest Solutions Page 2
1. (a) Jasmin’s total score was 3 · 2 + 4 · 5 = 6 + 20 or 26 points.

(b) Suppose that Sam had n throws that each scored 5 points; then Sam had 2n throws that
each scored 2 points.
Since Sam’s total score was 36 points, then 5 · n + 2 · 2n = 36 or 9n = 36, and so n = 4.
In total, Sam took n + 2n = 3n throws, which is 3 · 4 = 12 throws.
(c) Since Tia’s total score was 37 points, then 2t + 5f = 37.
Since 2t is even for all integer values of t, then 5f must be odd since their sum is 37 (which
is odd).
The value of 5f is odd exactly when f is odd.
When f = 1, we get 2t + 5 = 37 or 2t = 32, and so t = 16.
When f ≥ 9, 5f ≥ 45 and so 2t + 5f > 37.
The possible ordered pairs (t, f ) are (16, 1), (11, 3), (6, 5), and (1, 7).
(d) If a throws each score 6 points and b throws each score 21 points, then 6a+21b or 3(2a+7b)
points are scored.
Since a and b are non-negative integers, then 2a + 7b is a non-negative integer, and so the
total number of points scored, 3(2a + 7b), is a multiple of 3.
Since 182 is not a multiple of 3, then it is not possible to have a total score of 182 points.
2. (a) The total area of the shaded regions can be determined by subtracting the area of 4AED
from the area of rectangle ABCD.
The area of rectangle ABCD is 2 · 15 = 30.
The base length of 4AED is AD = 15, its height is equal to AB = 2, and so its area is
1
2
· 15 · 2 = 15.
Thus, the total area of the shaded regions is 30 − 15 = 15.
(b) Solution 1
B F C
Since 4AF D has base length AD = 15 and height equal to
2 G
AB = 2, then its area is 21 · 15 · 2 = 15.
A 15 D
The area of 4AF D is equal to the sum of the areas of
4AGD and 4F GD.
Since the area of 4AF D is 15, and the area of 4F GD is 5, then the area of 4AGD is
15 − 5 = 10.
The area of 4ABD is 12 · 15 · 2 = 15.
The area of 4ABD is equal to the sum of the areas of 4AGD and 4ABG.
Since the area of 4ABD is 15, and the area of 4AGD is 10, then the area of 4ABG (the
shaded area) is 15 − 10 = 5.
Solution 2
Consider 4AF D and 4ABD. Each has base AD and height AB and thus the two trian-
gles have equal areas.
The area of 4AF D is equal to the sum of the areas of 4AGD and 4F GD.
The area of 4ABD is equal to the sum of the areas of 4AGD and 4ABG.
Thus, the area of 4ABG (the shaded area) is equal to the area of 4F GD, which is 5.
(c) The total area of the bottom two shaded regions (4ASR
B P C
and 4RQD) can be determined by subtracting the area of
2 S Q
P QRS from the area of 4AP D.
The base length of 4AP D is AD = 15, its height is equal to A R D
AB = 2, and so its area is 21 · 15 · 2 = 15.
Since the area of P QRS is 6, the total area of the two
bottom shaded regions is 15 − 6 = 9.
We can similarly show that the total area of the two top shaded regions (4BSP and
4P QC) can be determined by subtracting the area of P QRS from the area of 4BRC.
Since 4BRC also has area 15, then the total area of the two top shaded regions is also
15 − 6 = 9.
Thus, the total area of the shaded regions is 9 + 9 = 18.
3. (a) Switching the second and fourth digits of 6238 gives the cousin 6832.
Since 6832 is not in the list, then it is the missing cousin.
(b) Of the 16 six-digit integers in the given list, 15 are cousins of the original integer.
If the original six-digit integer is abcdef , then there are exactly 5 cousins for which a is
not the 1st digit.
These 5 cousins are a result of switching the 1st digit, a, with each of the other 5 digits,
b, c, d, e, f , and thus are bacdef , cbadef , dbcaef , ebcdaf , and f bcdea.
Each of the remaining 15 − 5 = 10 cousins of abcdef have first digit a.
Since the digits are distinct, then the 1st digit of the original integer is the integer that
occurs most frequently as the 1st digit in the given list, which is 7.
A similar argument can be made for each of the other digits.
That is, the nth digit of the original integer is the integer that occurs most frequently as
the nth digit in the given list.
Thus, the original integer has 2nd digit 2, 3rd digit 6, 4th digit 4, 5th digit 9, 6th digit 1,
and so the original integer is 726 491.
We may check that if the original integer is 726 491, then the cousins are indeed the
remaining 15 integers in the given list.
(c) The cousins of the three-digit integer cd3 are dc3, 3dc and c3d.
The difference between the two integers cd3 and dc3 could be negative, in which case it’s
not the right difference. If it is positive, then cd3 minus dc3 has ones digit 0, and thus
cannot equal d95.
Similarly, the difference between the two integers cd3 and c3d could be negative, in which
case it’s not the right difference. Since cd3 and c3d each have hundreds digit c, then cd3
minus c3d has hundreds digit 0, and thus cannot equal d95 (since d is a non-zero digit).
Therefore, cd3 minus 3dc is equal to d95.
Since the difference, d95, has ones digit 5, then c = 8. (You should confirm for yourself
that this is the only possible value for c before moving on.)
With c = 8, we get 8d3 minus 3d8 is equal to d95.
The three-digit integer 8d3 is equal to 800 + 10d + 3.
The three-digit integer 3d8 is equal to 300 + 10d + 8.
Thus, 8d3 minus 3d8 is equal to (800 + 10d + 3) − (300 + 10d + 8) = 500 − 5 = 495, which
is equal to d95, and so d = 4.
(We may confirm that 843 − 348 = 495.)
The value of c is 8 and the value of d is 4, and no other values are possible.
(d) Solution 1
The six cousins of the four-digit integer mn97 are nm97, 9nm7, 7n9m, m9n7, m79n, and
mn79.
The cousin nm97 is equal to 1000n + 100m + 90 + 7.
The cousin 9nm7 is equal to 9000 + 100n + 10m + 7.
The cousin 7n9m is equal to 7000 + 100n + 90 + m.
The cousin m9n7 is equal to 1000m + 900 + 10n + 7.
The cousin m79n is equal to 1000m + 700 + 90 + n.
The cousin mn79 is equal to 1000m + 100n + 70 + 9.
The sum, S, of the six cousins is thus
S = 1000(n + 9 + 7 + m + m + m) + 100(m + n + n + 9 + 7 + n) + 10(9 + m + 9 + n + 9 + 7)
+ (7 + 7 + m + 7 + n + 9)
= 1000(3m + n + 16) + 100(m + 3n + 16) + 10(m + n + 34) + (m + n + 30)
The sum, S, must be equal to the five-digit integer nmnm7, which has ones digit 7.
The ones digit of S = 1000(3m + n + 16) + 100(m + 3n + 16) + 10(m + n + 34) + (m + n + 30)
is equal to the ones digit of m + n + 30, which is equal to the ones digit of m + n (since
the ones digit of 30 is 0).
If the ones digit of m + n is 7, then m + n = 7 or m + n = 17.
Since m and n are distinct non-zero digits, then the possible pairs (m, n) are (1, 6), (2, 5),
(3, 4), (4, 3), (5, 2), (6, 1), (8, 9), and (9, 8).
If (m, n) = (1, 6), then the five-digit integer nmnm7 is 61 617, and
S = 1000(3m + n + 16) + 100(m + 3n + 16) + 10(m + n + 34) + (m + n + 30)
= 1000(3(1) + 6 + 16) + 100(1 + 3(6) + 16) + 10(1 + 6 + 34) + (1 + 6 + 30)
= 1000(25) + 100(35) + 10(41) + 37
= 28 947
and so (m, n) = (1, 6) is not a possible pair.
We continue to check the remaining possible pairs (m, n) in the table below.
(m, n) nmnm7 S
(2, 5) 52 527 30 747
(3, 4) 43 437 32 547
(4, 3) 34 347 34 347
(5, 2) 25 257 36 147
(6, 1) 16 167 37 947
(8, 9) 98 987 54 657
(9, 8) 89 897 56 457
The only pair of distinct, non-zero digits (m, n) is (4, 3).
Solution 2
The six cousins of the four-digit integer mn97 are nm97, 9nm7, 7n9m, m9n7, m79n, and
mn79. As in Solution 1, the sum, S, of the six cousins is
S = 1000(n + 9 + 7 + m + m + m) + 100(m + n + n + 9 + 7 + n) + 10(9 + m + 9 + n + 9 + 7)
+ (7 + 7 + m + 7 + n + 9)
= 1000(3m + n + 16) + 100(m + 3n + 16) + 10(m + n + 34) + (m + n + 30)
= 3000m + 1000n + 16 000 + 100m + 300n + 1600 + 10m + 10n + 340 + m + n + 30
= 3111m + 1311n + 17 970
The sum, S, must be equal to the five-digit integer nmnm7, which is equal to
10 000n + 1000m + 100n + 10m + 7 or 10 100n + 1010m + 7.
Setting these two expressions equal to one another and simplifying, we get
3111m + 1311n + 17 970 = 10 100n + 1010m + 7

8789n − 2101m = 17 963
799n − 191m = 1633 (after division by 11)
Since 799n = 1633 + 191m and m ≥ 1, then 799n ≥ 1633 + 191(1) or 799n ≥ 1824, and
1824
so n ≥ ≈ 2.3.
799
3352
Since m ≤ 9, then 799n ≤ 1633 + 191(9) or 799n ≤ 3352, and so n ≤ ≈ 4.2.
799
Since 2.3 ≤ n ≤ 4.2 and n is an integer, then the possible values of n are 3 and 4.
1563
Substituting n = 4, we get 799(4) = 1633 + 191m or 191m = 1563, and since is not
191
an integer, then n 6= 4.
Substituting n = 3, we get 799(3) = 1633 + 191m or 191m = 764, and so m = 4.
When m = 4 and n = 3, the five-digit integer nmnm7 is 34 347 and
S = 3111(4) + 1311(3) + 17 970 or S = 34 347, and so the only pair of distinct, non-
zero digits (m, n) is (4, 3).
4. (a) Since there are 9 ways to generate each of the three integers, there are a total of
9 · 9 · 9 = 729 ways to generate all three integers.
Next, we consider the different cases for which the product of the three integers is a prime
number.
The integers from 1 to 9 include the composite numbers 4, 6, 8, and 9.
If any one of the three integers generated is composite, then the product of the three
integers is composite.
Thus, the integers must be chosen from 1, 2, 3, 5, and 7.
Of these five integers, 2, 3, 5, and 7 are prime numbers.
If two or more of the three integers generated are prime numbers, then the product of the
three integers is composite.
Thus, at most one integer is a prime number.
If no integer generated is a prime number (all three are equal to 1), then the product is 1,
which is not prime.
Thus, Amarpreet’s product is a prime number exactly when one of the integers is 2, 3, 5,
or 7, and each of the other two integers is equal to 1.
There are 4 ways to choose one of the prime numbers, p, and 3 ways to arrange the integers
1, 1, p, and so there are 4 · 3 = 12 ways to generate three integers whose product is a
prime number.
12 4
Therefore, the probability that the product is a prime number is = .
729 243
(b) Solution 1
We begin by considering the different cases for which the product of the four integers is
divisible by 5, but not divisible by 7.
The only integer from 1 to 9 that is divisible by 5 is 5, and so the product of the four
integers is divisible by 5 exactly when at least one of the four integers is a 5.
Similarly, the only integer from 1 to 9 that is divisible by 7 is 7, and so the product of
the four integers is not divisible by 7 exactly when a 7 is not among the four integers
generated.
We proceed to count the number of ways to generate such arrangements of four integers
by considering the number of times that 5 appears in the arrangement.
Case 1: there are four 5s
Since each of the four integers must be 5, there is 1 such possibility in this case.
Case 2: there are exactly three 5s
The three 5s can be arranged in 4 different ways: 555 , 55 5, 5 55, 555.
There are 7 choices for the integer that is not a 5 since it can be any of the nine integers,
except 5 and 7.
Thus, there are 4 · 7 = 28 possibilities in this case.
Case 3: there are exactly two 5s
The two 5s can be arranged in 6 different ways: 55 , 5 5 , 5 5, 55 , 5 5, 55.
Once the two 5s have been placed, there are 7 choices for each of the remaining two inte-
gers (since it is possible that these integers are equal to one another).
Thus, there are 6 · 7 · 7 = 294 possibilities in this case.
Case 4: there is exactly one 5
The 5 can be placed in 4 different ways.
Once the 5 has been placed, there are 7 choices for each of the remaining three integers
(since it is possible that these integers are equal to one another).
Thus, there are 4 · 7 · 7 · 7 = 1372 possibilities in this case.
In total, there are 1 + 28 + 294 + 1372 = 1695 ways to select the four integers.
Since there are 9 ways to generate each of the four integers, there are a total of
94 = 6561 ways to generate all four integers.
Therefore, the probability that Braxton’s product is divisible by 5, but not divisible by 7
1695 565
is = .
6561 2187
Solution 2
Let A be the event that the product generated by Braxton is divisible by 5, and A be the
event that the product is not divisible by 5.
Let B be the event that the product generated by Braxton is divisible by 7, and B be the
event that the product is not divisible by 7. P(A and B)
We are asked to find the probability of A and B, which we
P(A) P(B)
write as P (A and B).
Consider the Venn diagram shown in Figure 1.
The shaded region of this diagram represents P (A and B).
Figure 1
We wish to determine a more efficient method for determining P (A and B) than that which
was shown in Solution 1.
To do so, we proceed to use Venn diagrams to help express P (A and B) in an equivalent
form.
Consider the next two Venn diagrams shown below.

In Figure 2, the shaded region represents P (B).
In Figure 3, the shaded region represents P (A and B).
P(B) P(A and B)

P(A) P(B) P(A) P(B)
Figure 2 Figure 3
Notice that if the region that is shaded in Figure 3 is removed (unshaded) in Figure 2,
then the resulting Venn diagram is equivalent to that in Figure 1.
That is, mathematically, P (A and B) = P (B) − P (A and B).
(Without the use of the Venn diagrams, we may have similarly noticed that since exactly
one of A and A occurs, then P (B) = P (A and B) + P (A and B), and so
P (B) − P (A and B) = P (A and B) + P (A and B) − P (A and B) = P (A and B). )
Thus, the probability that the product is divisible by 5, but not divisible by 7, is equal
to the probability that the product is not divisible by 7, minus the probability that the
product is both not divisible by 5 and not divisible by 7.
integers is not divisible by 7 exactly when a 7 is not among the four integers generated.
8
The probability that the integer generated is not 7 is , and so the probability that the
9 4
8
product of the four integers is not divisible by 7 is P (B) = .
9
integers is not divisible by 5 exactly when a 5 is not among the four integers generated.
7
The probability that the integer generated is not divisible by both 5 and 7 is thus , and
9
so the probability that the product of the four integers is not divisible 5 and not divisible
4
7
by 7 is P (A and B) = .
9
Therefore, the probability that the product is divisible by 5, but not divisible by 7, is
4 4
8 7 84 − 74 4096 − 2401 1695 565
− = 4
= = = .
9 9 9 6561 6561 2187
Solution 3
Suppose that the four numbers generated are a, b, c, d.
When there are no restrictions, there are 9 choices for each of a, b, c, d, and so a total of
94 possible combinations of 4 digits that can be generated.
If abcd is not divisible by 7, then none of a, b, c, d can be 7; thus, there are 8 possibilities
for each of a, b, c, d, and so 84 of these 94 combinations give products that are not divisible
by 7.
If abcd is not divisible by 7 and not divisible by 5, then none of a, b, c, d can be 5; thus,
there are 7 possibilities for each of a, b, c, d, and so 74 of these 84 combinations whose
products are not divisible by 7 give products that are also not divisible by 5.
This means that there are 84 − 74 combinations whose product is not divisible by 7 but
whose product is divisible by 5.

Therefore, the probability that a random combination gives a product that is not divisible
84 − 74 4096 − 2401 1695 565
by 7 but is divisible by 5 is 4
= = = .
9 6561 6561 2187
(c) If any one of the 2023 integers generated is a 6, then the product is divisible by 6.
Thus, we exclude 6 from the choice of possible integers.
Further, an integer is divisible by 6 exactly when it is divisible by both 2 and 3.
With 6 excluded, the remaining integers that are divisible by 2 are 2, 4 and 8, and the
remaining integers that are divisible by 3 are 3 and 9.
Therefore, the product is also divisible by 6 when at least one of the integers generated is
2 or 4 or 8, and at least one of the integers generated is 3 or 9.
To summarize, the product is not divisible by 6 exactly when
• 6 is not among the integers generated, and
• there are integers from at most one of the two lists 2, 4, 8, and 3, 9.
Thus, there are exactly 3 cases which produce a product that is not divisible by 6, as
follows.
Case Is there a 6? Is there a 2, 4 or 8? Is there a 3 or 9?
1 no no no
2 no no yes
3 no yes no
For each of these 3 cases, we count the number of ways that the product generated is not
divisible by 6.
Case 1: there is no 6, 2, 4, 8, 3, or 9
In this case, there are 3 integers that may be chosen (1, 5 and 7), and so there are 32023
ways to generate this product.
Case 2: there is no 6, 2, 4, or 8
In this case, there are 5 integers that may be chosen (1, 3, 5, 7, and 9), and so there are
52023 ways to generate this product.
However, this count includes the cases in which only 1, 5 and 7 are chosen (3 and 9 are
not), which were counted in Case 1.
Thus, the number of ways to generate the products in Case 2, different from those in Case
1, is 52023 − 32023 .
Case 3: there is no 6, 3, or 9
In this case, there are 6 integers that may be chosen (1, 2, 4, 5, 7, and 8), and there are
62023 ways to generate this product.
However, this count includes the cases in which only 1, 5 and 7 are chosen (2, 4 and 8 are
not), which were counted in Case 1.
Thus, the number of ways to generate the products in Case 3, different from those in Case
1, is 62023 − 32023 .
Thus, the total number of ways to generate all products that are not divisible by 6 is
32023 + 52023 − 32023 + 62023 − 32023 = 52023 + 62023 − 32023
When there are no restrictions, there are 9 choices for each integer generated, and so there
is a total of 92023 possible ways to generate the product.
Thus, the probability, p, that the product is not divisible by 6 is
52023 + 62023 − 32023

p=
92023
and so p · 92023 = 52023 + 62023 − 32023 , which is equal to the sum and difference of integers
and thus an integer.
To determine the ones digit of the integer equal to p · 92023 , we determine the ones digit of
the integer equal to 52023 + 62023 − 32023 .
The ones digit of 5a is equal to 5 for all positive integers a, and thus the ones digit of 52023
is 5.
The ones digit of 6b is equal to 6 for all positive integers b, and thus the ones digit of 62023
is 6.
Consider the ones digit of successive integer powers of 3 beginning at 31 :
• 31 has ones digit 3
Since 31 and 35 have the same ones digit and we are performing the same action to get
from one step to the next, the results begin to repeat. Thus the block of 4 ones digits
(3, 9, 7, 1) must form a cycle.
Since 2023 = 4 · 505 + 3, then the ones digit of 32023 is the third number in the repeating
block, which is 7.
Finally, the ones digit of the integer equal to p · 92023 is the ones digit of 5 + 6 − 7, which
is equal to 4.
cemc.uwaterloo.ca
Tuesday, April 12, 2022


Solutions

1. (a) Regular hexagon ABCDEF has side length 2x, and so AB = 2x.
Since 4OAB is equilateral, then OA = OB = AB = 2x.
The radius of the circle is equal to OA and thus is 2x.
(b) Since M is the midpoint of AB and OA = OB, then A F
OM is perpendicular to AB.
M
Since M is the midpoint of AB, then AM = 21 AB = x.
O
Using the Pythagorean Theorem in right-angled B E
4OAM , we get OA2 = OM 2 + AM 2 or
2
(2x) √ = OM + x , and so OM 2 = 3x2 or
2 2
OM = 3x (since OM > 0). C D

Alternatively, notice that 4OAM is a√30◦ -60◦ -90◦ trian-
√ Figure 2
gle, and so AM : OA : OM = 1 : 2 : 3 = x : 2x : 3x.
(c) The diagonals AD, BE and CF divide ABCDEF into six congruent equilateral triangles.
Thus the area of ABCDEF is six times the area of 4OAB (having base AB and height
OM ), or √ √
6 × 12 × AB × OM = 3 × 2x × 3x = 6 3x2
(d) The area of the shaded region is determined by subtracting the area of ABCDEF from
the area of the circle with centre O and radius 2x.
Thus, the area of the shaded region is
√ √ √
π(2x)2 − 6 3x2 = 4πx2 − 6 3x2 = (4π − 6 3)x2
√ 123
The area of this shaded region is 123, and so (4π − 6 3)x2 = 123 or x2 = √ .
r 4π − 6 3
123
Since x > 0, we get x = √ and so x = 7.5 when rounded to the nearest tenth.
4π − 6 3
2. (a) With 1 kg of muffin batter, exactly 24 mini muffins and 2 large muffins can be made. Thus
with 2 kg of muffin batter, exactly 2 × 24 = 48 mini muffins and 2 × 2 = 4 large muffins
can be made, and so n = 4.
(b) Solution 1
With 2 kg of muffin batter, exactly 36 mini muffins and 6 large muffins can be made.
With 2 kg of muffin batter, exactly 48 mini muffins and 4 large muffins can be made.
Adding these, we get that with 2 kg + 2 kg = 4 kg of muffin batter, exactly 36 + 48 = 84
mini muffins and 6 + 4 = 10 large muffins can be made, and so x = 4.
Solution 2
Let m be the number of kilograms of muffin batter needed to make 1 mini muffin.
Let ` be the number of kilograms of muffin batter needed to make 1 large muffin.
Since 1 kg of muffin batter makes exactly 24 mini muffins and 2 large muffins, then
1 = 24m + 2`.
Since 2 kg of muffin batter makes exactly 36 mini muffins and 6 large muffins, then
2 = 36m + 6`.
Subtracting the second equation from 3 times the first equation, we get
1
3 × 1 − 2 = 3 × (24m + 2`) − (36m + 6`) or 1 = 36m, and so m = .
36
1 1
Substituting m = into the second equation, we get 2 = 36 + 6` or 1 = 6`, and
36 36
1
so ` = .
6
1 1 7
Since kg of muffin batter is needed to make 1 mini muffin, then 84 · = kg of muf-
36 36 3
fin batter is needed to make 84 mini muffins.
1 1 5
Since kg of muffin batter is needed to make 1 large muffin, then 10 · = kg of muffin
6 6 3
batter is needed to make 10 large muffins.
7 5 12
Thus, exactly 84 mini muffins and 10 large muffins can be made with + = = 4 kg
3 3 3
of muffin batter, and so x = 4.
1
(c) In part (b) Solution 2, it was determined that 36 kg of muffin batter is needed to make
1
1 mini muffin, and 6 kg of muffin batter is needed to make 1 large muffin.
Therefore, the number of kilograms of muffin batter needed to make 1 large muffin is
1 1
6 times the number of kilograms of batter needed to make 1 mini muffin 6 × 36 = 6 .
In other words, the batter needed to make 1 large muffin can make 6 mini muffins.
Then, using the amount of batter that makes 7 large muffins, 6 × 7 = 42 mini muffins can
be made.
3. (a) If the first number in a sequence is 3 and the sequence is generated by the function
x2 − 3x + 1, then the second number in the sequence is
32 − 3(3) + 1 = 1,
and the third number in the sequence is
12 − 3(1) + 1 = −1,
and the fourth number in the sequence is
(−1)2 − 3(−1) + 1 = 5.
The first four numbers in the sequence are 3, 1, −1, 5.
(b) Let the first and second numbers in the sequence generated by the function x2 − 4x + 7
be f and s, respectively.
Then, the first three numbers in the sequence are f, s, 7.
Since the third number in the sequence is 7, then s2 − 4s + 7 = 7.
Solving this equation, we get s2 − 4s = 0 or s(s − 4) = 0, which has solutions s = 0 and
s = 4, and so the first three numbers in the sequence could be f, 0, 7 or f, 4, 7.
If the second number in the sequence is 0, then f 2 − 4f + 7 = 0.
The discriminant of this equation is (−4)2 − 4(1)(7) = −12 (less than zero) and so there
are no real solutions.
Thus, there is no first number in this sequence for which the second number is 0.
If the second number in the sequence is 4, then f 2 − 4f + 7 = 4, or f 2 − 4f + 3 = 0 and
so (f − 1)(f − 3) = 0, which has solutions f = 1 and f = 3.
Therefore, if 7 is the third number in a sequence generated by the function x2 − 4x + 7,
then the first three numbers in the sequence could be 1, 4, 7 or 3, 4, 7, and so the possible
first numbers in the sequence are 1 and 3.
(c) The first two numbers in the sequence are c, c, and so c2 − 7c − 48 = c.
Solving this equation, we get c2 − 8c − 48 = 0 or (c + 4)(c − 12) = 0, which has solutions
c = −4 and c = 12.
(d) The first number in the sequence is a and the second number is b, and so
a2 − 12a + 39 = b.
The second number in the sequence is b and the third number is a, and so
b2 − 12b + 39 = a.
Subtracting the second equation from the first and simplifying, we get
(a2 − 12a + 39) − (b2 − 12b + 39) = b−a

a2 − b2 − 12a + 12b = b−a
a2 − b2 − 11a + 11b = 0
(a − b)(a + b) − 11(a − b) = 0
(a − b)(a + b − 11) = 0
Since a 6= b, then a − b 6= 0 and so a + b − 11 = 0 or b = 11 − a.

Substituting into the first equation, we get a2 − 12a + 39 = 11 − a or a2 − 11a + 28 = 0.
Factoring gives (a − 4)(a − 7) = 0 and so the possible values of a are 4 and 7.
(Note that the two possible sequences are 4, 7, 4, . . . and 7, 4, 7, . . . .)
4. (a) Written as a product of its prime factors, 240 = 24 31 51 and so
f (240) = (1 + 4)(1 + 1)(1 + 1) = (5)(2)(2) = 20
(b) Suppose f (N ) = 6 and N is refactorable. Then 6 divides N .

N is divisible by 6 exactly when its prime factors include at least one 2 and at least one 3
(since 6 = 2 × 3).
Assume that N contains an additional prime factor p, distinct from 2 and 3.
In this case, the divisors of N are 1, 2, 3, 6, p, 2p, 3p, and 6p, which is too many divisors.
Thus, it must be that N is a positive integer of the form N = 2u 3v , where u and v are
positive integers, and so f (N ) = (1 + u)(1 + v) = 6.
Since u and v are positive integers, then 1 + u ≥ 2 and 1 + v ≥ 2 and so there are exactly
two possibilities: u = 1 and v = 2 or u = 2 and v = 1.
When u = 1 and v = 2, N = 21 32 = 18 and when u = 2 and v = 1, N = 22 31 = 12.
The refactorable numbers N with f (N ) = 6 are 12 and 18.
(c) Since N is refactorable, then f (N ) = 256 = 28 divides N .
Thus for some integer w ≥ 8, N is of the form N = 2w pa11 pa22 · · · pakk , where
2 < p1 < p2 < · · · < pk are prime numbers, and a1 , a2 , · · · , ak are positive integers.
In this case, f (N ) = (1 + w)(1 + a1 )(1 + a2 ) · · · (1 + ak ) = 28 , and so each of the factors
1 + w, 1 + a1 , 1 + a2 , · · · , 1 + ak is a power of 2 (since their product is 28 ).
This means that each of the exponents w, a1 , a2 , · · · , ak is one less than a power of two.
Since w ≥ 8, then 1 + w ≥ 9 and so 1 + w ≥ 24 (the smallest power of 2 greater than 8).
Since f (N ) = (1 + w)(1 + a1 )(1 + a2 ) · · · (1 + ak ) = 28 , then 24 ≤ 1 + w ≤ 28 and
1 ≤ (1 + a1 )(1 + a2 ) · · · (1 + ak ) ≤ 24 , and so w must be equal to 15, 31, 63, 127, or 255,
and each of a1 , a2 , · · · , ak must be equal to 1, 3, 7, or 15.
For example, if 1 + w = 28 , then w = 256 − 1 = 255 and N = 2255 .
If 1 + w = 24 , then w = 15 and N is of the form N = 215 pa11 pa22 · · · pakk .
Thus, the smallest N having the greatest number of distinct prime factors is N = 215 31 51 71 111 ,
and we may confirm that f (N ) = (1 + 15)(1 + 1)(1 + 1)(1 + 1)(1 + 1) = 24 · 2 · 2 · 2 · 2 = 28 ,
as required.
Comparing these first two values of N , we recognize that 215 31 51 71 111 is significantly less
than 2255 .
Further, we recognize that
• all remaining possible values of N have exactly 2, 3 or 4 distinct prime factors,
• where exponents are equal, smaller prime factors give smaller values of N ,
• the greatest exponents must occur on the smallest prime factors, and
• we recall that each exponent is one less than a power of 2.
In the table below, we use the above information to determine the smallest possible values
of N having 1, 2, 3, 4, and 5 distinct prime factors.
Further, we compare the size of each of these values of N within each of the 5 groups.
Number of distinct Values of N
prime factors of N
1 2255
2 215 315 < 231 37 < 263 33 < 2127 31
3 215 33 53 < 215 37 51 < 231 33 51 < 263 31 51
4 215 33 51 71 < 231 31 51 71
5 215 31 51 71 111
Finally, we compare the smallest values of N from each of the 5 rows in the table above.
Since 215 33 51 71 < 215 31 51 71 111 < 215 33 53 < 215 315 < 2255 , then the smallest refactorable
number N with f (N ) = 256, is N = 215 33 51 71 = 30 965 760.
(d) Let m be a positive integer of the form m = pa11 pa22 · · · pakk , where p1 < p2 < · · · < pk are
prime numbers, and a1 , a2 , · · · , ak are positive integers.
We begin by stating a value of N and then proceed to show that it works.
(pa1 −1) (pa2 2 −1) (pak −1)
For each m = pa11 pa22 · · · pakk , choose N = p1 1 p2 · · · pk k .
n
On the contest paper, the useful fact states that 2 ≥ n + 1 for all positive integers n.
Since p ≥ 2 for all prime numbers p, then pn ≥ 2n ≥ n + 1.
Since pn ≥ n + 1, then pn − 1 ≥ n, and so pai i − 1 ≥ ai for all integers i where 1 ≤ i ≤ k.
(pa1 −1)
Thus, pa11 divides p1 1 since each is a power with base p1 and pa11 − 1 ≥ a1 .
a2
(p −1)
Similarly, pa22 divides p2 2 , and so on.
(pa1 −1) (pa2 2 −1) (pak −1)
Therefore, m = pa11 pa22 · · · pakk divides N = p1 1 p2 · · · pk k .
Further, f (N ) = (1 + pa11 − 1) (1 + pa22 − 1) · · · (1 + pakk − 1) = pa11 pa22 · · · pakk = m.
Thus for every integer m > 1, there exists a refactorable number N such that f (N ) = m.
cemc.uwaterloo.ca
April 2021
April 2021
Solutions

1. (a) The total cost to rent a car is $180.00.

$180.00
If 4 people rent a car, the cost per person is = $45.00.
4
(b) Since the members of the group equally share the total cost to rent the vehicle, the smaller
the group, the greater the cost per person.
To rent an SUV, the smallest group size required is 5 passengers and the total cost is
$200.00.
$200.00
Thus, the maximum possible cost per person to rent an SUV is = $40.00.
5
(c) Let the total cost to rent a van be v.
When renting a van, the maximum possible cost per person occurs when the number of
v
passengers is 9 (the fewest number possible), and so this maximum cost is .
9
The minimum possible cost per person occurs when the number of passengers is 12 (the
v
greatest number possible), and so this minimum cost is .
12
v v 4v − 3v
Then, − = $6.00 or = $6.00, and so v = $6.00 × 36.
9 12 36
Thus, the total cost to rent a van is $216.00.
2. (a) Trapezoid ABCD is drawn, as shown. y

The slope of line segments AB and CD are each zero
and thus they are parallel. D (2, 5) C (11, 5)
The length of AB is the difference between the
x-coordinates of A and B, or 12.
The length of CD is the difference between the x
x-coordinates of C and D, or 11 − 2 = 9. A (0, 0) B (12, 0)
The height of the trapezoid is equal to the vertical distance between AB and CD, which
is 5.
5 5 105
The area of trapezoid ABCD is (AB + CD) or (21) = .
2 2 2
(b) The line passing through B and D intersects the y-axis y
at E. Let the coordinates of E be (0, e), as shown.
5−0 1 E (0, e) D (2, 5) C (11, 5)
The slope of the line through B and D is =− .
2 − 12 2
Solution 1 B (12, 0)
Since E, D and B lie on the same line, then the slope of x
ED is equal to the slope of BD. A (0, 0)
e−5 1 e−5 1
Equating slopes, we get = − or = , and so e − 5 = 1 or e = 6.
0−2 2 2 2
Thus, point E has coordinates (0, 6).
Solution 2
1
The line passing through B(12, 0) and D(2, 5) has slope − , and thus has equation
2
1
y − 5 = − (x − 2).
2
1 1
Rearranging, we get y − 5 = − x + 1 or y = − x + 6.
2 2
Since this line has y-intercept 6, then point E has coordinates (0, 6).
Solution 3
1
The line passing through B(12, 0) and D(2, 5) has slope − , and thus has
2
1
equation y − 5 = − (x − 2).
2
1
This line passes through E(0, e), and so e − 5 = − (0 − 2) or e = 1 + 5 = 6.
2
Thus, point E has coordinates (0, 6).
(c) Sides AD and BC are extended to intersect at F , as y F ( j, k)
shown.
Solution 1
Let the coordinates of F be (j, k).
Since A, D and F lie on the same line, then the slope of D (2, 5) C (11, 5)
AD is equal to the slope of AF .
5 k 5
Equating slopes, we get = or k = j.
2 j 2 x
Since B, C and F lie on the same line, then the slope of A (0, 0) B (12, 0)
BC is equal to the slope of BF .
5−0 k−0 k
Equating slopes, we get = or −5 = , and so k = −5(j − 12).
11 − 12 j − 12 j − 12
5 5
Substituting k = j, we get j = −5(j − 12) or j = −2(j − 12), and so 3j = 24 or j = 8.
2 2
5
When j = 8, k = (8) = 20, and so F has coordinates (8, 20).
2
Solution 2
5
The line passing through A(0, 0) and D(2, 5) has slope and y-intercept 0, and thus has
2
5
equation y = x.
2
The line passing through B(12, 0) and C(11, 5) has slope −5 and thus has equation
y = −5(x − 12).
These two lines intersect at F , and so the coordinates of F can be determined by solving
5
the equation x = −5(x − 12). Solving, we get x = −2(x − 12) or 3x = 24, and so x = 8.
2
5
When x = 8, y = (8) = 20, and so F has coordinates (8, 20).
2
(d) Let P have coordinates (r, s).
Assume AB is the base of 4P AB.
1
In this case, if the height of 4P AB is h, then the area of 4P AB is (AB)h = 6h.
2
The area of 4P AB is 42, and so 6h = 42 or h = 7.
That is, P (r, s) is located a vertical distance of 7 units from the line through A and B, or
7 units from the x-axis.
There are two possibilities: P (r, s) is located 7 units above the x-axis, and thus lies on
the horizontal line y = 7, or P (r, s) is located 7 units below the x-axis, and thus lies on
the horizontal line y = −7.
In the first case, P has coordinates (r, 7) and in the second case, P has coordinates (r, −7).
Recall that P lies on the line passing through B and D.
1
The line passing through B(12, 0) and D(2, 5) has slope − , and thus has
2
1
equation y − 5 = − (x − 2).
2
1
If P (r, 7) lies on this line, then 7 − 5 = − (r − 2) or −4 = r − 2, and so r = −2.
2
1
Similarly, if P (r, −7) lies on this line, then −7 − 5 = − (r − 2) or 24 = r − 2, and so in
2
this case, r = 26.
The points P that lie on the line passing through B and D, so that the area of 4P AB
is 42, are (−2, 7) and (26, −7).
3. (a) Since an = 2n for n ≥ 1, then a5 = 25 = 32.

Since b2 = 1, b3 = 3, and bn = bn−1 + 2bn−2 for n ≥ 3, then b4 = b3 + 2b2 = 3 + 2(1) = 5
and b5 = b4 + 2b3 = 5 + 2(3) = 11.
Therefore, a5 = 32 and b5 = 11.
(b) Since b1 = p · (a1 ) + q · (−1)1 and a1 = 2, then b1 = 2p − q.
From the definition of sequence B, we know b1 = 1, and so 2p − q = 1.
Since b2 = p · (a2 ) + q · (−1)2 and a2 = 22 = 4, then b2 = 4p + q.
From the definition of sequence B, we know b2 = 1, and so 4p + q = 1.
This gives two equations in two unknowns, p and q.
Adding these two equations, we get 6p = 2, and so p = 13 .
Substituting, we get 2 13 − q = 1 or q = 23 − 1 = − 31 .

1
Thus, the real numbers p and q for which bn = p · (an ) + q · (−1)n for all n ≥ 1 are p = 3
and q = − 13 .
(c) Using algebraic manipulation, and the definitions an = 2n and bn = 13 (an ) − 13 (−1)n , each
for n ≥ 1, we obtain the following equivalent equations,
Sn = b1 + b2 + b3 + · · · + bn
= 13 (a1 ) − 13 (−1) + 31 (a2 ) − 13 (−1)2 + 31 (a3 ) − 13 (−1)3 + · · · + 31 (an ) − 13 (−1)n

= 13 (a1 + a2 + a3 + · · · + an ) − 31 ((−1) + (−1)2 + (−1)3 + · · · + (−1)n )

= 13 (2 + 22 + 23 + · · · + 2n ) − 13 (−1 + 1 − 1 + · · · + (−1)n )
Next, we consider each of the two expressions within parentheses, separately.

The expression 2 + 22 + 23 + · · · + 2n is the sum of n terms of a geometric sequence with
first term a = 2 and common ratio r = 2.
1 − 2n

2 3 n
Thus, 2 + 2 + 2 + · · · + 2 = 2 = −2(1 − 2n ).
1−2
The expression −1 + 1 − 1 + · · · + (−1)n is an alternating sum of the terms −1 and 1.
This simplifies to 0 if there are an even number of terms, that is, if n is even, and simplifies
to −1 if n is odd.
Summarizing, we have

 31 (−2(1 − 2n )) , if n is even
Sn =
 1 (−2(1 − 2n )) + 1 , if n is odd
3 3
and simplifying, we get 

2
 3 (2n − 1) , if n is even

Sn =
 32 (2n − 1) +
 1
3
, if n is odd
We want the smallest positive integer n that satisfies Sn ≥ 162021 and note that the value
of Sn increases as n increases.
2021
Since 16 = 24 , then 162021 = (24 ) = 28084 and so we want the smallest positive integer n
that satisfies Sn ≥ 28084 .
When n is even, we get When n is odd, we get
2 n
3
(2 −1) ≥ 28084 2 n
3
(2 − 1) + 13 ≥ 28084
2 − 1 ≥ 3 · 28083
n 2 n
3
(2 − 1) ≥ 28084 − 13
2n ≥ 3 · 28083 + 1 2n − 1 ≥ 3 · 28083 − 1
2
2n ≥ 3 · 28083 + 1
2
Since 2n is an even integer, 3 · 28083 + 1 is an odd integer and 3 · 28083 + 12 is between an even
integer and an odd integer, and thus the inequalities 2n ≥ 3 · 28083 + 1 and 2n ≥ 3 · 28083 + 12
are both equivalent to saying 2n > 3 · 28083 .
Since 3 · 28083 > 2 · 28083 , then simplifying, we get 3 · 28083 > 28084 .
Thus, we want the smallest positive integer n that satisfies 2n > 3 · 28083 > 28084 .
When n ≤ 8084, this inequality is not true.
When n = 8085, we get 28085 = 22 · 28083 = 4 · 28083 which is greater than 3 · 28083 , as
required.
Thus, the smallest positive integer n that satisfies Sn ≥ 162021 is n = 8085.
4. (a) In 4XY Z, x = 20, y = 21, and ∠XZY = 90◦ , Y

as shown.
√ Using the Pythagorean Theorem, we get
20 cm
z = 202 + 212 = 29 (since z > 0).
The value of A is Z X
21 cm
1 1
A = (y)(x) = (21)(20) = 210
2 2
The value of P is
P = z + x + y = 29 + 20 + 21 = 70
(b) When A = 336, we get 12 xy = 336, and so xy = 672.
By the Pythagorean Theorem, x2 + y 2 = 502 , which when manipulated algebraically gives
the following equivalent equations
x2 + y 2 = 2500
(x + y)2 − 2xy = 2500
(x + y)2 = 2500 + 2xy
(x + y)2 = 2500 + 2(672)
(x + y)2 = 3844
√
Since x + y > 0, then x + y = 3844 = 62.
Thus, we get P = x + y + z = 62 + 50 = 112.
(The triangle satisfying these conditions has side lengths 14 cm, 48 cm and 50 cm.)
(c) Since A = 3P , we get 12 xy = 3(x + y + z), and so xy = 6(x + y + z).

When xy = 6(x + y + z) is manipulated algebraically, we get the following equivalent
equations
xy = 6(x + y + z)
xy − 6x − 6y = 6z
(xy − 6x − 6y)2 = (6z)2
(xy)2 − 12x(xy) − 12y(xy) + 72xy + 36x2 + 36y 2 = 36z 2
(xy)2 − 12x(xy) − 12y(xy) + 72xy + 36x2 + 36y 2 = 36(x2 + y 2 ) (since x2 + y 2 = z 2 )
xy(xy − 12x − 12y + 72) = 0
xy − 12x − 12y + 72 = 0 (since xy 6= 0)
x(y − 12) − 12y = −72
x(y − 12) − 12y + 144 = −72 + 144
x(y − 12) − 12(y − 12) = 72
(x − 12)(y − 12) = 72
Since x and y are positive integers, then x − 12 and y − 12 are a factor pair of 72.
The product (x − 12)(y − 12) is positive (since 72 > 0), and thus x − 12 < 0 and y − 12 < 0
or x − 12 > 0 and y − 12 > 0.
If x − 12 < 0 and y − 12 < 0, then x < 12 and y < 12.
There are exactly two Pythagorean triples (x, y, z) in which x < 12 and y < 12.
In the first case, (x, y, z) = (3, 4, 5), which gives A = 21 (3)(4) = 6, P = 3 + 4 + 5 = 12, and
so A 6= 3P .
In the second case, (x, y, z) = (6, 8, 10), which gives A = 21 (6)(8) = 24, P = 6+8+10 = 24,
and so A 6= 3P .
Therefore, x − 12 > 0 and y − 12 > 0, and so x and y are each greater than 12.
In the table below, we use the positive factor pairs of 72 to determine all possible integer
values of x, y and z.
Further, we initially make the assumption that x ≤ y, recognizing that by the symmetry
of the equation, the values of x and y may be interchanged with one another and doing
so gives the same value for z and the same triangle.
p
Factor pair x − 12 y − 12 x y z = x2 + y 2 (x, y, z)
1 and 72 1 72 13 84 85 (13, 84, 85) or (84, 13, 85)
2 and 36 2 36 14 48 50 (14, 48, 50) or (48, 14, 50)
3 and 24 3 24 15 36 39 (15, 36, 39) or (36, 15, 39)
4 and 18 4 18 16 30 34 (16, 30, 34) or (30, 16, 34)
6 and 12 6 12 18 24 30 (18, 24, 30) or (24, 18, 30)
8 and 9 8 9 20 21 29 (20, 21, 29) or (21, 20, 29)
It is worth noting that instead of factoring the equation xy − 12x − 12y + 72 = 0 as we
did, we could have chosen to rewrite it as
x(y − 12) = 12y − 72

12y − 72
x =
y − 12
12(y − 12) + 144 − 72
x =
y − 12
72
x = 12 +
y − 12
and considered that since x is a positive integer, then y − 12 is a divisor of 72.
(d) Since A = kP , we get 12 xy = k(x + y + z), and so xy = 2k(x + y + z).
When xy = 2k(x + y + z) is manipulated algebraically, we get the following equivalent
equations
xy = 2k(x + y + z)
xy − 2kx − 2ky = 2kz
(xy − 2kx − 2ky)2 = (2kz)2
(xy)2 − 4kx(xy) − 4ky(xy) + 8k 2 xy + 4k 2 x2 + 4k 2 y 2 = 4k 2 z 2
(xy)2 − 4kx(xy) − 4ky(xy) + 8k 2 xy + 4k 2 x2 + 4k 2 y 2 = 4k 2 (x2 + y 2 ) (∵ x2 + y 2 = z 2 )
xy(xy − 4kx − 4ky + 8k 2 ) = 0
xy − 4kx − 4ky + 8k 2 = 0 (since xy 6= 0)
x(y − 4k) − 4ky = −8k 2
x(y − 4k) − 4ky + 16k 2 = −8k 2 + 16k 2
x(y − 4k) − 4k(y − 4k) = 8k 2
(x − 4k)(y − 4k) = 8k 2
Since x, y and k are positive integers, then x − 4k and y − 4k are a factor pair of 8k 2 .
We begin by assuming that k = 2.
Substituting, we get (x − 8)(y − 8) = 32.
The product (x − 8)(y − 8) is positive (since 32 > 0), and thus x − 8 < 0 and y − 8 < 0
or x − 8 > 0 and y − 8 > 0.
If x − 8 < 0 and y − 8 < 0, then x < 8 and y < 8 which is not possible since P = 510, and
so x − 8 > 0 and y − 8 > 0.
If (x − 8)(y − 8) = 32 and x ≤ y, then x − 8 is equal to 1, 2 or 4, which gives x = 9, 10, 12
and y − 8 is equal to 32, 16, 8, and so y = 40, 24, 16, respectively.
Using the Pythagorean Theorem, we get z = 41, 26, 20, respectively.
For each of the three possibilities, P = x + y + z 6= 510 and so we conclude k 6= 2.
It can similarly be shown that k cannot equal 3, and thus k ≥ 5 (since k is a prime
number).
Since k is a prime number and k ≥ 5, the positive factor pairs of 8k 2 are
(1, 8k 2 ), (2, 4k 2 ), (4, 2k 2 ), (8, k 2 ), (k, 8k), (2k, 4k)
and the negative factor pairs of 8k 2 are
(−1, −8k 2 ), (−2, −4k 2 ), (−4, −2k 2 ), (−8, −k 2 ), (−k, −8k), (−2k, −4k)
If for example x − 4k = −1 and y − 4k = −8k 2 , then y = 4k − 8k 2 which is less than zero

for all k ≥ 5.
This is not possible since y > 0.
Assuming x ≤ y, it can similarly be shown that when x − 4k and y − 4k are equal to a
negative factor pair of 8k 2 , then y ≤ 0 for all values of k ≥ 5.
Thus, x − 4k and y − 4k must equal a positive factor pair of 8k 2 .
Beginning with the fact that the perimeter of the triangle is 510 cm, we get the following
equivalent equations
x + y + z = 510
x + y = 510 − z
x + 2xy + y 2 = 5102 − 1020z + z 2 (squaring both sides)
2
2xy = 5102 − 1020z (since x2 + y 2 = z 2 )

4A = 5102 − 1020z (A = 21 xy and so 4A = 2xy)
4(510k) = 5102 − 1020z (since A = kP and P = 510)
2k = 255 − z
2k = 255 − (510 − x − y)
x + y − 2k = 255
From the first factor pair, we get x − 4k = 1 and y − 4k = 8k 2 , and so

(x, y) = (1 + 4k, 8k 2 + 4k) (assuming x ≤ y).
Substituting x = 1 + 4k and y = 8k 2 + 4k into x + y − 2k = 255 and simplifying, we get
8k 2 + 6k = 254 or k(4k + 3) = 127, which has no solutions since 127 is a prime number.
We continue our analysis of the remaining 5 factor pairs in the table below.
As before, we make the assumption that x ≤ y, recognizing that the values of x and y
may be interchanged with one another and doing so gives the same value(s) for k.
Factor x y x + y − 2k = 255 Value(s) of k
pair simplified
2, 4k 2 2 + 4k 4k 2 + 4k 2
4k + 6k = 253 No k (LS is even and the RS is odd)
4, 2k 2 4 + 4k 2k 2 + 4k 2k 2 + 6k = 251 No k (LS is even and the RS is odd)
8, k 2 8 + 4k k 2 + 4k k(k + 6) = 247 k = 13
k, 8k 5k 12k 15k = 255 k = 17
2k, 4k 6k 8k 12k = 255 No k (LS is even and the RS is odd)
Therefore, the values of k which satisfy the given conditions are k = 13 and k = 17.
cemc.uwaterloo.ca


Solutions

1. (a) The cost of 12 bags of avocados is $5.00 × 12 = $60.00.

Thus, the chef spent $135.00 − $60.00 = $75.00 on mangoes.
$75.00
The cost of each box of mangoes is $12.50, and so the chef purchased = 6 boxes
$12.50
of mangoes.
(b) Solution 1
A bag of avocados sells for $5.00, and so a 10% discount represents a savings of 100
10
× $5.00
or 0.10 × $5.00 which equals $0.50.
Thus, the discounted price for a bag of avocados is $5.00 − $0.50 = $4.50.
A box of mangoes sells for $12.50, and so a 20% discount represents a savings of 100
20
×$12.50
or 0.20 × $12.50 which equals $2.50.
Thus, the discounted price for a box of mangoes is $12.50 − $2.50 = $10.00.
On Saturdays, the total cost for 8 bags of avocados and 4 boxes of mangoes is
$4.50 × 8 + $10.00 × 4 = $36.00 + $40.00 = $76.00
Solution 2
A bag of avocados sells for $5.00, and so a 10% discount is equivalent to paying
100% − 10% = 90% of the regular price.
90
Thus, the discounted price for a bag of avocados is 100 × $5.00 = 0.90 × $5.00 = $4.50.
A box of mangoes sells for $12.50, and so a 20% discount is equivalent to paying
100% − 20% = 80% of the regular price.
80
Thus, the discounted price for a box mangoes is 100 × $12.50 = 0.80 × $12.50 = $10.00.
On Saturdays, the total cost for 8 bags of avocados and 4 boxes of mangoes is
$4.50 × 8 + $10.00 × 4 = $36.00 + $40.00 = $76.00
(c) Avocados are sold in bags of 6 and the chef needs 100 avocados.
Since 6 × 16 = 96 and 6 × 17 = 102, the chef will need to purchase 17 bags of avocados
(16 bags is not enough).
Mangoes are sold in boxes of 15 and the chef needs 70 mangoes.
Since 15 × 4 = 60 and 15 × 5 = 75, the chef will need to purchase 5 boxes of mangoes.
The total cost of the purchase was $5.00 × 17 + $12.50 × 5 = $85.00 + $62.50 = $147.50.
(d) Avocados are sold for $5.00 per bag, and so the cost to purchase any number of bags is a
whole number.
Mangoes are sold for $12.50 per box, and so the cost to purchase mangoes is a whole
number only when an even number of boxes are bought (1 box costs $12.50, 2 boxes cost
$25.00, 3 boxes cost $37.50, and so on).
Since the chef spends exactly $75.00 (a whole number), then the chef must purchase an
even number of boxes of mangoes.
If the chef purchases 2 boxes of mangoes, the cost is $25.00, which leaves
$75.00 − $25.00 = $50.00 to be used to purchase $50.00 ÷ $5.00 = 10 bags of avoca-
dos.
In this case, the chef has 10 × 6 = 60 avocados and 2 × 15 = 30 mangoes.
Each tart requires 1 avocado and 2 mangoes and so the chef can make 30 ÷ 2 = 15 tarts
(he has more than 15 avocados but only 30 mangoes).
If the chef purchases 4 boxes of mangoes, the cost is 4 × $12.50 = $50.00, which leaves
$75.00 − $50.00 = $25.00 to be used to purchase $25.00 ÷ $5.00 = 5 bags of avocados.
In this case, the chef has 5 × 6 = 30 avocados and 4 × 15 = 60 mangoes.

Each tart requires 1 avocado and 2 mangoes and so the chef can make 30 tarts.
If the chef purchases 6 boxes of mangoes, the cost is 6 × $12.50 = $75.00, which leaves no
money to purchase avocados.
Purchasing more than 6 boxes of mangoes will cost the chef more than $75.00.
Therefore, if the chef purchases 30 avocados (5 bags) and 60 mangoes (4 boxes), she will
have spent exactly $75.00, have twice as many mangoes as avocados, and be able to make
the greatest number of tarts, 30.
2. (a) The parabola y = 41 x2 and the parabolic rectangle are each symmetrical about the y-axis,
and thus a second vertex of the rectangle lies on the parabola and has coordinates (−6, 9).
A third vertex of the parabolic rectangle lies on the x-axis vertically below (6, 9), and thus
has coordinates (6, 0).
Similarly, the fourth vertex also lies on the x-axis vertically below (−6, 9), and thus has
coordinates (−6, 0).
(b) If one vertex of a parabolic rectangle is (−3, 0), then a second vertex has coordinates (3, 0),
and so the rectangle has length 6.
The vertex that lies vertically above (3, 0) has x-coordinate 3.
This vertex lies on the parabola y = 14 x2 and thus has y-coordinate equal to 41 (3)2 = 94 .
The width of the rectangle is equal to this y-coordinate 49 , and so the area of the parabolic
rectangle having one vertex at (−3, 0) is 6 × 49 = 54 4
= 272
.
(c) Let a vertex of the parabolic rectangle be the point (p, 0), with p > 0.
A second vertex (also on the x-axis) is thus (−p, 0), and so the rectangle has length 2p.
The width of this rectangle is given by the y-coordinate of the point that lies on the
parabola vertically above (p, 0), and so the width is 41 p2 .
The area of a parabolic rectangle having length 2p and width 14 p2 is 2p × 14 p2 = 12 p3 .
If such a parabolic rectangle has length 36, then 2p = 36, and so p = 18.
The area of this rectangle is 12 (18)3 = 2916.
If such a parabolic rectangle has width 36, then 14 p2 = 36 or p2 = 144, and so p = 12
(since p > 0).
The area of this rectangle is 12 (12)3 = 864.
The areas of the two parabolic rectangles that have side length 36 are 2916 and 864.
(d) Let a vertex of the parabolic rectangle be the point (m, 0), with m > 0.
A second vertex (also on the x-axis) is thus (−m, 0), and so the rectangle has length 2m.
The width of this rectangle is given by the y-coordinate of the point that lies on the
parabola vertically above (m, 0), and so the width is 41 m2 .
The area of a parabolic rectangle having length 2m and width 14 m2 is 2m × 41 m2 = 12 m3 .
If the length and width of such a parabolic rectangle are equal, then
1 2
4
m
= 2m
2
m = 8m
2
m − 8m = 0
m(m − 8) = 0
Thus m = 8 (since m > 0), and so the area of the parabolic rectangle whose length and
width are equal is 12 (8)3 = 256.
3. (a) For n ≥ 3 and k ≥ 0, the value of T (n, k) is constant

for all possible locations of the k interior points and all
possible triangulations.
Thus, we may use the triangulation shown to determine
that T (3, 2) = 5.
(b) We begin by drawing triangulations to determine the values of T (4, k) for k = 0, 1, 2, 3.
T(4, 0) = 2 T(4, 1) = 4 T(4, 2) = 6 T(4, 3) = 8
Although we would obtain these same four answers by positioning the interior points in
different locations, or by completing the triangulations in different ways, the diagrams
above were created to help visualize a pattern.
From the answers shown, we see that T (4, k + 1) = T (4, k) + 2, for k = 0, 1, 2.
We must justify why this observation is true for all k ≥ 0 so that we may use the result
to determine the value of T (4, 100).
Notice that each triangulation (after the first) was created by placing a new interior point
inside the previous triangulation.
Further, each square is divided into triangles, and so each new interior point is placed
inside a triangle of the previous triangulation (since no 3 points may lie on the same line).
For example, in the diagrams shown to the right,
we observe that P lies in triangle t of the previous
triangulation.
Also, each of the triangles outside of t is untouched by
the addition of P , and thus they continue to contribute t P
the same number of triangles (5) to the value of T (4, 3) T(4, 2) = 6 T(4, 3) = 8
as they did to the value of T (4, 2).
Triangle t contributes 1 to the value of T (4, 2).
To triangulate the region defined by triangle t, P must be joined to each of the 3 vertices
of triangle t (no other triangulation of this region is possible).
Thus, the placement of P divides triangle t into 3 triangles for every possible location of
P inside triangle t.
That is, t contributes 1 to the value of T (4, 2), but the region defined by t contributes 3
to the value of T (4, 3) after the placement of P .
To summarize, the value of T (4, k + 1) is 2 more than the value of T (4, k) for all k ≥ 0
since:
• the (k + 1)st interior point may be placed anywhere inside the triangulation for T (4, k)
(provided it is not on an edge)
• specifically, the (k + 1)st interior point lies inside a triangle of the triangulation which
gives T (4, k)
• this triangle contributed 1 to the value of T (4, k)
• after the (k + 1)st interior point is placed inside this triangle and joined to each of the
3 vertices of the triangle, this area contributes 3 to the value of T (4, k + 1)
• this is a net increase of 2 triangles, and thus T (4, k + 1) = T (4, k) + 2, for all k ≥ 0.
T (4, 0) = 2 and each additional interior point increases the number of triangles by 2.
Thus, k additional interior points increases the number of triangles by 2k, and so
T (4, k) = T (4, 0) + 2k = 2 + 2k for all k ≥ 0.
Using this formula, we get T (4, 100) = 2 + 2(100) = 202.
(c) In the triangulation of a regular n-gon with no interior points, we may choose any one of
the n vertices and join this vertex to each of the remaining n − 3 non-adjacent vertices.
All such triangulations of a regular n-gon with no interior points creates n − 2 triangles,
and so T (n, 0) = n − 2 for all n ≥ 3 (since T (n, 0) is constant).
The reasoning used in part (b) extends to any regular polygon having n ≥ 3 vertices.
That is, each additional interior point that is added to the triangulation for n ≥ 3 vertices
and k ≥ 0 interior points gives a net increase of 2 triangles.
Thus, T (n, k + 1) = T (n, k) + 2 for all regular polygons having n ≥ 3 vertices and k ≥ 0
interior points.
So then k additional interior points increases the number of triangles by 2k, and so
T (n, k) = T (n, 0) + 2k = (n − 2) + 2k for all k ≥ 0.
Using this formula T (n, k) = (n − 2) + 2k, we get T (n, n) = (n − 2) + 2n = 3n − 2 and
3n − 2 = 2020 when n = 2022 3
= 674.
4. (a) Solution 1
If x0 is even, then x20 is even, and so x1 = x20 + 1 is odd.
If x1 is odd, then x21 is odd, and so x2 = x21 + 1 is even.
Thus if x0 is even, then x2 is even and so x2 − x0 is even.
If x0 is odd, then x20 is odd, and so x1 = x20 + 1 is even.
If x1 is even, then x21 is even, and so x2 = x21 + 1 is odd.
Thus if x0 is odd, then x2 is odd and so x2 − x0 is even.
Therefore, for all possible values of x0 , x2 − x0 is even.
Solution 2
Using the definition twice and simplifying, we get
x2 = (x1 )2 + 1
x2 = ((x0 )2 + 1)2 + 1
x2 = (x0 )4 + 2(x0 )2 + 2
x2 = (x0 )4 + 2((x0 )2 + 1)
x2 − x0 = (x0 )4 + 2((x0 )2 + 1) − x0
To show that x2 − x0 is even, we may show that (x0 )4 + 2((x0 )2 + 1) − x0 is even (since
the expressions are equal).
Since 2((x0 )2 + 1) is the product of some integer and 2, this term is even for all possible
values of x0 .
If x0 is even, then (x0 )4 is even, and so (x0 )4 + 2((x0 )2 + 1) − x0 is the sum and difference
of three even terms and thus is even.
If x0 is odd, then (x0 )4 is odd, (x0 )4 − x0 is even, and so (x0 )4 + 2((x0 )2 + 1) − x0 is even.
Thus, x2 − x0 is even for all possible values of x0 .
Solution 3
Using the definition twice and simplifying, we get
x2 = (x1 )2 + 1
x2 = ((x0 )2 + 1)2 + 1
x2 = (x0 )4 + 2(x0 )2 + 2
x2 − x0 = (x0 )4 + 2(x0 )2 − x0 + 2
x2 − x 0 = (x0 )4 + (x0 )2 + (x0 )2 − x0 + 2
x2 − x0 = (x0 )2 ((x0 )2 + 1) + x0 (x0 − 1) + 2
To show that x2 − x0 is even, we may show that (x0 )2 ((x0 )2 + 1) + x0 (x0 − 1) + 2 is even
(since they are equal).
Since x0 − 1 is one less than x0 , then x0 and x0 − 1 are consecutive integers and so one of
them is even.
Thus, the product x0 (x0 − 1) is even.
Similarly, (x0 )2 is one less than (x0 )2 + 1, and thus these are consecutive integers and so
one of them is even.
Therefore, the product (x0 )2 ((x0 )2 + 1) is even.
Since (x0 )2 ((x0 )2 + 1) + x0 (x0 − 1) + 2 is the sum of three even integers, x2 − x0 is even
for all possible values of x0 .
(b) An integer is divisible by 10 exactly when its units (ones) digit is 0.
The difference x2026 − x2020 has units digit 0 exactly when x2026 and x2020 have equal units
digits.
Thus, we will show that for all possible values of x0 , the units digit of x2026 is equal to the
units digit of x2020 , and so x2026 − x2020 is divisible by 10.
When a non-negative integer is divided by 10, the remainder is one of the integers from 0
through 9, inclusive.
Thus for every possible choice for x0 , there exists some non-negative integer k, so that x0
can be expressed in exactly one of the following ways: 10k, 10k + 1, 10k + 2, . . . , 10k + 8,
10k + 9.
If for example x0 has units digit 4, then x0 = 10k + 4 for some non-negative integer k, and
so x1 = (10k + 4)2 + 1 = 100k 2 + 80k + 17 = 10(10k 2 + 8k + 1) + 7, and thus has units
digit 7.
Since x1 is determined by x0 only (x1 = (x0 )2 + 1), the units digit of x1 is uniquely
determined by the units digit of x0 .
For example, if the units digit of x0 is 4, then the units digit of x1 is equal to the units
digit of (4)2 + 1, which equals 7.
More generally, if the units digit of xi (for a non-negative integer i) is equal to u, then the
units digit of xi+1 is equal to the units digit of u2 + 1.
(Can you explain why this is true?)
For example, if we know that x15 = 29, then the units digit of x16 is equal to the units
digit of 92 + 1 = 82, which is 2.
Given that we know all possible units digits of x0 , this provides an efficient method for
determining the units digits of x1 , x2 , x3 , and so on.
In the table below, we list the units digits for the terms x1 through x7 for each of the 10
possible units digits of x0 , 0 through 9 inclusive.
x0 0 1 2 3 4 5 6 7 8 9
x1 1 2 5 0 7 6 7 0 5 2
x2 2 5 6 1 0 7 0 1 6 5
x3 5 6 7 2 1 0 1 2 7 6
x4 6 7 0 5 2 1 2 5 0 7
x5 7 0 1 6 5 2 5 6 1 0
x6 0 1 2 7 6 5 6 7 2 1
x7 1 2 5 0 7 6 7 0 5 2
Looking at the table, we see that for each of the possible units digits for x0 , the units digit
of x1 is equal to the units digit of x7 .
Thus beginning at x1 , each column in the table will repeat every 6 terms, and so indepen-
dent of the starting value x0 , xi+6 and xi have equal units digits for all integers i ≥ 1.
Since 2026 − 2020 = 6, then x2026 and x2020 have equal units digits, and so x2026 − x2020
has units digit 0, and thus is divisible by 10.
(c) Since x115 − 110 = (x115 − 5) − 105, then x115 − 110 is divisible by 105 exactly when x115 − 5
is divisible by 105.
Further, 105 = 3 × 5 × 7 and each of 3, 5, 7 is a prime number, and so x115 − 5 is divisible
by 105 exactly when it is divisible by 3, 5 and 7.
Every xi is a multiple of 3, 1 more than a multiple of 3, or 2 more than a multiple of 3.
Suppose that xi is a multiple of 3. Then xi = 3k for some non-negative integer k, and so
xi+1 = (xi )2 + 1 = (3k)2 + 1 = 3(3k 2 ) + 1
which is 1 more than a multiple of 3.

If xi is 1 more than a multiple of 3, then xi = 3k + 1 for some non-negative integer k, and
so
xi+1 = (xi )2 + 1 = (3k + 1)2 + 1 = 9k 2 + 6k + 2 = 3(3k 2 + 2k) + 2
If xi is 2 more than a multiple of 3, then xi = 3k + 2 for some non-negative integer k, and
so
xi+1 = (xi )2 + 1 = (3k + 2)2 + 1 = 9k 2 + 12k + 5 = 3(3k 2 + 4k + 1) + 2
Each possible choice of x0 is a multiple of 3, 1 more than a multiple of 3, or 2 more than
a multiple of 3.
If x0 is a multiple of 3, then x1 is 1 more than a multiple of 3 and x2 , x3 , x4 , . . . and so on
are each 2 more than a multiple of 3.
If x0 is 1 or 2 more than a multiple of 3, then x1 , x2 , x3 , x4 , . . . and so on are each 2 more
than a multiple of 3.
Therefore, x2 , x3 , x4 , . . . and so on are all 2 more than a multiple of 3 (independent of x0 ),
and so for all i ≥ 2, xi is a number of the form 3k + 2 for some non-negative integer k.
Therefore, x115 − 5 = 3k + 2 − 5 = 3(k − 1) is divisible by 3 for all possible choices of
x0 = n.
Thus, we need to only determine when x115 − 5 is divisible by 5 and 7.
For which of the possible values of x0 is x115 − 5 divisible by 5?
x115 − 5 is divisible by 5 exactly when x115 is divisible by 5.
Every xi is either a multiple of 5, 1 more than a multiple of 5, 2 more than a multiple of
5, 3 more than a multiple of 5, or 4 more than a multiple of 5.
With respect to division by 5, the table below gives the remainders of xi+1 given each of
the 5 possible remainders of xi , 0 through 4 inclusive.
xi xi+1 = (xi )2 + 1
5k 25k 2 + 1 = 5(5k 2 ) + 1
5k + 1 25k 2 + 10k + 2 = 5(5k 2 + 2k) + 2
5k + 2 25k 2 + 20k + 5 = 5(5k 2 + 4k + 1)
5k + 3 25k 2 + 30k + 10 = 5(5k 2 + 6k + 2)
5k + 4 25k 2 + 40k + 17 = 5(5k 2 + 8k + 3) + 2
From this table, we make the following observations:
• if xi is a multiple of 5, then xi+1 is 1 more than a multiple of 5
• if xi is 1 more than a multiple of 5, then xi+1 is 2 more than a multiple of 5
• if xi is 2 more than a multiple of 5, then xi+1 is a multiple of 5
• if xi is 3 more than a multiple of 5, then xi+1 is a multiple of 5
Using these observations, we summarize the remainders of x1 , x2 , x3 , x4 when dividing by
5 for each of the possible remainders for x0 .
x0 0 1 2 3 4
x1 1 2 0 0 2
x2 2 0 1 1 0
x3 0 1 2 2 1
x4 1 2 0 0 2
With respect to division by 5, we see in the table above that for each of the possible
remainders for x0 , the remainder for x1 is equal to that of x4 .
Thus beginning at x1 , each column in the table repeats every 3 terms, and so independent
of the starting value x0 , xi+3 and xi have equal remainders after division by 5 for all
integers i ≥ 1.
Since 115 = 3(37) + 4, then x115 and x4 have the same remainders after division by 5, and
so x115 is divisible by 5 for all choices of x0 = n which are either 2 more than a multiple
of 5 or 3 more than a multiple of 5.
Finally, we want to determine for which of the possible values of x0 is x115 − 5 divisible
by 7.
x115 − 5 divisible by 7 exactly when x115 is 5 more than a multiple of 7.
Every xi is exactly one of: a multiple of 7, 1 more than a multiple of 7, 2 more than a
multiple of 7, and so on up to 6 more than a multiple of 7.
With respect to division by 7, the table below gives the remainders of xi+1 given each of
the 7 possible remainders of xi , 0 through 6 inclusive.
xi xi+1 = (xi )2 + 1
7k 49k 2 + 1 = 7(7k 2 ) + 1
7k + 1 49k + 14k + 2 = 7(7k 2 + 2k) + 2
2
7k + 2 49k 2 + 28k + 5 = 7(7k 2 + 4k) + 5

7k + 3 49k 2 + 42k + 10 = 7(7k 2 + 6k + 1) + 3
7k + 4 49k 2 + 56k + 17 = 7(7k 2 + 8k + 2) + 3
7k + 5 49k 2 + 70k + 26 = 7(7k 2 + 10k + 3) + 5
7k + 6 49k 2 + 84k + 37 = 7(7k 2 + 12k + 5) + 2
From this table, we make the following observations:

• if xi is a multiple of 7, then xi+1 is 1 more than a multiple of 7
Using these observations, we summarize the remainders of x1 , x2 , x3 , x4 when dividing by
7 for each of the possible remainders for x0 .
x0 0 1 2 3 4 5 6
x1 1 2 5 3 3 5 2
x2 2 5 5 3 3 5 5
x3 5 5 5 3 3 5 5
x4 5 5 5 3 3 5 5
Looking at the table, we see that if x0 is 0, 1, 2, 5, or 6 more than a multiple of 7, then xi
is 5 more than a multiple of 7 for all i ≥ 3.
Also, if x0 is 3 or 4 more than a multiple of 7, then xi is 3 more than a multiple of 7 for
all i ≥ 1 (and thus never 5 more than a multiple of 7).
Therefore, x115 − 5 is a multiple of 7 exactly when x0 is not 3 or 4 more than a multiple
of 7.
Summary:
x115 − 110 is divisible by 105 exactly when
• x0 is 2 or 3 more than a multiple of 5, and
• x0 is a multiple of 7 or 1, 2, 5, or 6 more than a multiple of 7.
The values of x0 in the range 1 ≤ x0 ≤ 35 satisfying these properties are:
2, 7, 8, 12, 13, 22, 23, 27, 28, 33
The values of x0 in the range 36 ≤ x0 ≤ 100 satisfying these properties must each be a
mutiple of 5 × 7 = 35 greater than one the numbers in the above list.
Thus, there are 10 possible values for x0 in the original list, 10 more from 2 + 35 = 37 to
33 + 35 = 68, and 9 more from 2 + 2(35) = 72 to 28 + 2(35) = 98, so 29 in total (note that
33 + 2(35) > 100).
If Parsa chooses an integer n with 1 ≤ n ≤ 100 at random and sets x0 = n, the probability
29
that x115 − 110 is divisible by 105 is 100 .
cemc.uwaterloo.ca


Solutions

1. (a) The radius of each hole is 2 cm, and so the diameter of each hole is 4 cm.
Since there are 4 holes, then their diameters combine for a total distance of 4 × 4 = 16 cm
along the 91 cm midline.
Thus, the 5 equal spaces along the midline combine for a total distance of 91−16 = 75 cm,
75
and so the distance along the midline between adjacent holes is = 15 cm.
5
(b) Let the radius of each hole be r cm, and so the diameter of each hole is 2r cm.
Since there are 4 holes, then their diameters combine for a total distance of 4 × 2r = 8r cm
along the midline.
The distance along the midline between adjacent holes is equal to the radius, r cm, and
so these 5 equal spaces along the midline combine for a total distance of 5r cm.
The total length of the midline is 91 cm, and so 8r + 5r = 91 or 13r = 91, and so the
91
radius of each hole is = 7 cm.
13
(c) Solution 1
As in part (b), if the diameter of each hole is 2r cm, then the 4 diameters combine for a
total distance of 4 × 2r = 8r cm along the midline.
If the distance along the midline between adjacent holes is 5 cm, then the 5 equal spaces
along the midline combine for a total distance of 25 cm.
The total length of the midline is 91 cm, and so we get 8r + 25 = 91 or 8r = 66, and so
66
the radius of each hole is = 8.25 cm.
8
However, the vertical distance from the midline to each edge of the metal is 8 cm, and
since the holes must be circles, the radius of each hole cannot be 8.25 cm, and so the
distance between adjacent holes cannot be 5 cm.
Solution 2
The minimum possible distance between adjacent holes is determined by maximizing the
radius of each of the circles.
Since the holes must be circles, and the vertical distance from the midline to each edge of
the metal is 8 cm, the maximum radius is 8 cm.
Since there are 4 holes, then their diameters combine for a maximum total distance of
4 × 16 = 64 cm along the midline.
The total length of the midline is 91 cm, and so the 5 equal spaces combine for a minimum
total length of 91 − 64 = 27 cm.
27
Thus, the minimum distance between adjacent holes is = 5.4 cm, which is greater than
5
the required 5 cm.
2. (a) To add a bump, the line segment of length 21 is first broken into three segments, each
21
having length = 7.
3
The middle segment of these three segments is removed, and two new segments each having
length 7 are added.
Thus, after a bump is added to a segment of length 21, the new path will have length
4 × 7 = 28.
(b) A path with exactly one bump has four line segments of equal length.
240
If such a path has length 240, then each of the four line segments has length = 60.
4
Thus, the original line segment had three line segments each of length 60, and so the
length of the original line segment was 3 × 60 = 180.
(c) To add the first bump, the line segment of length 36 is broken into three segments, each
36
having length = 12.
3
length 12 are added.
Thus, after the first bump is added to a segment of length 36, the new path will have
length 4 × 12 = 48.
Next, a bump is added to each of the four segments having length 12.
Consider adding a bump to one of these four segments.
12
The segment of length 12 is broken into three segments, each having length = 4.
3
The middle segment is removed, two new segments each having length 4 are added, and
so the new length is 4 × 4 = 16.
There are four such segments to which this process happens, and so the total path length
of the resulting figure is 4 × 16 = 64.
(d) To add a bump, the line segment of length n is first broken into three segments, each
n
having length .
3
n
length are added.
3
n 4
Thus, after a bump is added to a segment of length n, Path 1 will have length 4 × = n.
3 3
n
To create Path 2, a bump is added to each of the four segments having length .
3
Consider adding a bump to one of these four segments.
n 1 n
The segment of length is broken into three segments, each having length × .
3 3 3
1 n
The middle segment is removed, two new segments each having length × are added,
3 3
1 n 4
and so the new length is 4 × × = 2 n.
3 3 3
There are four such segments to
2 which this process happens, and so the total length of
4 42 4
Path 2 is 4 × 2 n = 2 n or n.
3 3 3
To summarize, when a bump is added to a line segment of length n, the length of the
4
resulting path, Path 1, is n.
3
2
4
When bumps are then added to Path 1, the length of the resulting Path 2 is n.
3
4
This process will continue with each new path having a total length that is of the
3
previous path length. 3 4
4 4
That is, Path 3 will have length n, Path 4 will have length n, and Path 5 will
3 3
5
4
have length n.
3
If the length of Path 5 is an integer, then n must be divisible by 35 (since there are no
factors of 3 in 45 ).
The smallest integer n which is divisible by 35 is 35 = 243.
The smallest possible integer n for which the length of Path 5 is an integer is 243.
36 + 64 100
3. (a) The arithmetic mean of 36 and 64 is = = 50.
√ 2 √2
The geometric mean of 36 and 64 is 36 · 64 = 62 · 82 = 6 · 8 = 48.
x+y
(b) If the arithmetic mean of two positive real numbers x and y is 13, then = 13.
√ 2
If the geometric mean of two positive real numbers x and y is 12, then xy = 12.
Multiplying the first equation by 2 gives x + y = 26 and so x = 26 − y.
Substituting x = 26 − y into the second equation and squaring both sides
gives (26 − y)y = 122 or y 2 − 26y + 144 = 0.
Factoring the left side of this equation, we get (y − 8)(y − 18) = 0 and so y = 8 or y = 18.
When y = 8, x = 26 − 8 = 18, and when y = 18, x = 8.
That is, the numbers 8 and 18 have an arithmetic mean of 13 and a geometric mean of 12.
x+y √
(c) We are required to solve the equation − xy = 1, for positive integers x and y with
2
x < y ≤ 50.
Simplifying first, we get
x+y √
− xy = 1
2 √
x − 2 xy + y = 2
√ 2 √ √
( x) − 2 xy + ( y)2 = 2 (since x, y > 0)
√ √
( x − y)2 = 2
√ √ √
x− y = ± 2
√ √ √
y− x = 2 (since y > x)
√ √ √
y = x+ 2
√
y = x + 2 + 2 2x
√ √
Since x is a positive integer, then y = 2 + x + 2 2x is a positive integer exactly when 2x
is a positive integer.
This occurs exactly when x = 2m2 for integer values of m.
Since x < y ≤ 50, then 2m2 < 50 or m2 < 25, and so m is any integer from 1 to 4 inclusive.
We determine the corresponding values of x and y in the table below.
√
m x = 2m2 y = 2 + x + 2 2x
1 2 8
2 8 18
3 18 32
4 32 50
The pairs (x, y) satisfying the required conditions are (2, 8), (8, 18), (18, 32), and (32, 50).
4. (a) We begin by multiplying the first equation by 5 and the second equation by 3 to get
15x + 20y = 50
15x + 18y = 3c
By subtracting the second equation from the first, we get
(15x + 20y) − (15x + 18y) = 50 − 3c

or 2y = 50 − 3c. Solving for y, we get

3
y = 25 − c
2
Substituting this into the first of the original two equations, we get

3
3x + 4 25 − c = 10
2
Multiplying the 4 through the parentheses gives 3x + 100 − 6c = 10 which simplifies to
3x = 6c − 90 or x = 2c − 30. Therefore, in terms of c, we have

3
(x, y) = 2c − 30, 25 − c
2
(b) Similar to part (a), we will first solve for x and y in terms of d.
This will give us enough information to determine for which d these values of x and y are
integers.
Multiplying the first equation by 4 we get 4x + 8y = 12.
We can subtract 4x + dy = 6 from 4x + 8y = 12 to get 8y − dy = (8 − d)y = 6.
6
This means y = .
8−d
From the first equation, we get x = 3 − 2y.
Substituting the expression for y into this equation and simplifying, we get

6 12
x=3−2 =3−
8−d 8−d
6 12
We need to find values of d so that y = and x = 3 − are both integers.
8−d 8−d
12
Since 3 is an integer, x will be an integer exactly when is an integer.
8−d
For this to happen, we need 8 − d to be a divisor of 12.
6
However, for y = to be an integer, we need 8 − d to be a divisor of 6.
8−d
Since any divisor of 6 is also a divisor of 12, this means that if y is an integer, then x is
an integer.
Therefore, we need only find integers d so that y is an integer.
The divisors of 6, and hence, the possible values of 8 − d, are −6, −3, −2, −1, 1, 2, 3, and 6.
Therefore, the possible values of d are 2, 5, 6, 7, 9, 10, 11, and 14.
(We may check that each of these values of d give integer values for x and y.)
(c) In order to simplify things, we will first show that, regardless of the values of x, y and k,
if n is an integer, then y must be equal to −1.
To see this, we begin by multiplying the first equation by −2 and the second equation by
3 to get
−(18n + 12)x + (6n + 4)y = −6n2 − 12n − (6k + 10)
(18n + 12)x + (9n2 + 6n)y = −3n2 + (6k + 6)
We now add the two equations.
When doing this, the −(18n + 12) cancels with the (18n + 12) to give
(9n2 + 12n + 4)y = −9n2 − 12n − 4

Observe that (3n + 2)2 = 9n2 + 12n + 4, so we can rewrite this equation as
y(3n + 2)2 = −(3n + 2)2
Suppose (3n + 2)2 = 0. Then 3n + 2 = 0 so n = − 32 , which is not an integer.

Therefore, if we assume n is an integer, we can divide the above equation by (3n + 2)2 to
get
−(3n + 2)2
y= = −1
(3n + 2)2
We are interested in finding positive integer values of k for which there exist integers n so
that the system of equations has a solution (x, y) where x and y are integers.
Thus, going forward, we assume n is an integer which we have shown means y = −1.
The first equation simplifies to
(9n + 6)x − (3n + 2)(−1) = 3n2 + 6n + (3k + 5)
which can be rearranged to get
(9n + 6)x = 3n2 + 3n + 3k + 3 (1)
The second equation becomes
(6n + 4)x + (3n2 + 2n)(−1) = −n2 + (2k + 2)
which rearranges to
(6n + 4)x = 2n2 + 2n + 2k + 2 (2)
Dividing equation (1) by 3 or equation (2) by 2 gives
(3n + 2)x = n2 + n + 1 + k
We want n, k, and x to all be integers, which means we need integers n and k so that
n2 + n + 1 + k is a multiple of 3n + 2.
To simplify this, we will show that n2 + n + 1 + k is a multiple of 3n + 2 if and only if
3(n2 + n + 1 + k) is a multiple of 3n + 2.
Assuming n2 + n + 1 + k is a multiple of 3n + 2, it is certainly true that 3(n2 + n + 1 + k)
is a multiple of 3n + 2.
Notice that 3n + 2 is 2 more than a multiple of 3, so 3n + 2 is not a multiple of 3.
This means 3n + 2 does not have a prime factor of 3, so 3(n2 + n + k + 1) being a multiple
of 3n + 2 means that n2 + n + 1 + k is a multiple of 3n + 2.
To summarize, we want to understand integer pairs (n, k) for which n2 + n + 1 + k is a
multiple of 3n + 2.
We have shown that finding such integer pairs is the same as finding integer pairs (n, k)
for which 3n2 + 3n + 3 + 3k is a multiple of 3n + 2.
By rearranging and factoring, this is the same as finding integer pairs (n, k) so that
n(3n + 2) + n + 3 + 3k is a multiple of 3n + 2.
Noticing that
n(3n + 2) + n + 3 + 3k n + 3 + 3k
=n+
3n + 2 3n + 2
the expression on the right side is an integer if and only if n + 3 + 3k is a multiple of 3n + 2.

Therefore, we need to find integer pairs (n, k) such that n + 3 + 3k is a multiple of 3n + 2.
Using the same reasoning as before, since 3n + 2 does not have a prime factor of 3, we
have that n + 3 + 3k is a multiple of 3n + 2 if and only if 3(n + 3 + 3k) = 3n + 9 + 9k is a
multiple of 3n + 2.
Notice that
3n + 9 + 9k (3n + 2) + (7 + 9k) 7 + 9k
= =1+
3n + 2 3n + 2 3n + 2
7 + 9k
so 3n + 9 + 9k is a multiple of 3n + 2 if and only if is an integer, or 7 + 9k is a
3n + 2
multiple of 3n + 2.
Putting all of this together, given that n and k are integers, the system of equations has
a solution (x, y) where x and y are both integers precisely when 7 + 9k is a multiple of
3n + 2.
Phrasing the question in these terms, we are looking for a positive integer k for which
there are exactly eight integers n so that 3n + 2 is a factor of 9k + 7.
This means we need a positive integer k so that exactly eight of the factors of 9k + 7 are
two more than a multiple of 3.
Beginning with k = 1, we get 9k + 7 = 16.
The factors of 16 are −16, −8, −4, −2, −1, 1, 2, 4, 8, and 16, of which only −16, −4, −1, 2,
and 8 are 2 more than a multiple of 3.
When k = 2, 9k + 7 = 25.
The factors of 25 are −25, −5, −1, 1, 5, and 25.
There are fewer than eight factors in total, so k = 2 does not work.
When k = 3, 9k + 7 = 34.
The factors of 34 are −34, −17, −2, −1, 1, 2, 17, 34, of which only −34, −1, 2, and 17 are 2
more than a multiple of 3.
When k = 4, 9k + 7 = 43 which is prime, so it only has four factors in total which means
k = 4 does not work.
When k = 5, 9k + 7 = 52.
The factors of 52 are −52, −26, −13, −4, −2, −1, 1, 2, 4, 13, 26, and 52, of which only
−52, −13, −4, −1, 2, and 26 are 2 more than a multiple of 3.
When k = 6, 9k + 7 = 61 which is prime, so it only has four factors in total which means
k = 6 does not work.
When k = 7, 9k + 7 = 70.
The factors of 70 are
−70, −35, −14, −10, −7, −5, −2, −1, 1, 2, 5, 7, 10, 14, 35, and 70
Of these, the numbers −70, −10, −7, −1, 2, 5, 14 and 35 are two more than a multiple of 3.
There are exactly 8 numbers in this list.
Therefore, if k = 7, there are exactly eight integers n (−70, −10, −7, −1, 2, 5, 14, and 35),
with the property that the system of equations has a solution (x, y) where x and y are
integers.
(It is worth noting that there are other values of k which satisfy the given properties.)
cemc.uwaterloo.ca

Friday, April 13, 2018

Solutions

17 + 13 + 20 + 12 + 18 + 10 90
1. (a) The average of Aneesh’s first six test scores was = = 15.
6 6
(b) After Jon’s third test, his average score was 14, and so the sum of the scores on his first
three tests was 14 × 3 = 42.
The sum of his scores on his first two tests was 17 + 12 = 29, and so the score on his third
test was 42 − 29 = 13.
17 + 12 + 13
(We may check that the average of 17, 12 and 13 is = 14.)
3
(c) Dina wrote six tests followed by n more tests, for a total of n + 6 tests.
After Dina’s first 6 tests, her average score was 14, and so the sum of the scores on her
first 6 tests was 14 × 6 = 84.
Dina scored 20 on each of her next n tests, and so the sum of the scores on her next n
tests was 20n.
Therefore, the sum of the scores on these n + 6 tests was 84 + 20n.
After Dina’s n + 6 tests, her average score was 18, and so the sum of the scores on her
n + 6 tests was 18(n + 6).
Thus, 84 + 20n = 18(n + 6) or 84 + 20n = 18n + 108 or 2n = 24, and so n = 12.
2. (a) The distance from Botown to Aville is 120 km.

120 4
Jessica drove this distance at a speed of 90 km/h, and so it took Jessica = hours
90 3
4
or × 60 = 80 minutes.
3
(b) The distance from Botown to Aville is 120 km.
The car predicted that Jessica would drive this distance at a speed of 80 km/h, and so it
120 3 3
predicted that it would take Jessica = hours or × 60 = 90 minutes.
80 2 2
The ETA displayed by her car at 7:00 a.m. was 8:30 a.m..
(c) Jessica drove from 7:00 a.m. to 7:16 a.m. (for 16 minutes) at a speed of 90 km/h, and so
16
she travelled a distance of × 90 = 24 km.
60
At 7:16 a.m., Jessica had a distance of 120 km − 24 km = 96 km left to travel.
The car predicted that Jessica would drive this distance at a speed of 80 km/h, and so it
96 6 6
predicted that it would take Jessica = hours or × 60 = 72 minutes to complete
80 5 5
the trip.
The ETA displayed by her car at 7:16 a.m. was 72 minutes later or 8:28 a.m..
(d) As in part (b), the car predicted that it would take Jessica 90 minutes or 1.5 hours to
travel from Botown to Aville.
Let the distance that Jessica travelled at 100 km/h be d km, and so the distance that
Jessica travelled at 50 km/h was (120 − d) km.
d
The time that Jessica drove at 100 km/h was hours.
100
120 − d
The time that Jessica drove at 50 km/h was hours.
50
Since the time predicted by her car is equal to the actual time that it took Jessica to travel
d 120 − d
from Botown to Aville, then + = 1.5.
100 50
Solving for d, we get d + 2(120 − d) = 1.5 × 100 or −d + 240 = 150, and so d = 90 km.
Therefore, Jessica drove a distance of 90 km at a speed of 100 km/h.
3. (a) We are given that T1 = 1, T2 = 2 and T3 = 3.

Evaluating, we get
T4 = 1 + T1 T2 T3 = 1 + (1)(2)(3) = 7, and
T5 = 1 + T1 T2 T3 T4 = 1 + (1)(2)(3)(7) = 43.
(b) Solution 1
Each term after the second is equal to 1 more than the product of all previous terms in
the sequence. Thus, Tn = 1 + T1 T2 T3 · · · Tn−1 .
For all integers n ≥ 2, we use the fact that Tn = 1 + T1 T2 T3 · · · Tn−1 to get
RS = Tn2 − Tn + 1
= Tn (Tn − 1) + 1
= Tn (1 + T1 T2 T3 · · · Tn−1 − 1) + 1
= Tn (T1 T2 T3 · · · Tn−1 ) + 1
= T1 T2 T3 · · · Tn−1 Tn + 1
= Tn+1
= LS
Solution 2
For all integers n ≥ 2, we use the fact that Tn = 1 + T1 T2 T3 · · · Tn−1 to get
LS = Tn+1
= 1 + T1 T2 T3 · · · Tn−1 Tn
= 1 + (T1 T2 T3 · · · Tn−1 )Tn
= 1 + (Tn − 1)Tn
= Tn2 − Tn + 1
= RS
(c) Using the result from part (b), we get Tn + Tn+1 = Tn + Tn2 − Tn + 1 = Tn2 + 1, for all
integers n ≥ 2.
Similarly,
Tn Tn+1 − 1 = Tn (Tn2 − Tn + 1) − 1
= Tn3 − Tn2 + Tn − 1
= Tn2 (Tn − 1) + Tn − 1
= (Tn − 1)(Tn2 + 1)
Since Tn + Tn+1 = Tn2 + 1 and Tn2 + 1 is a factor of Tn Tn+1 − 1, then Tn + Tn+1 is a factor
of Tn Tn+1 − 1 for all integers n ≥ 2.
2
(d) Using the result from part (b), we get T2018 = T2017 − T2017 + 1.
2 2
Since T2017 is a positive integer greater than 1, then T2017 − T2017 + 1 > T2017 − 2T2017 + 1
2 2
and T2017 − T2017 + 1 < T2017 .
2 2 2
That is, T2017 − 2T2017 + 1 < T2017 − T2017 + 1 < T2017 , and so (T2017 − 1)2 < T2018 < T2017
2
.
2 2
Since T2017 − 1 and T2017 are two consecutive positive integers, then (T2017 − 1) and T2017
are two consecutive perfect squares, and so T2018 lies between two consecutive perfect
squares and thus is not a perfect square.
4.(a)(i) By completing the square, the equations defining the two parabolas become
y = x2 − 8x + 17 = x2 − 8x + 16 + 1 = (x − 4)2 + 1, and
y = −x2 + 4x + 7 = −(x2 − 4x + 4) + 11 = −(x − 2)2 + 11.

Thus, the parabola defined by the equation y = x2 − 8x + 17 has vertex V1 (4, 1), and the
parabola defined by the equation y = −x2 + 4x + 7 has vertex V2 (2, 11).
(a)(ii) First, we determine the coordinates of the points of intersection P and Q.
When the two parabolas intersect,
x2 − 8x + 17 = −x2 + 4x + 7
2x2 − 12x + 10 = 0
x2 − 6x + 5 = 0
(x − 5)(x − 1) = 0,
and so the two parabolas intersect at P (5, 2) and Q(1, 10).

Next, we want to show why quadrilateral V1 P V2 Q is a parallelogram.
To do this, we will use the property that if the diagonals of a quadrilateral bisect each
other, then the quadrilateral is a parallelogram.

4 + 2 1 + 11
The midpoint of diagonal V1 V2 is , or (3, 6), and the midpoint of diagonal
2 2
5 + 1 2 + 10
P Q is , or (3, 6).
2 2
Since the midpoint of each diagonal is the same point, (3, 6), then the diagonals bisect
each other and so quadrilateral V1 P V2 Q is a parallelogram.
(Note that we could have also shown that each pair of opposite sides of V1 P V2 Q is parallel.)
(b)(i) By completing the square, the equation defining the parabola y = −x2 + bx + c becomes
y = −x2 + bx + c
= −(x2 − bx) + c
b2 b2
= −(x2 − bx + − ) + c
4 4
2
b b2
= −(x2 − bx + ) + + c
4 4
2
b b
= −(x − )2 + + c.
2 4
b b2

The vertex of this parabola is V3 , + c and the vertex of the parabola defined by
2 4
the equation y = x2 is V4 (0, 0).
First, we determine the conditions on b and c so that the points of intersection R and S
exist and are distinct from one another.
When the two parabolas intersect, −x2 + bx + c = x2 or 2x2 − bx − c = 0.
This equation has two distinct real roots when its discriminant is greater than 0, or when
b2 − 4(2)(−c) > 0.
−b2
The points of intersection, R and S, exist and are distinct from one another when c > .
8
Next, we determine conditions on b and c so that each of R and S are distinct from both
V3 and V4 .
2
The roots √ of the equation 2x − bx − c = 0 are given by the quadratic formula, and so
b ± b2 + 8c
x= .
4 √
b + b2 + 8c
We let the x-coordinate of R be x1 = and the x-coordinate of S be
√ 4
b − b2 + 8c
x2 = .
4 √
b ± b2 + 8c √
Each of the points R and S is not distinct from V4 when = 0 or b = ∓ b2 + 8c
4
or b2 = b2 + 8c, and so c = 0.
Thus, we require that c 6= 0. √
b ± b2 + 8c b
Similarly, each of the points R and S is not distinct from V3 when = or
√ √ 4 2
b ± b2 + 8c = 2b or ± b2 + 8c = b or b2 + 8c = b2 , and so c = 0.
As before, we require that c 6= 0.
(Note that since R and V4 lie on the same parabola, then if their x-coordinates are not
equal, then they are distinct points – that is, we need not consider their y-coordinates.
The same is true for points S and V4 , R and V3 , and S and V3 .)
b b2

Finally, we require that the vertices of the parabolas, V3 , + c and V4 (0, 0), be dis-
2 4
tinct from one another.
Vertices V3 and V4 are distinct provided that if their x-coordinates are equal, then their
y-coordinates are not equal (V3 and V4 lie on different parabolas and so we must consider
both x- and y-coordinates).
b b2 02
If = 0 or b = 0, then +c = + c = c, and since we have the requirement (from
2 4 4
earlier) that c 6= 0, then vertices V3 and V4 are certainly distinct when c 6= 0.
−b2
If the two conditions c > and c 6= 0 are satisfied, then for all pairs (b, c), the points
8
R and S exist, and the points V3 , V4 , R, S are distinct.
(b)(ii) We begin by assuming that the conditions on b and c from part (b)(i) above are satisfied.
Thus, the points R and S exist, and the points V3 , V4 , R, S are distinct.
For quadrilateral V3 RV4 S to be a rectangle, it is sufficient to require that it be a parallel-
ogram that has at least one pair of adjacent sides that are perpendicular to each other.
From (b)(i), the parabolas intersect at R(x1 , x21 ) and S(x2 , x22 ) (R and S each lie on the
parabola y = x2 , and thus the y-coordinates are x21 and x22 , respectively).
Recall that x1 and x2 are the distinct real roots of the quadratic equation 2x2 − bx − c = 0.
−B
The sum of the roots of the general quadratic equation Ax2 + Bx + C = 0 is equal to ,
A
b
and so x1 + x2 = .
2
The product of the roots of the general quadratic equation Ax2 + Bx + C = 0 is equal
C −c
to , and so x1 x2 = .
A 2
First, we will show that quadrilateral V3 RV4 S is a parallelogram since its diagonals bisect
each other.
2
!
b
+ 0 b4 + c + 0 b b2 c b b2 + 4c

2
The midpoint of diagonal V3 V4 is , or , + or , .
2 2 4 8 2 4 8
x1 + x2 x21 + x22

The midpoint of diagonal RS is , .
2 2
2
b 2 2 2 b −c
However, x1 + x2 = and x1 + x2 = (x1 + x2 ) − 2x1 x2 = −2 , and so the
2 2 2
b 2
!
b

2 2

+ c b b c b b + 4c
midpoint of RS is 2 , 2 or , + or , .
2 2 4 8 2 4 8
Since the midpoint of diagonal V3 V4 is equal to the midpoint of diagonal RS, then the
diagonals bisect each other, and so V3 RV4 S is a parallelogram.
Next, we require that any one pair of adjacent sides of quadrilateral V3 RV4 S be perpen-
dicular to each other. (This will mean that all pairs of adjacent sides are perpendicular.)
x2 − 0
The slope of V4 S is 2 = x2 since x2 6= 0 (S(x2 , x22 ) and V4 (0, 0) are distinct points).
x2 − 0
x2 − 0
Similarly, the slope of V4 R is 1 = x1 since x1 6= 0 (R(x1 , x21 ) and V4 (0, 0) are distinct
x1 − 0
points).
Sides V4 S and V4 R are perpendicular to each other if the product of their slopes, x1 x2 , is
equal to −1.
−c −c
Since x1 x2 = , then = −1, and so c = 2.
2 2
−b2
In addition to the condition that c = 2, the two conditions from part (b)(i), c > and
8
c 6= 0, must also be satisfied.
Clearly if c = 2, then c 6= 0.
−b2 −b2
Further, when c = 2, c > becomes 2 > or b2 > −16 which is true for all real
8 8
values of b.
The points R and S exist, the points V3 , V4 , R, S are distinct, and quadrilateral V3 RV4 S
is a rectangle for all pairs (b, c) where c = 2 and b is any real number.
cemc.uwaterloo.ca


Solutions

1. (a) Since ABCD is a cyclic quadrilateral, ∠DCB + ∠DAB = 180◦ or (2u)◦ + 88◦ = 180◦ or
2u = 92 and so u = 46.
(b) Since ST QR is a cyclic quadrilateral, ∠SRQ + ∠ST Q = 180◦ or x◦ + 58◦ = 180◦ and so
x = 122.
Since P QRS is a cyclic quadrilateral, ∠SP Q + ∠SRQ = 180◦ or y ◦ + x◦ = 180◦ or
y + 122 = 180 and so y = 58.
(c) In 4JKL, KJ = KL and so ∠KLJ = ∠KJL = 35◦ (4JKL is isosceles).
In 4JKL, ∠JKL = 180◦ − 2(35◦ ) = 110◦ .
Since JKLM is a cyclic quadrilateral, ∠JM L + ∠JKL = 180◦ or ∠JM L + 110◦ = 180◦ ,
and so ∠JM L = 70◦ .
In 4JLM , LJ = LM and so ∠M JL = ∠JM L = 70◦ (4JLM is isosceles).
In 4JLM , ∠JLM = 180◦ − 2(70◦ ) = 40◦ , and so w = 40.
(d) Solution 1
Since DEF G is a cyclic quadrilateral, ∠DGF + ∠DEF = 180◦ or ∠DGF + z ◦ = 180◦
and so ∠DGF = 180◦ − z ◦ .
Since F GH is a straight angle, then ∠DGF +∠DGH = 180◦ or (180◦ −z ◦ )+∠DGH = 180◦
or ∠DGH = 180◦ − 180◦ + z ◦ = z ◦ .
Solution 2
Since DEF G is a cyclic quadrilateral, ∠DGF + ∠DEF = 180◦ .
Since F GH is a straight angle, then ∠DGF + ∠DGH = 180◦ .
Therefore, ∠DGF + ∠DGH = ∠DGF + ∠DEF , and so ∠DGH = ∠DEF = z ◦ .
2. (a) We complete Row 5 of the table, as shown.

After having completed n rows, n2 integers have been Row 1 1
written. Row 2 1 2 3
Therefore, after having completed 5 rows, 52 = 25 Row 3 1 2 3 4 5
integers have been written. Row 4 1 2 3 4 5 6 7
The 25th integer written in the table is the last integer Row 5 1 2 3 4 5 6 7 8 9
..
in Row 5, which is 9. .
(b) After having completed 10 rows, 102 = 100 integers have been written in the table.
Therefore, the 100th integer written in the table is the last integer in Row 10.
Row n ends at the nth odd integer.
Beginning at 1, the first odd integer is 1 = 2(1) − 1, the second odd integer is 3 = 2(2) − 1,
the third odd integer is 5 = 2(3) − 1.
Beginning at 1, the nth odd integer is 2n − 1 (1 less than the nth even integer 2n).
Row 10 ends at the 10th odd integer, which is 2(10) − 1 = 19.
Therefore, the 100th integer written in the table is 19.
(c) After completing 44 rows, 442 = 1936 integers have been written in the table.
After completing 45 rows, 452 = 2025 integers have been written in the table.
Therefore, the 2017th integer written in the table is in Row 45.
The final integer written in Row 45 is the 45th odd integer, which is 2(45) − 1 = 89.
The 2025th integer written in the table is 89, and so the 2017th integer written in the table
is 2025 − 2017 = 8 less than 89, or 89 − 8 = 81.
(d) Each row, after the first, contains all of the integers that were written in the previous row,
followed by the next two consecutive integers.
For example, Row 3 contains all integers that were written in Row 2 (that is, 1, 2, 3),
followed by the next two consecutive integers, 4 and 5.
Therefore, the first time an integer is written in the table, it appears as the last integer in
a row, or as the second last integer in a row.
Since each row ends at the nth odd integer, then the last integer in each row is odd, and
the second last integer in each row is even.
The first time the integer 96 appears in the table, it is the second last integer written in
a row (since 96 is even), and 97 is the last integer written in that same row.
Since Row n ends with the nth odd integer, 2n − 1, then when 2n − 1 = 97, we get 2n = 98
and so n = 49.
That is, Row 49 ends with the integer 97 and so the second last integer written in Row 49
is 96.
Since 96 first appears in Row 49 of the table, then 96 appears in every row following
Row 49 and does not appear in any row before Row 49.
Therefore, the integer 96 appears in 200 − 48 = 152 of the first 200 rows of the table.
3. (a) When the line y = −15 intersects the parabola with equation y = −x2 + 2x, the
x-coordinates of the two points of intersection satisfy the equation −15 = −x2 + 2x.
Solving this equation, we get x2 − 2x − 15 = 0 or (x + 3)(x − 5) = 0, and so x = −3 or
x = 5.
Since both points of intersection lie on the line y = −15, then the coordinates of the two
points of intersection are (−3, −15) and (5, −15).
(b) The point with x-coordinate 4, on the parabola with equation y = −x2 − 3x, has y-
coordinate −42 − 3(4) = −28.
Therefore, the line intersects the parabola at the point (4, −28).
−28 − 8 −36
The line passes through the point (0, 8), and so the line has slope = = −9.
4−0 4
The line has slope −9 and y-intercept 8, and so the equation of the line is y = −9x + 8.
When the line y = −9x + 8 intersects the parabola with equation y = −x2 − 3x, the
x-coordinates of the two points of intersection satisfy the equation −9x + 8 = −x2 − 3x.
Solving this equation, we get x2 − 6x + 8 = 0 or (x − 2)(x − 4) = 0, and so x = 2 or x = 4.
Therefore, the line intersects the parabola at x = 4 and at x = 2, and so a = 2.
(c) The point with x-coordinate p, on the parabola with equation y = −x2 + kx, has y-
coordinate −p2 + kp.
Therefore, the line intersects the parabola at the point (p, −p2 + kp).
Similarly, the line also intersects the parabola at the point (q, −q 2 + kq).
The slope of the line passing through the points (p, −p2 + kp) and (q, −q 2 + kq) is
(−p2 + kp) − (−q 2 + kq)
, where p 6= q and so p − q 6= 0.
p−q
Simplifying this slope, we get
(−p2 + kp) − (−q 2 + kq) q 2 − p2 + kp − kq

=
p−q p−q
(q − p)(q + p) + k(p − q)
=
p−q
(q − p)(q + p) k(p − q)
= +
p−q p−q
−(p − q)(q + p) k(p − q)
= +
p−q p−q
= −(q + p) + k
= k−q−p
The line has slope k − q − p and passes through the point (p, −p2 + kp).
Therefore, the equation of the line is y − (−p2 + kp) = (k − q − p)(x − p).
(The equation of a line having slope m and passing through the point (x1 , y1 ) is
y − y1 = m(x − x1 ). This is called the point-slope form of a line.)
Finally, we determine the y-intercept of the line by substituting x = 0 into the equation
of the line y − (−p2 + kp) = (k − q − p)(x − p) and solving for y.
y − (−p2 + kp) = (k − q − p)(x − p)

y − (−p2 + kp) = (k − q − p)(0 − p)
y + p2 − kp = −kp + pq + p2
y = −p2 + kp − kp + pq + p2
y = pq
The y-intercept of the line that intersects the parabola with equation y = −x2 + kx at
x = p and at x = q with p 6= q, is pq.
1 1
(d) When the curve x = 3 y 2 + y intersects the parabola with equation y = −x2 + kx,
k k
the x-coordinates of the two points of intersection ((0, 0) and T ) satisfy the equation
1 1
x = 3 (−x2 + kx)2 + (−x2 + kx), where k 6= 0.
k k
Simplifying this equation, we get
1 1
x = 3
(−x2 + kx)2 + (−x2 + kx)
k k
k3x = (−x2 + kx)2 + k 2 (−x2 + kx)
k3x = x4 − 2kx3 + k 2 x2 − k 2 x2 + k 3 x
0 = x4 − 2kx3
0 = x3 (x − 2k)
Since x3 (x − 2k) = 0, then the x-coordinates of the points of intersection of the curve and
the parabola are x = 0 and x = 2k.
Therefore, the x-coordinate of point T is x = 2k, and the y-coordinate of T is
−(2k)2 + k(2k) = −4k 2 + 2k 2 = −2k 2 .
Since the y-coordinate of point T does not contain a linear term in the variable k and does
not contain a constant term, then the equation of the parabola on which all such points
T lie, contains a quadratic term only.

That is, all points T (2k, −2k 2 ) lie on a parabola with an equation of the form y = ax2 .
Substituting, we get −2k 2 = a(2k)2 or −2k 2 = 4ak 2 or −2 = 4a (since k 6= 0), and so
1
a=− .
2
1
(We may verify that x = 2k and y = −2k 2 satisfies y = ax2 + bx + c only if a = − and
2
b = c = 0.)
1
Therefore, the equation of the required parabola is y = − x2 .
2
4. (a) Let N = abcdef ghi be the largest 9-digit zigzag number.
We will determine N by assigning each of the digits 1, 2, 3, 4, 5, 6, 7, 8, and 9 to
a, b, c, d, e, f, g, h, and i.
We begin by trying to construct a zigzag number with a = 9, since any other choice will
give us a smaller zigzag number.
It cannot be that b = 8 since any of the remaining possible choices for c gives a > b > c.
That is, the middle of the first three adjacent digits would be less than the first digit and
greater than the third digit.
In our attempt to find the maximum possible zigzag number, we choose b to equal the
next largest possible number, 7.
Since a > b, then it must be that b < c, which means that c = 8 (since 9 has already been
assigned).
At this point, we have
N = 978def ghi.
If the largest 9-digit zigzag number starts with 9, then it must begin with 978.
We continue to assign values to digits, ensuring that for each group of three adjacent
digits, either the middle digit is greater than each of the other two digits or the middle
digit is less than each of the other two digits.
Since b < c, then c > d.
This tells us that d < e and since N is as large as possible, we choose d = 5 and e = 6 to
give
N = 97856f ghi.
Since d < e, then e > f and thus f < g.
Of the remaining digits, we choose f = 3 and g = 4 to make N as large as possible.
Finally, since f < g, then g > h and thus h < i so we choose h = 1 and i = 2.
Therefore, the largest 9-digit zigzag number is N = 978563412.
(b) We proceed by proving several facts.
Fact 1: G(6, 2) = L(6, 5)
Consider a 6-digit zigzag number counted by G(6, 2), say n = 251634.
We form a new 6-digit number by subtracting each of the digits of n from 7 to obtain
N = 526143. Note that N is a 6-digit zigzag number and one that is counted by L(6, 5).
Consider now an arbitrary 6-digit zigzag number counted by G(6, 2), say n = 2bcdef ,
where b, c, d, e, f are the digits 1, 3, 4, 5, 6 in some order so that 2 < b, b > c, c < d, d > e,
and e < f .
We form a new 6-digit number by subtracting each of the digits of n from 7 to obtain
N = 5(7 − b)(7 − c)(7 − d)(7 − e)(7 − f ).
Since b, c, d, e, f are the digits 1, 3, 4, 5, 6 in some order, then 7 − b, 7 − c, 7 − d, 7 − e, 7 − f
are the digits 6, 4, 3, 2, 1 in some order.

Since 2 < b, then −2 > −b and so 7 − 2 > 7 − b or 5 > 7 − b.
Since b > c, then −b < −c and so 7 − b < 7 − c.
Similarly, 7 − c > 7 − d and 7 − d < 7 − e and 7 − e > 7 − f .
Therefore, N is a 6-digit zigzag number and one that is counted by L(6, 5).
Also, if 2bcdef and 2BCDEF are two different zigzag numbers counted by G(6, 2), then
one of the following must be true: b 6= B or c 6= C or d 6= D or e 6= E or f 6= F .
This means that 5(7−b)(7−c)(7−d)(7−e)(7−f ) and 5(7−B)(7−C)(7−D)(7−E)(7−F )
will be different numbers counted by G(6, 2) as at least one pair of corresponding digits
will be unequal.
In other words, each zigzag number counted by G(6, 2) corresponds to a different zigzag
number counted by L(6, 5), and so G(6, 2) ≤ L(6, 5). (There could be zigzag numbers
counted by L(6, 5) that are not achieved by using this process.)
However, we can apply the same process to an arbitrary zigzag number counted by L(6, 5)
to obtain one counted by G(6, 2), and thus show that L(6, 5) ≤ G(6, 2).
In other words, G(6, 5) = L(6, 2).
Fact 2: G(6, a) = L(6, 7 − a) for a = 1, 2, 3, 4, 5, 6
Using a similar argument to that in Fact #1, we can show that G(6, a) = L(6, 7 − a) for
each of a = 1, 2, 3, 4, 5, 6.
We can now prove the statement from (ii):
G(6, 1) + G(6, 2) + G(6, 3) + G(6, 4) + G(6, 5) + G(6, 6)

= L(6, 6) + L(6, 5) + L(6, 4) + L(6, 3) + L(6, 2) + L(6, 1)
= L(6, 1) + L(6, 2) + L(6, 3) + L(6, 4) + L(6, 5) + L(6, 6)
as required.
We note further that we can generalize Fact 2 in a way that will be useful in (c).
Fact 3: G(n, a) = L(n, n + 1 − a) for each pair of integers a and n with 1 ≤ a ≤ n ≤ 9

To see this, we use a similar argument to that from Fact 1, instead subtracting each digit
of a zigzag number counted by G(n, a) from n+1 to obtain a corresponding zigzag number
counted by L(n, (n + 1) − a).
We now prove that G(6, 3) = L(5, 3) + L(5, 4) + L(5, 5).
Fact 4: The number of zigzag numbers of the form 35cdef equals L(5, 4)
We use G(6, 35) to represent the number of zigzag numbers of the form 35cdef . (We will
use analogous notation later as well.)
Consider a 6-digit zigzag number counted by G(6, 35), say n = 351624.
We form a 5-digit number by deleting the first digit of n to obtain 51624 and then replac-
ing the 4, 5 and 6 with 3, 4 and 5, respectively, to obtain N = 41523.
Note that N is a 5-digit zigzag number and one that is counted by L(5, 4).
Consider now an arbitrary 6-digit zigzag number counted by G(6, 35), say n = 35cdef ,
where c, d, e, f are the digits 1, 2, 4, 6 in some order so that 3 < 5, 5 > c, c < d, d > e, and
e < f.
We form a new 6-digit number by deleting the first digit of n to obtain 5cdef (whose digits
are 1, 2, 4, 5, 6 in some order).

Note that we have 5 > c and c < d and d > e and e < f .
We then replace the digits 4, 5, 6 with 3, 4, 5 respectively to obtain 4c0 d0 e0 f 0 .
Since the digits of the 5cdef are 1, 2, 4, 5, 6 and the digits of 4c0 d0 e0 f 0 are 1, 2, 3, 4, 5 (ar-
ranged in the same order), then we have not changed the relative ordering of the digits of
the number, so we will still have 5 > c0 and c0 < d0 and d0 > e0 and e0 < f 0 .
Therefore, N is a 5-digit zigzag number and one that is counted by L(5, 4).
Also, if 35cdef and 35CDEF are two different zigzag numbers counted by G(6, 3), then
one of the following must be true: b 6= B or c 6= C or d 6= D or e 6= E or f 6= F .
This means that 4c0 d0 e0 f 0 and 4C 0 D0 E 0 F 0 will be different numbers counted by L(5, 4).
(For example, if c0 = C 0 , then we must have had c = C.)
In other words, each zigzag number counted by G(6, 35) corresponds to a different zigzag
number counted by L(5, 4), and so G(6, 35) ≤ L(5, 4).
However, we can apply the reverse process to an arbitrary zigzag number counted by L(5, 4)
(change 3, 4, 5 to 4, 5, 6 and put a 3 on the front) to obtain one counted by G(6, 35), and
thus show that L(5, 4) ≤ G(6, 35).
In other words, G(6, 35) = L(5, 4).
Fact 5: G(6, 3a) = L(5, a − 1) for each of a = 4, 5, 6
Using a similar argument to that in Fact #4, we can show that G(6, 3a) = L(5, a − 1) for
each of a = 4, 5, 6.
We can now prove the statement from (i), noting that the second digit of a zigzag number
counted by G(6, 3) must be 4, 5 or 6 and so each zigzag number counted by G(6, 3) must
be of the form 34cdef or 35cdef or 36cdef :
G(6, 3) = G(6, 34) + G(6, 35) + G(6, 36) = L(5, 3) + L(5, 4) + L(5, 5)
as required.
Before concluding (b), we note the following generalization of Fact 5 that will be use-
ful in (c):
Fact 6: G(n, ab) = L(n − 1, b − 1) for all integers n, a, b with 1 ≤ a < b ≤ n ≤ 9

To see this, we use a similar argument to that from Fact 5, again removing the first digit
a and reducing all digits from a + 1 to n, inclusive, by 1 to obtain digits a to n − 1.
(c) Let T be the total number of 8-digit zigzag numbers. An 8-digit zigzag number starts
with one of 1, 2, 3, 4, 5, 6, 7, 8 and its second digit is either less than or greater than its first
digit.
Therefore, T equals the sum of
` = L(8, 1) + L(8, 2) + L(8, 3) + L(8, 4) + L(8, 5) + L(8, 6) + L(8, 7) + L(8, 8)
and
g = G(8, 1) + G(8, 2) + G(8, 3) + G(8, 4) + G(8, 5) + G(8, 6) + G(8, 7) + G(8, 8)
From Fact 3, L(8, 1) = G(8, 8) and L(8, 2) = G(8, 7) and L(8, 3) = G(8, 6) and so on.
Therefore,
` = G(8, 8) + G(8, 7) + G(8, 6) + G(8, 5) + G(8, 4) + G(8, 3) + G(8, 2) + G(8, 1) = g

and so T = 2(G(8, 1) + G(8, 2) + G(8, 3) + G(8, 4) + G(8, 5) + G(8, 6) + G(8, 7) + G(8, 8)).
We calculate the value of each of the terms on the right side by building a table of values
of G(n, k) for 3 ≤ n ≤ 8 and 1 ≤ k ≤ 8.
Note that if k ≥ n, then G(n, k) = 0 as no n-digit zigzag number can begin with k > n
and if k = n, an n-digit zigzag number starting with n = k cannot have its second digit
greater than its first.
Also, the only 3-digit zigzag numbers are 132, 231, 213 and 312 because the two other
ways of arranging the digits 1, 2 and 3 (123 and 321) do not satisfy the zigzag property.
Thus G(3, 1) = 1, G(3, 2) = 1 and G(3, 3) = 0.
This gives us the following start to the table:
k 1 2 3 4 5 6 7 8
n
3 1 1 0 0 0 0 0 0
4 0 0 0 0 0
5 0 0 0 0
6 0 0 0
7 0 0
8 0
To complete the table, we build a relationship between the values of G in one row and the
values of G in the previous row.
For a positive digit n with 4 ≤ n ≤ 8 and a positive digit k < n, we have:
G(n, k) = G(n, k(k + 1)) + G(n, k(k + 2)) + · · · + G(n, k(n − 1)) + G(n, kn)
(since the second digit must be larger than k)
= L(n − 1, k) + L(n − 1, k + 1) + · · · + L(n − 1, n − 2) + L(n − 1, n − 1) (by Fact 6)
= G(n − 1, n − k) + G(n − 1, n − k − 1) + · · · + G(n − 1, 2) + G(n − 1, 1) (by Fact 3)
Using this formula,
G(4, 1) = G(3, 3) + G(3, 2) + G(3, 1) = 2

G(4, 2) = G(3, 2) + G(3, 1) = 2
G(4, 3) = G(3, 1) = 1
G(5, 1) = G(4, 4) + G(4, 3) + G(4, 2) + G(4, 1) = 5
G(5, 2) = G(4, 3) + G(4, 2) + G(4, 1) = 5
G(5, 3) = G(4, 2) + G(4, 1) = 4
..
.
Proceeding in this way, we complete the table row by row and obtain:
k 1 2 3 4 5 6 7 8
n
3 1 1 0 0 0 0 0 0
4 2 2 1 0 0 0 0 0
5 5 5 4 2 0 0 0 0
6 16 16 14 10 5 0 0 0
7 61 61 56 46 32 16 0 0
8 272 272 256 224 178 122 61 0
Finally, T = 2(272 + 272 + 256 + 224 + 178 + 122 + 61) = 2770, and so the number of
8-digit zigzag numbers is 2770.
cemc.uwaterloo.ca


Solutions

1. (a) Since 5 baskets of raisins fill 2 tubs, then 5 × 6 = 30 baskets of raisins fill 2 × 6 = 12 tubs.
Therefore, 12 tubs of raisins fill 30 baskets.
(b) Since 5 scoops of raisins fill 1 jar, then 5 × 6 = 30 scoops of raisins fill 1 × 6 = 6 jars.
Since 3 scoops of raisins fill 1 cup, then 3 × 10 = 30 scoops of raisins fill 1 × 10 = 10 cups.
Since 30 scoops fill 6 jars, and 30 scoops fill 10 cups, then 10 cups of raisins fill 6 jars.
(c) Solution 1
From part (b), we know that 10 cups of raisins fill 6 jars.
Thus, 10 × 5 = 50 cups of raisins fill 6 × 5 = 30 jars.
Since 30 jars of raisins fill 1 tub, then 50 cups of raisins fill 1 tub, or 50 × 2 = 100 cups of
raisins fill 1 × 2 = 2 tubs.
Since 2 tubs of raisins fill 5 baskets, then 100 cups of raisins fill 5 baskets.
This tells us that 100 ÷ 5 = 20 cups of raisins fill 5 ÷ 5 = 1 basket.
Solution 2
Since 5 baskets fill 2 tubs, then 52 tubs fill 1 basket.
Since 30 jars of raisins fill 1 tub, then 25 × 30 = 12 jars of raisins fill 25 tubs and so fill
1 basket.
Since 5 scoops of raisins fill 1 jar, then 12 × 5 = 60 scoops of raisins fill 12 jars and so fill
1 basket.
Since 3 scoops of raisins fill 1 cup, then 20 × 1 = 20 cups fill 20 × 3 = 60 scoops and so fill
1 basket.
Therefore, 20 cups of raisins fill 1 basket.
2. (a) Since M is the midpoint of chord AB, then AM = 12 (AB) = 5.

Also, since M is the midpoint of chord AB, then OM is perpendicular to AB.
Using the Pythagorean Theorem in 4OM A, we √ get OM 2 = OA2 − AM 2
or OM 2 = 132 − 52 = 169 − 25 = 144, and so OM = 144 = 12 (since OM > 0).
(b) Let the circle have centre O and chord P Q, as shown.
Since the radius is 25, then OQ = 25.
The perpendicular distance from O to the chord is given by OR,
and so OR = 7. O
2 2 2 Q
In 4ORQ, the Pythagorean Theorem gives RQ √ = OQ − OR or
RQ2 = 252 − 72 = 625 − 49 = 576, and so RQ = 576 = 24 (since R
RQ > 0). P
Since OR is perpendicular to the chord P Q, then R is the midpoint of P Q, and so
P Q = 2(RQ) = 2(24) = 48.
Therefore, the length of the chord is 48.
(c) Join O to S and O to U , as shown.
The radius of the circle is 65, and so OS = OU = 65. T
Since OM is perpendicular to chord ST , then M is the midpoint M
of the chord and so M S = 21 (ST ) = 12 (112) = 56.
In 4OM S, the Pythagorean Theorem gives OM 2 = OS√2 −M S 2 or S O
V
OM 2 = 652 −562 = 4225−3136 = 1089, and so OM = 1089 = 33
(since OM > 0). N
Since M N = OM +ON = 72, then ON = 72−OM = 72−33 = 39. U
In 4ON U , the Pythagorean Theorem gives N U 2 = OU 2 − √
ON 2
or N U 2 = 652 − 392 = 4225 − 1521 = 2704, and so N U = 2704 = 52 (since N U > 0).
Finally, since ON is perpendicular to chord U V , then N is the midpoint of the chord and
so U V = 2(N U ) = 2(52) = 104.
Therefore, the length of the chord U V is 104.
3. (a) Since 405 = 34 × 5, then 405 is divisible by 34 but is not divisible by 35 .

Thus, f (405) = 4.
(b) First, we find all factors of 3 which exist in the product 1 × 2 × 3 × 4 × 5 × 6 × 7 × 8 × 9 × 10.
The multiples of 3 are the only numbers which contain factors of 3.
The multiples of 3 in the given product are 3, 6 and 9.
Rewriting the given product, we get
1 × 2 × 3 × 4 × 5 × 6 × 7 × 8 × 9 × 10
= 1 × 2 × 3 × 4 × 5 × (2 × 3) × 7 × 8 × (3 × 3) × 10
= 34 × (1 × 2 × 4 × 5 × 2 × 7 × 8 × 10).
Since the product in parentheses does not include any factors of 3, then the largest power
of 3 which divides the given product is 34 , and so f (1×2×3×4×5×6×7×8×9×10) = 4.
(c) First, we count the number of factors of 3 included in 100!.
Every multiple of 3 includes least 1 factor of 3.
The product 100! includes 33 multiples of 3 (since 33 × 3 = 99).
Counting one factor of 3 from each of the multiples of 3 (these are 3, 6, 9, 12, 15, 18, . . . , 93, 96, 99),
we see that 100! includes at least 33 factors of 3.
However, each multiple of 32 = 9 includes a second factor of 3 (since 9 = 32 , 18 = 32 × 2,
etc.) which was not counted in the previous 33 factors.
The product 100! includes 11 multiples of 9 (since 11 × 9 = 99), and thus there are at least
11 additional factors of 3 in 100!.
Similarly, 100! includes 3 multiples of 33 = 27, each of which contribute an additional
factor of 3 (these are 27 = 33 , 54 = 33 × 2, and 81 = 34 ).
Finally, there is one multiple of 34 = 81 which contributes one more factor of 3.
Since 35 > 100, then 100! does not include any multiples of 35 and so we have counted all
possible factors of 3.
Thus, 100! includes exactly 33 + 11 + 3 + 1 = 48 factors of 3, and so 100! = 348 × t for
some positive integer t that is not divisible by 3.
Counting in a similar way, the product 50! includes 16 multiples of 3, 5 multiples of 9, and
1 multiple of 27, and thus includes 16 + 5 + 1 = 22 factors of 3.
Therefore, 50! = 322 × r for some positive integer r that is not divisible by 3.
Also, 20! includes 6 + 2 = 8 factors of 3, and thus 20! = 38 × s for some positive integer s
that is not divisible by 3.
100! 348 × t 348 × t 318 × t
Therefore, N = = 22 = = .
50!20! (3 × r)(38 × s) (330 × rs) rs
318 × t
Since we are given that N is equal to a positive integer, then is a positive integer.
rs
Since r and s contain no factors of 3 and 318 × t is divisible by rs, then it must be the
case that t is divisible by rs.
318 × t t t
In other words, we can re-write N = as N = 318 × where is an integer.
rs rs rs
t
Since each of r, s and t does not include any factors of 3, then the integer is not
rs
divisible by 3.
100!
Therefore, the largest power of 3 which divides is 318 , and so f (N ) = 18.
50!20!
(d) Since f (a) = 8, then the exponent of the largest power of 3 that divides a is 8.
That is, a = 38 m for some positive integer m and 3 does not divide m.
Since f (b) = 7, then the exponent of the largest power of 3 that divides b is 7.
That is, b = 37 n for some positive integer n and 3 does not divide n.
Substituting and simplifying, we get
a + b = 38 m + 37 n = 37 (3m + n)
Since 3 divides 3m but 3 does not divide n, then 3 does not divide the sum 3m + n.
That is, 3m + n is not a multiple of 3 and so the largest power of 3 that divides a + b is 37 .
Therefore, f (a + b) = 7.
4. (a) (i) For every 10 cents that one restaurant’s price is higher than the other restaurant’s
price, it loses one customer to the other restaurant.
On Monday, LP charges $9.30 − $7.70 = $1.60 more per pizza than what EP charges.
1.60
Therefore, LP loses 0.10 = 16 customers to EP and thus has 50 − 16 = 34 customers.
(ii) The cost for LP to make each pizza is $5.00, and so LP’s profit is $9.30−$5.00 = $4.30
for each pizza sold.
On Monday, LP’s total profit is $4.30 × 34 = $146.20.
(b) Solution 1
Let LP’s price per pizza on Tuesday be $L, where L > 0 and L is an integer multiple of
0.10.
If LP charges $L per pizza, then its profit is $(L − 5) per pizza sold.
We note that if L < 5, then LP’s profit per pizza sold is negative (that is, LP is losing
money on each pizza it sells).
Since EP charges $7.20 per pizza, then the number of customers that LP has is
7.20 − L
50 + .
0.10
7.20 − L
We note that if L < 7.20 (LP charges less per pizza than EP charges), then >0
0.10
7.20 − L
and LP will have more than 50 customers. In fact, LP gains customers.
0.10
7.20 − L
Similarly, if L > 7.20 (LP charges more per pizza than EP charges), then <0
0.10
L − 7.20
and LP will have fewer than 50 customers. In fact, LP loses customers.
0.10
LP’s profit on Tuesday is given by the product of its number of customers and its profit
per pizza sold.
7.20 − L
That is, LP’s profit in dollars, P , is P = 50 + × (L − 5).
0.10
5 + 7.2 − L
Simplifying, we get P = × (L − 5) = 10(12.2 − L)(L − 5).
0.10
Therefore, P is a quadratic function of L.
The graph of this quadratic function, P = 10(12.2 − L)(L − 5), is a parabola opening
downward and thus the maximum profit occurs at its vertex.
The zeros of this parabola occur when 12.2 − L = 0 (that is, when L = 12.2) and when
L − 5 = 0 (that is, when L = 5).

The vertex of the parabola occurs on its axis of symmetry, which is the vertical line passing
through the midpoint of its zeros, L = 12.2 and L = 5.
12.2 + 5 17.2
That is, the maximum profit occurs when L = = = 8.60.
2 2
On Tuesday, LP should charge $8.60 per pizza to maximize their profit.
Solution 2
On Tuesday, EP charges $7.20 per pizza.
Suppose that, on Tuesday, LP charges $(7.20 + 0.10d) per pizza for some integer d. (Note
that LP’s price must be an integer multiple of 10 cents higher or lower than EP’s price.)
If d > 0, then LP will lose d customers to EP.
If d < 0, then LP will gain −d customers from EP.
In other words, on Tuesday, LP will have 50 − d customers.
Since it costs LP $5.00 to make each pizza, then LP’s profit per pizza is equal to
$(7.20 + 0.10d) − $5.00 = $(2.20 + 0.10d).
Therefore, in dollars, LP’s profit on Tuesday is the product of its number of customers
and its profit per pizza sold, or P = (2.20 + 0.10d)(50 − d) = 0.10(22 + d)(50 − d).
Therefore, P is a quadratic function of d.
The graph of this quadratic function, P = 0.10(22 + d)(50 − d), is a parabola opening
The zeros of this parabola occur when 22 + d = 0 (that is, when d = −22) and when
50 − d = 0 (that is, when d = 50).
through the midpoint of its zeros, d = −22 and d = 50.
(−22) + 50
That is, the maximum profit occurs when d = = 14.
2
On Tuesday, LP should charge $(7.20 + 0.10(14)) = $8.60 per pizza to maximize their
profit.
(c) Solution 1
Suppose that EP set its price per pizza at $E, where E > 0 and E is an integer multiple
of 0.20.
After EP sets its price at $E, LP maximizes its profit by setting its price per pizza at $L,
where L > 0 and L is an integer multiple of 0.10.
Let EP’s profit be PE and LP’s profit be PL .
First we determine the price per pizza, $L, that LP will choose in order to maximize its
profit, PL , given that LP knows that EP has set its price per pizza at $E.
LP’s profit per pizza sold is $(L − 5) and, using a similar method as in (b), its number of
E−L
customers is 50 + .
0.10
E−L
Thus, LP’s total profit, in dollars, is given by PL = 50 + × (L − 5).
0.10
5+E−L
Simplifying, we get PL = × (L − 5) = 10(5 + E − L)(L − 5).
0.10
We think about E as fixed and L as variable, making this a quadratic function in L.
The graph of this quadratic function, PL = 10(5 + E − L)(L − 5), is a parabola opening
The zeros of this parabola occur when 5 + E − L = 0 (that is, L = 5 + E) and when
L − 5 = 0 (that is, L = 5).
through the midpoint of its zeros, L = 5 + E and L = 5.
5+E+5 10 + E
That is, the maximum profit for LP occurs when L = = = 5 + 21 E.
2 2
(Since E is a multiple of 0.20, then L is a multiple of 0.10.)
Thus, if EP first sets its price per pizza at $E, then LP should charge $(5 + 12 E) per pizza
to maximize its profit.
Since EP realizes what LP is doing, we can assume that EP now knows that LP will set
their price per pizza at $(5 + 21 E).
Thus, EP may determine its price per pizza, $E, that will maximize its profit.
L−E
EP’s profit per pizza sold is $(E − 5) and its number of customers is 50 + .
0.10
(Since L and E are both multiples of 0.10, then this number
is an integer.)
L−E
Thus, EP’s total profit is given by PE = 50 + × (E − 5).
0.10
5+L−E
Simplifying, we get PE = × (E − 5) = 10(5 + L − E)(E − 5).
0.10
Since L = 5 + 21 E, the quadratic function becomes PE = 10(5 + (5 + 12 E) − E)(E − 5), or
PE = 10(10 − 21 E)(E − 5).
This is again a parabola opening downward and so its maximum profit occurs at its vertex.
The zeros of this parabola occur when E = 20 and when E = 5.
20 + 5
Thus, the maximum profit for EP occurs when E = = 12.50.
2
However, since E must equal an integer multiple of 0.20, then E cannot equal $12.50.
Since the quadratic relation PE is quadratic in E and the resulting parabola opens down-
ward, then values of E closest to the vertex give the largest values corresponding values
of PE .
Therefore, to maximize EP’s profit, we choose the closest values to E = 12.50 that are
multiples of 20 cents.
These values are E = 12.40 (which gives L = 11.20), and E = 12.60 (which gives
L = 11.30).
We note that E = 12.40 and E = 12.60 are symmetric about the axis of symmetry,
E = 12.50, and thus give equal values of PE = 281.20. Further, there are no values of E
which satisfy the given conditions and for which PE is greater in value, since there are no
multiples of 20 cents between $12.40 and $12.50 or between $12.60 and $12.50.
When EP sets its price at E = 12.40, LP’s profit is PL = 10(5 + E − L)(L − 5) or
PL = 10(5 + 12.40 − 11.20)(11.20 − 5) = 10(6.20)(6.20) = 384.40.
When EP sets its price at E = 12.60, LP’s profit is PL = 10(5 + E − L)(L − 5) or
PL = 10(5 + 12.60 − 11.30)(11.30 − 5) = 10(6.30)(6.30) = 396.90.
To maximize its profit, EP could charge $12.40 or $12.60 per pizza, which result in profits
for LP of $384.40 and $396.90, respectively.
Solution 2
On Wednesday, suppose that EP charges $2e per pizza, where e is a multiple of 0.10.
Based on this fixed (but unknown) price, LP chooses its price on Wednesday to maximize
its profit.
Suppose that, on Wednesday, LP charges $(2e+0.10n) per pizza for some integer n. (Note
that LP’s price must be an integer multiple of 10 cents higher or lower than EP’s price.)
As in (b), on Wednesday, LP will have 50 − n customers.

Since it costs LP $5.00 to make each pizza, then LP’s profit per pizza is equal to
$(2e + 0.10n) − $5.00 = $(2e + 0.10n − 5).
Therefore, in dollars, LP’s profit on Wednesday is
PL = (2e+0.10n−5)(50−n) = 0.10(20e+n−50)(50−n) = −0.10n2 +(10−2e)n+(100e−250)
We treat e as a constant and n as a variable. Therefore, PL is a quadratic function of n.

Since the coefficient of n2 is negative, the graph of this quadratic function is a parabola
opening downward and thus the maximum profit for LP occurs at its vertex.
10 − 2e
The vertex occurs when n = − = 50 − 10e.
2(−0.10)
In this case, LP’s profit, in dollars, is
PL = 0.10(20e + (50 − 10e) − 50)(50 − (50 − 10e)) = 0.10(10e)(10e) = 10e2
Now, on Wednesday, EP realizes what LP is doing and so sets its initial price, $2e, to
maximize EP’s profit (knowing that LP will pick its price afterwards to optimize LP’s
profit).
Since EP’s price is set at $2e per pizza, then its profit per pizza is $(2e − 5).
Since LP has 50 − n customers and there are 100 customers in total, then EP has
100 − (50 − n) = 50 + n = 50 + (50 − 10e) = 100 − 10e customers. (From above, we
can assume that n = 50 − 10e.)
Therefore, in dollars, EP’s total profit on Wednesday is
PE = (100 − 10e)(2e − 5) = −20e2 + 250e − 500 = −20(e2 − 12.5e + 25)
Completing the square, we obtain
PE = −20((e − 6.25)2 − 6.252 + 25) = −20(e − 6.25)2 + 281.25
This is the equation of a parabola opening downwards. Thus, the maximum value of PE
occurs when e = 6.25. However, we require that e be a multiple of 0.10.
To find the maximum value(s) of PE including this constraint, we take the closest values
of e to the vertex that are multiples of 0.10. These are e = 6.20 and e = 6.30.
Since e = 6.20 and e = 6.30 are symmetric about the vertex e = 6.25, then they give
the same profit PE , namely PE = 281.20. Since we have stayed as closed to the vertex as
possible, this is EP’s maximum possible profit given the constraints.
When e = 6.20, EP’s price is $12.40 and LP’s profit is $10e2 = $10(6.20)2 = $384.40.
When e = 6.30, EP’s price is $12.60 and LP’s profit is $10e2 = $10(6.30)2 = $396.90.
To maximize its profit, EP should charge $12.40 or $12.60 per pizza, which result in profits
for LP of $384.40 and $396.90, respectively.
cemc.uwaterloo.ca


Solutions

1. (a) The distance from the end of any car (boxcar or engine car) to the end of the next car is
the sum of the 2 m distance between the cars and the 15 m length of a boxcar, or 17 m.
Thus the distance from the end of the engine car to the end of the 10th boxcar (the end
of the train) is 10 × 17 = 170 m.
Since the engine car has a length of 26 m, then the total length of a train with 10 boxcars
is 26 + 170 = 196 m.
(b) The total length of a train with n boxcars is (26 + 15n + 2n) m (one 26 m engine car,
n 15 m boxcars, and a 2 m distance in front of each of the n boxcars).
That is, the total length of a train with n boxcars is (26 + 17n) m.
If a train has a length of 2015 m, then 26 + 17n = 2015 or 17n = 1989 and so n = 117.
A train with a total length of 2015 m has 117 boxcars.
(c) A train with 14 boxcars has length 26 + 17(14) = 264 m.
The length of time during which a portion of the train is in Canada and a portion of the
train is in the United States at the same time is equal to the total length of time it takes
the train to cross the border. (When the train is crossing the border, portions of the train
are in both countries at the same time.)
The train begins to cross the border when the front of the engine car reaches the border.
The train finishes crossing the border when the end of the last boxcar reaches the border
and so the front is 264 m farther.
That is, the length of time required for the train to cross the border is equal to the length
of time it takes the train to travel a distance equal to the length of the train, or 264 m.
Since the train is travelling at a speed of 1.6 m/s, then the time required to travel 264 m
is 264
1.6
= 165 s.
Therefore, the length of time during which a portion of the train is in Canada and a
portion of the train is in the United States at the same time is 165 s.
2. (a) The two-digit positive integers AB and BA equal 10A + B and 10B + A, respectively.
Solving AB − BA = 72, we get (10A + B) − (10B + A) = 72 or 9A − 9B = 72 and so
A − B = 8.
Since A and B are positive digits, then the only possibility for which A − B = 8 occurs
when A = 9 and B = 1. (Verify for yourself that this is indeed the only possibility.)
Therefore, the positive integer AB is 91.
(We may check that AB − BA = 91 − 19 = 72.)
(b) The two-digit positive integers M N and N M equal 10M + N and 10N + M , respectively.
Solving M N − N M = 80, we get (10M + N ) − (10N + M ) = 80 or 9M − 9N = 80 and
so 9(M − N ) = 80.
Since M and N are positive digits, then M −N is an integer and so 9(M −N ) is a multiple
of 9.
However, 80 is not a multiple of 9 and so 9(M − N ) 6= 80.
Therefore, it is not possible that M N − N M = 80.
(c) The three-digit positive integers P QR and RQP equal 100P +10Q+R and 100R+10Q+P ,
respectively.
Simplifying P QR − RQP , we get (100P + 10Q + R) − (100R + 10Q + P ) or 99P − 99R
or 99(P − R).
Since P and R are positive digits, the maximum possible value of P − R is 8 (which occurs
when P is as large as possible and R is as small as possible, or P = 9 and R = 1).
Since P > R, the minimum possible value of P − R is 1 (which occurs when P = 9 and
R = 8, for example).
That is, 1 ≤ P − R ≤ 8 and so there are exactly 8 possible integer values of P − R.

(Verify for yourself that there are values for P and R so that P − R is equal to each of
the integers from 1 to 8.)
Since P QR − RQP = 99(P − R) and there are exactly 8 possible values of P − R, then
there are exactly 8 possible values of P QR − RQP .
(We note that the value of P QR − RQP does not depend on the value of the digit Q.)
3. (a) Diagram 1 illustrates that T (3) = 9.
To determine T (4), add 1 line segment to Diagram 1 as shown
in Diagram 2.
We are told that this new (4th ) line segment must intersect
each of the existing 3 line segments exactly once, creating 3
new points of intersection (labelled 1, 2, 3).
This 4th line segment also adds 2 new endpoints (labelled 4
and 5) distinct from the previous 3 new points.
In addition, each of the points which exist in the illustration of Diagram 1
T (3) (Diagram 1) continue to exist in the illustration of T (4)
(Diagram 2) and are distinct from each of the new points
which were added.
Therefore, we get 4 1 2 3 5
T (4) = T (3) + 3 + 2
= 9+3+2
= 14
Diagram 2
Diagram 2 illustrates that T (4) = 14.
To determine T (5), add 1 line segment to Diagram 2 as shown
in Diagram 3.
We are told that this new (5th ) line segment must intersect
6
each of the existing 4 line segments exactly once, creating 4
new points of intersection (labelled 1, 2, 3, 4). 3
4
This 5th line segment also adds 2 new endpoints (labelled 5 2

5 1
and 6) distinct from the previous 4 new points.
In addition, each of the points which exist in the illustration of
T (4) (Diagram 2) continue to exist in the illustration of T (5) Diagram 3
(Diagram 3) and are distinct from each of the new points
which were added.
Therefore, we get
T (5) = T (4) + 4 + 2
= 14 + 4 + 2
= 20
Therefore, T (4) = 14 and T (5) = 20.

(b) As in part (a), consider finding T (n) with the help of (in terms of) T (n − 1) for any integer
n ≥ 2.
To determine T (n), add 1 line segment to any illustration of T (n − 1).
This new (nth ) line segment must intersect each of the existing n − 1 line segments exactly
once, creating n − 1 new points of intersection.
This nth line segment also adds 2 new endpoints (distinct from the previous n − 1 points).
In addition, each of the points which exist in the illustration of T (n − 1) continue to exist
in the illustration of T (n) and are distinct from each of the new points which were added.
Therefore, we get T (n) = T (n − 1) + (n − 1) + 2 or T (n) = T (n − 1) + n + 1 and so
T (n) − T (n − 1) = n + 1 for all n ≥ 2.
(c) From part (b), T (n) − T (n − 1) = n + 1 and so T (n) = T (n − 1) + n + 1.
That is, the addition of an nth line segment increases T (n − 1) by n + 1.
For example since T (1) = 2, then T (2) = T (1) + 3 = 2 + 3.
For small values of n, we determine T (n) in the table below.
n T (n) = T (n − 1) + n + 1, n ≥ 2
2 T (2) = T (1) + 3 = 2 + 3
3 T (3) = T (2) + 4 = 2 + 3 + 4
4 T (4) = T (3) + 5 = 2 + 3 + 4 + 5
5 T (5) = T (4) + 6 = 2 + 3 + 4 + 5 + 6
6 T (6) = T (5) + 7 = 2 + 3 + 4 + 5 + 6 + 7
.. ..
. .
We may use the pattern in the table above to establish an equation for T (n).
What is the pattern?
Consider for example the row for n = 5.
T (5) is the sum of the positive integers from 2 to n + 1 = 5 + 1 = 6.
This is true for each of the rows shown in the table.
That is, T (n − 1) = 2 + 3 + 4 + · · · + n for any positive integer n ≥ 3.
(Verify that this is true for each of the rows shown in the table.)
Since the addition of an nth line segment increases T (n − 1) by n + 1, then
T (n) = T (n − 1) + n + 1 and so T (n) = (2 + 3 + 4 + · · · + n) + n + 1.
Reorganizing this equation for T (n), we get
T (n) = 1 + 2 + 3 + 4 + · · · + n + n.
n(n + 1)
Since the sum of the first n positive integers 1 + 2 + 3 + 4 + · · · + n is equal to ,
2
n(n + 1)
then T (n) = + n.
2
Solving T (n) = 2015, we get
n(n + 1)
+n = 2015
2
n(n + 1) + 2n = 4030
n2 + 3n = 4030
n2 + 3n − 4030 = 0
(n − 62)(n + 65) = 0
and so n = 62 (since n > 0). (We could use the quadratic formula if we didn’t see how to
factor the quadratic.)
Therefore, n = 62 is the only value of n for which T (n) = 2015.
4. (a) Since 125 = 53 , then the positive divisors of 125 are 50 , 51 , 52 , 53 or 1, 5, 25, 125.
Therefore, for each positive integer a from 1 to 125 inclusive, the possible values of
gcd(a, 125) are 1, 5, 25, 125.
The gcd(a, 125) = 125 exactly when a is divisible by 125. Since 1 ≤ a ≤ 125 and there is
only one multiple of 125 in this range, then gcd(a, 125) = 125 only when a = 125.
gcd(a, 125) = 25 exactly when a is divisible by 25 and not by 125.
These a are each of the form a = 25k for some positive integer k.
The possible values of a in the range 1 ≤ a < 125 are 25, 50, 75, and 100.
gcd(a, 125) = 5 exactly when a is divisible by 5 and not by 25.
There are 25 multiples of 5 between 1 and 125, inclusive.
5 of these have a gcd with 125 of 25 or 125, as above.
The remaining 25 − 5 = 20 multiples of 5 must have a gcd with 125 of 5.
gcd(a, 125) = 1 exactly when a is not divisible by 5.
Since there are 25 multiples of 5 between 1 and 125, inclusive, then there are 125−25 = 100
integers in this range that are not multiples of 5.
In summary, when the positive integers a with 1 ≤ a ≤ 125 are considered, there is/are
• 1 integer a for which gcd(a, 125) = 125
• 4 integers a for which gcd(a, 125) = 25
Therefore,
P (125) = gcd(1, 125) + gcd(2, 125) + · · · + gcd(124, 125) + gcd(125, 125)

= 1(125) + 4(25) + 20(5) + 100(1)
= 425
(b) Since r is a prime number, the positive divisors of r2 are 1, r, r2 .

Since s is a prime number, the positive divisors of s are 1, s.
Since r and s are different prime numbers, then gcd(r2 , s) = 1. (There are no common
positive divisors other than 1 in these two lists.)
Therefore, P (r2 s) = P (r2 )P (s), from the given fact.
To calculate P (s), we proceed as in (a).
For each positive integer a with 1 ≤ a ≤ s, the possible values of gcd(a, s) are 1 and s.
The only multiple of s in the given range is a = s, so there is only one a (namely a = s)
for which gcd(a, s) = s.
There are thus s − 1 integers a for which gcd(a, s) = 1.
Therefore,
P (s) = gcd(1, s) + gcd(2, s) + · · · + gcd(s − 1, s) + gcd(s, s)
= 1(s) + (s − 1)1
= 2s − 1
In a similar way, there is 1 integer a with 1 ≤ a ≤ r2 for which gcd(a, r2 ) = r2 , and r − 1

integers a with gcd(a, r2 ) = r, and r2 − r integers a with gcd(a, r2 ) = 1.
Therefore,
P (r2 ) = gcd(1, r2 ) + gcd(2, r2 ) + · · · + gcd(r2 − 1, r2 ) + gcd(r2 , r2 )
= 1(r2 ) + (r − 1)r + (r2 − r)1
= 3r2 − 2r
= r(3r − 2)
Therefore, P (r2 s) = P (r2 )P (s) = r(3r − 2)(2s − 1), as required.

(c) We prove that P (r2 s) can never be equal to a power of a prime number by assuming
that P (r2 s) equals tn for some prime number t and positive integer n, and obtaining a
contradiction.
From (b), we obtain r(3r − 2)(2s − 1) = tn .
Since the right side is a power of a prime number, then the only divisors of the right side
are powers of t.
Therefore, each of the factors on the left side must be powers of t.
Since r is a factor on the left side and r is itself a prime, then r = t.
Therefore, 3r − 2 = 3t − 2 must also be a power of t.
If 3t − 2 = t, then t = 1, which is not prime.
If 3t − 2 = t2 and t is prime, then t2 − 3t + 2 = 0 or (t − 2)(t − 1) = 0 and so t = 2 or
t = 1. Since t is to be a prime number, then if 3t − 2 = t2 , we must have t = 2.
This also tells us that if t = 2, then 3t − 2 is only a power of a prime when 3t − 2 = t2 .
Can 3t − 2 = tu for some integer u > 2 and prime t > 2?
If u > 2, then tu > t2 , since t > 1.
If t > 2, then t2 − 3t + 2 = (t − 2)(t − 1) > 0 so t2 > 3t − 2.
Thus, if u > 2, then tu > t2 > 3t − 2, so 3t − 2 6= tu .
Therefore, if 3t − 2 is a power of t, then t = 2.
If t = 2, then t and 3t − 2 are both powers of t = 2. Furthermore, t = 2 is the only prime
for which this can work.
But in this case, 2s − 1 is odd and so cannot be a power of t = 2.
This contradicts our original assumption.
Therefore, P (r2 s) = r(3r − 2)(2s − 1) cannot be the power of a prime number.
(d) We note that 243 = 35 . We try some different possible forms for m.
Suppose that m = rs for some prime numbers r and s.
Then P (m) = P (rs) = (2r − 1)(2s − 1).
Are there prime numbers r and s for which (2r − 1)(2s − 1) = 243?
Note that if p is prime, then p ≥ 2 so 2p − 1 ≥ 3.
Therefore, we could have 2r − 1 = 3 and 2s − 1 = 81 (which gives r = 2 and s = 41 which
are both prime) or 2r − 1 = 9 and 2s − 1 = 27 (which gives r = 5 (prime) and s = 14 (not
prime)).
Therefore, P (82) = P (2 · 41) = 3 · 81 = 243.
Since 243 is a power of a prime, then from part (c), P (r2 s) 6= 243 for all primes r and s.
Let’s next see if P (r3 s) can equal 243.
Using a similar derivation to that in (a), we can see that
P (r3 ) = 1(r3 ) + (r − 1)r2 + (r2 − r)r + (r3 − r2 )(1) = 4r3 − 3r2
Therefore, P (r3 s) = P (r3 )P (s) = (4r3 − 3r2 )(2s − 1) = r2 (4r − 3)(2s − 1).
If r2 (4r − 3)(2s − 1) = 243, then since r is prime and a factor of the left side, we must
have r = 3.
Therefore, 32 (4(3) − 3)(2s − 1) = 243 or 81(2s − 1) = 243 or 2s − 1 = 3, which gives s = 2.
Thus, P (54) = P (33 · 2) = 243.
Therefore, m = 82 and m = 54 both satisfy P (m) = 243.
cemc.uwaterloo.ca


Solutions
c 2014 University of Waterloo

p √
1. (a) Using the given definition, 8 7 = 8 + 4(7) = 36 = 6.
√
(b) Since 16 n = 10, then 16 + 4n = 10 or 16 + 4n = 100 (by squaring both sides) or
4n = 84 and so n = 21. p √
We check that indeed 16 n = 16 21 = 16 + 4(21) = 100 = 10.
p √
(c) We first determine the value insidepthe brackets: √9 18 = 9 + 4(18) = 81 = 9.
So then (9 18) 10 = 9 10 = 9 + 4(10) = 49 = 7.
√ √
(d) Using the definition, k k = k + 4k √ = 5k.
So we are asked to solve the equation 5k = k.
Squaring both sides we get, 5k = k 2 and so k 2 − 5k = 0 or k(k − 5) = 0, and so k = 0 or
k = 5. p √
Checking k = 0, we obtain k k = 0 0 = p0 + 4(0) = √0 = 0 = k, as required.
Checking k = 5, we obtain k k = 5 5 = 5 + 4(5) = 25 = 5 = k, as required.
Thus, the only possible solutions are k = 0 and k = 5.
2. (a) The song’s position on week 1 (w = 1), is P (1) = 3(1)2 − 36(1) + 110 = 77.
(b) The song’s position is given by the quadratic function P = 3w2 − 36w + 110, the graph of
which is a parabola opening upward.
The minimum value of this parabola is achieved at its vertex.
To find the coordinates of the vertex, we may complete the square.
P = 3w2 − 36w + 110

= 3(w2 − 12w) + 110
= 3(w2 − 12w + 36 − 36) + 110
= 3(w2 − 12w + 36) − 108 + 110
= 3(w − 6)2 + 2
Therefore, the vertex of the parabola occurs at w = 6 and P = 2.

(i) The best position that the song “Recursive Case” reaches is position #2.
(ii) The song reaches its best position on week 6.
(c) To determine the last week that “Recursive Case” appears on the Top 200 list, we want
to find the largest w such that P = 3w2 − 36w + 110 ≤ 200.
Using the vertex form from part (b), we have 3(w − 6)2 + 2 ≤ 200 or 3(w − 6)2 ≤ 198 or
(w − 6)2 ≤ 66.
To determine the largest positive integer w such that (w − 6)2 ≤ 66, we want to find the
largest square that is less than or equal to 66.
Since 82 ≤ 66 and 92 > 66, then the largest w satisfies w − 6 = 8 and so w = 14.
The last week that “Recursive Case” appears on the Top 200 list is week 14.
To check this we note that,
P (14) = 3(14 − 6)2 + 2 = 194 ≤ 200 but P (15) = 3(15 − 6)2 + 2 = 245 > 200.
E
3. (a) We will denote the area of a figure using vertical bars.
For example, |4BCE| is the area of 4BCE. D
18
Since ABCD has equal side lengths (it is a square) and
EA = EB = EC = ED, then the 4 triangular faces of A
pyramid ABCDE are congruent and so all have equal area. C F
20
The surface area of pyramid ABCDE is equal to the sum of
the base area and the areas of the 4 triangular faces or B
|ABCD| + |4EAB| + |4EBC| + |4ECD| + |4EDA| = |ABCD| + 4|4EAB|.
Square ABCD has side length 20 and so |ABCD| = 20 × 20 = 400. E
To determine |4EAB|, we construct altitude EJ as shown.
4EAB is isosceles and so EJ bisects AB with AJ = JB = 10.
4EAJ is a right-angled triangle and so by the Pythagorean 18 18
2 2 2 2 2 2
Theorem,
√ EA = AJ √+ EJ or 18 = 10 + EJ , so then
EJ = 224 or EJ = 4 14 (since EJ > 0).
B 10 J 10 A
1 1
√ √
The area of 4EAB is 2 (AB)(EJ) = 2 (20)(4 14) = 40 14.
√ √
Thus the surface area of ABCDE is |ABCD|+4|EAB| = 400+4(40 14) = 400+160 14.
E
(b) As in part (b), J is positioned
√ such that EJ is an altitude of
4EAB and so EJ = 4 14. Since EF is perpendicular to D
18
the base of the pyramid, then EF is perpendicular to F J, as
shown. A
Further, F is the centre of the base ABCD and J is F
C
J
the midpoint of AB, so then F J is parallel to CB and
20
F J = 21 × CB = 21 × 20 = 10. B
2 2 2 2
By the√ Pythagorean
√ Theorem, EJ = EF + F J or 224 = EF + 100 and so
EF = 124 = 2 31 (since EF > 0). √
Therefore, the height EF of the pyramid ABCDE is 2 31.
(c) Points G and H are the midpoints of ED and EA, respectively, and so
EG = GD = EH = HA = 9.
Thus, GH is a midsegment of 4EDA and so GH is parallel to DA and GH = 21 ×DA = 10.
(Note that this result follows from the fact that 4EGH is similar to 4EDA. Can you
prove this?)
Since GH is parallel to DA and DA is parallel to CB, then GH is parallel to CB.
That is, quadrilateral BCGH (whose area we are asked to find) is a trapezoid.
To determine the area of trapezoid BCGH, we need the lengths of the parallel sides
(GH = 10 and CB = 20) and we need the perpendicular distance between these two
parallel sides.
We will proceed by showing that HT (in the diagram below) is such a perpendicular height
of the trapezoid and also by determining its length.
Join H to I, the midpoint of EB, so that HI is a midsegment E
of 4EAB with HI = 10. G

9
Position P on the base of the pyramid such that HP is D H
perpendicular to the base. 9
Similarly, position M on the base such that IM is perpen- I K A

P
dicular to the base. C
20
Let M P extended intersect the edge BC at T and the edge M
AD at K, as shown. T
B
By symmetry, HP = IM and so HP M I is a rectangle with P M = HI = 10.

Further, since AB is parallel to HI and HI is parallel to KT (both are perpendicu-
lar to HP and IM ), then AB is parallel to KT . So then ABT K is a rectangle and
KT = AB = 20.
Also by symmetry, P M is centred on line segment KT such that KP = M T = 20−10 2
=5
(EA = EB and E lies vertically above the centre of the square base).
E
Therefore P T = P M + M T = 10 + 5 = 15.
Next, let the midpoint of HI be L and position N on the 9
D H
base of the pyramid such that LN is perpendicular to the L
9
base.
I A
Since F is the centre of the square and J is the midpoint of F
edge AB, then F J passes through N . C N
J
Since 4EF J is similar to 4LN J (by AA ∼), then
LN LJ 1 B
= = (since HI is a midsegment of 4EAB).
EF EJ 2 √ √
Therefore, LN = 12 (EF ) = 12 (2 31) = 31.
√
Since HI is parallel to the base of the pyramid, ABCD, then HP = LN = 31 (both
are perpendicular to the base). √ E
In right-angled 4HP T , HT 2 = HP 2 + P T 2 = ( 31)2 + 152 . 9
G
So HT 2 = 256 and HT = 16 (since HT > 0). D H
Since the plane containing 4HP T is perpendicular to the 9
base ABCD, then HT is perpendicular to BC. K A
That is, HT is the height of trapezoid BCGH. P
C
Finally, |BCGH| = HT 2
(GH + CB) = 16 2
(10 + 20) = 240. 20
T
B
4. (a) If (4, y, z) is an APT, then 42 + y 2 = z 2 + 1 or z 2 − y 2 = 15 and so (z − y)(z + y) = 15.

Since y and z are positive integers, then (z + y) is a positive integer and thus (z − y) is
also a positive integer (since the product of the two factors is 15).
That is, (z − y) and (z + y) are the possible pairs of factors of 15, of which there are two:
1 and 15, and 3 and 5.
Since z + y > z − y, we have the following two systems of equations to solve:
z−y =1 z−y =3
z + y = 15 z+y =5
Adding the first pair of equations, we get 2z = 16 or z = 8 and so y = 7.
Adding the second pair of equations, we get 2z = 8 or z = 4 and so y = 1.
Since y > 1, this second solution is not an APT.
The only APT with x = 4 is (4, 7, 8).
(b) Let positive integers u, v, w be the lengths of the sides of 4U V W (with side length u
opposite vertex U , v opposite V , and w opposite W ).
Without loss of generality, assume (u, v, w) forms an APT such that u2 + v 2 = w2 + 1 with
u > 1 and v > 1.
The area of 4U V W is given by A = 21 uv sin W (?) (we will derive this formula at the
end of the solution).
Assume that this area, A, is an integer.
u2 + v 2 − w2
In 4U V W , the cosine law gives w2 = u2 + v 2 − 2uv cos W or cos W = .
2uv
However, (u, v, w) is an APT and thus u2 + v 2 = w2 + 1 and so u2 + v 2 − w2 = 1.

1
Substituting, we get cos W = and since sin2 W = 1 − cos2 W , then
2 2uv
2 1
sin W = 1 − .
2uv
2 !
1
Squaring (?) and substituting for sin2 W , we get A2 = 14 u2 v 2 1 − .
2uv
2 2

1 (2uv) − 1 1 (2uv) − 1
Simplifying, this last equation we get A2 = 4 u2 v 2 or A2 = 4 u2 v 2
(2uv)2 4u2 v 2
2 2 2 2 2
or 16A = (2uv) − 1 and so 4u v − 16A = 1.
The left side of this equation has a common factor of 4 and so is an even integer for all
integers u, v, A.
However, the right side of the equation is 1 (an odd integer) and so this equation has no
solutions.
Our assumption that A is an integer is false and thus the area of any triangle whose sides
lengths form an APT is not an integer.
Derivation of A = 12 uv sin W
In 4U V W , construct the perpendicular from V meeting side U W (or side U W extended)
at T .
In this case, 4U V W has height V T and base U W .
VT
In right-angled 4V W T , we have sin W = and so V T = u sin W .
VW
(Note that if ∠W is obtuse, then height V T lies outside the triangle and we have
VT
sin(180◦ − W ) = and so V T = u sin W since sin(180◦ − W ) = sin W ).
VW
Thus, the area of 4U V W is 12 (U W )(V T ) = 21 uv sin W .
(c) Since (5t + p, bt + q, ct + r) is an APT, then (5t + p)2 + (bt + q)2 = (ct + r)2 + 1.
Expanding, we get 25t2 + 10pt + p2 + b2 t2 + 2bqt + q 2 = c2 t2 + 2crt + r2 + 1, or
(25 + b2 )t2 + (10p + 2bq)t + (p2 + q 2 ) = c2 t2 + 2crt + (r2 + 1).
Each side of this equation is a quadratic polynomial in the variable t.
Since this equation is true for all positive integers t, then the corresponding coefficients
on the left side and the right side are equal.
That is,
25 + b2 = c2 (1)
10p + 2bq = 2cr (2)
p2 + q 2 = r 2 + 1 (3)
Equation (1) is the Pythagorean relationship, and since b, c are positive integers then the
Pythagorean triple (5, 12, 13) satisfies (1) with b = 12 and c = 13.
(These are in fact the only values of b and c that work but we don’t need to show this.)
Substituting b and c into (2) and simplifying we get 5p + 12q = 13r (4).
Squaring equation (4), we get 25p2 + 120pq + 144q 2 = 169r2 (5).
From equation (3), we get r2 = p2 + q 2 − 1, and substituting this into (5) gives
25p2 + 120pq + 144q 2 = 169(p2 + q 2 − 1) or 144p2 − 120pq + 25q 2 = 169 or (12p − 5q)2 = 169
and so 12p − 5q = ±13.
We search for positive integers p and q with p ≥ 100 satisfying 12p − 5q = ±13 or
5q = 12p ± 13.
Since 12p is even for all integers p and 13 is odd, then 12p ± 13 is odd and so 5q must also
be odd.
The units digit of 5q is either 0 or 5 for all positive integers q and since 5q is odd, then its
units digit is 5.
The units digit of 5q − 13 is 2 and the units digit of 5q + 13 is 8, and so the units digit of
12p is either 2 or 8 (since 12p = 5q ± 13).
A value of p ≥ 100 such that 12p has units digit 2 is p = 101.
Substituting, we get 12(101) = 5q − 13 or 5q = 1225 and so q = 245.
Substituting p and q into (4) gives r = 265.
A value of p ≥ 100 such that 12p has units digit 8 is p = 104.
Substituting, we get 12(104) = 5q + 13 or 5q = 1235 and so q = 247.
Substituting p and q into (4) gives r = 268.
Therefore, two possible 5-tuples (b, c, p, q, r) satisfying the given conditions are
(12, 13, 101, 245, 265) and (12, 13, 104, 247, 268).
(Note that there are an infinite number of possible solutions.)
cemc.uwaterloo.ca


Solutions

1. (a) Pulling out two of each type of bill gives Rad, 2×($5+$10+$20+$50) = 2×($85) = $170.
Since his total sum of money is $175, the only other bill that Rad pulled out must have a
value of $175 − $170 = $5.
That is, Rad pulled out three $5 bills, two $10 bills, two $20 bills, and two $50 bills, for a
total of 3 + 2 + 2 + 2 = 9 bills.
(b) Sandy pulls out at least one of each type of bill and so she must have at least
$5 + $10 + $20 + $50 = $85.
Thus, we know four of the five bills that Sandy pulls out and that these four bills total
$85. The fifth bill that Sandy pulls could be any one of the four different types of bills.
If this fifth bill is a $5 bill, then Sandy’s total sum of money is $85 + $5 = $90.
Finally, if this fifth bill is a $50 bill, then Sandy’s total sum of money is $85 + $50 = $135.
Therefore, the sums of money that Sandy could have are $90, $95, $105, and $135.
(c) Lino could have at most three $50 bills since four $50 bills exceeds his total sum of money
(4 × $50 = $200 > $160).
If Lino had no $50 bills, then the bills (6 bills at most) each have value at most $20 and
would total $160.
However, this is not possible since 6 × $20 = $120 which is less than the required $160.
If Lino had one $50 bill, then the remaining bills (5 bills at most) would total
$160 − $50 = $110.
However, this is not possible since the largest denomination of the remaining bills is $20
and 5 × $20 = $100 which is less than the required $110.
Therefore, we proceed by considering the cases where Lino has two or three $50 bills.
These two cases are summarized in the table below.
Number Money Money Number Number Number Number of
of $50s in $50s remaining of $20s of $10s of $5s bills used
3 $150 $10 1 4
3 $150 $10 2 5
2 $100 $60 3 5
2 $100 $60 2 2 6
In each case from the table above, Lino has a total sum of $160 and has pulled out 6 or
fewer bills.
Since we are given that there are only four possibilities, then we have found them all.
2. (a) The equation of a parabola, written in the form y = (x − p)2 + q, has its vertex at (p, q).
Thus, the parabola with equation y = (x − 3)2 + 1, has its vertex at (3, 1).
(b) Solution 1
Under a translation, the shape of a parabola remains unchanged.
That is, the new parabola is congruent to the original parabola.
After a translation of 3 units to the left and 3 units up, the original vertex (3, 1) moves to
the point (3 − 3, 1 + 3) or (0, 4).
Since this new parabola is congruent to the original, but it has its vertex at (0, 4), then
its equation is y = (x − 0)2 + 4 or y = x2 + 4.
Solution 2
Under a translation of 3 units left and 3 units up, the equation y = (x − 3)2 + 1 becomes
y − 3 = ((x + 3) − 3)2 + 1
y − 3 = x2 + 1
y = x2 + 4
(c) At the point of intersection of these two parabolas, their y values must be equal. Thus,
(x − 3)2 + 1 = x2 + 4
x2 − 6x + 9 + 1 = x2 + 4
−6x = −6
x = 1
Substituting x = 1 into the equation y = x2 + 4, we determine the y value of the point of

intersection to be y = 5.
Therefore, the two parabolas intersect at the point (1, 5).
(d) At the point of intersection of these two parabolas, their y values must be equal. Thus,
(x − 3)2 + 1 = ax2 + 4
x2 − 6x + 9 + 1 = ax2 + 4
0 = ax2 − x2 + 6x − 6
0 = (a − 1)x2 + 6x − 6
Since the two parabolas intersect at exactly one point, then the resulting equation
(a − 1)x2 + 6x − 6 = 0 (which is quadratic since a < 0), has exactly one solution.
Thus, the discriminant of this equation must equal zero.
(Note: The discriminant of a quadratic equation of the form ax2 + bx + c = 0, is b2 − 4ac.)
Solving 62 − 4(−6)(a − 1) = 0, we get 36 + 24(a − 1) = 0 or 24a = −12, and so a = − 12 .
That is, the parabolas with equations y = ax2 + 4 and y = (x − 3)2 + 1 touch at exactly
one point when a = − 21 .
3. (a) If the sequence begins with 3 or more P’s, then it cannot be non-predictive since there are
only 2 Q’s.
Thus, there are two possible cases to consider; the sequence begins with exactly 1 P (that
is, the sequence begins PQ since the second letter is not P), or the sequence begins with
exactly 2 P’s (that is, the sequence begins PPQ since the third letter is not P).
Case 1: The sequence begins PQ
Any sequence that begins PQ is already non-predictive, and so the remaining 7 letters can
be in any order.
Therefore the number of non-predictive sequences beginning PQ, with m = 7 and n = 2,
is equal to the number of ways of arranging 6 P’s and 1 Q in the remaining 7 positions.
There are 7 distinct sequences having 6 P’s and 1 Q since the Q can be placed in any
one of the 7 remaining positions (while the P’s are placed in all other positions without
choice).
Thus the number of non-predictive sequences in this case is 7.
Case 2: The sequence begins PPQ
If the fourth letter in this sequence was P, then it would not be possible for the sequence
to be non-predictive since there are only 2 Q’s.
Therefore, the sequence must begin PPQQ.
Since the remaining 5 letters are all P’s, then there is no choice in forming the rest of the
sequence, so there is only one such sequence.
If m = 7 and n = 2, then the total number of non-predictive sequences beginning with P
is 7 + 1 = 8.
(b) Using a similar argument to that in part (a), if m > 2 and n = 2, then the non-predictive
sequences beginning with P must either begin PQ or PPQQ.
In the case in which the sequence begins PQ, there are (m − 1) P’s and 1 Q left to arrange
in the remaining m positions.
Since the Q can be placed in any one of the m remaining positions, and each placement
produces a distinct sequence, there are m non-predictive sequences that begin PQ.
In the case where the sequence begins PPQQ, there are only P’s left to arrange.
There is only 1 way to do this.
Thus, there is only 1 non-predictive sequence that begins PPQQ.
Therefore the number of non-predictive sequences with m > 2 and n = 2, beginning with
a P, is (m + 1).
Any sequence that begins with a Q is a non-predictive sequence.
Therefore, the total number of non-predictive sequences that begin with Q is equal to
the number of ways of arranging the remaining m P’s and 1 Q in the remaining (m + 1)
positions.
Since the Q can be placed in any one of the (m+1) positions, and each placement produces
a distinct sequence, there are (m + 1) non-predictive sequences that begin with a Q.
Therefore when n = 2, for every m > 2 the number of non-predictive sequences that begin
with P is equal to the number of non-predictive sequences that begin with Q.
(c) If the sequence begins with 4 or more P’s, then it cannot be non-predictive since there are
only 3 Q’s. (The number of Q’s can never “catch up” to the number of P’s.)
Consider the cases in which the sequence begins with 0, 1, 2, or 3 P’s.
Case 1: The sequence begins with a Q (it begins with 0 P’s)
Any sequence that begins with a Q is a non-predictive sequence.
Therefore, the total number of non-predictive sequences that begin with Q is equal to the
number of ways of arranging the remaining 10 P’s and 2 Q’s in 12 positions.
We count these arrangements by considering the number of distinct ways to place the 2
Q’s in the 12 positions and filling in the remaining positions with the 10 P’s.
There are 12 possible positions to place the first Q, followed by 11 possible positions to
place the second, or 12 × 11 arrangements.
However, since the 2 Q’s cannot be distinguished from one another (i.e. they are identical),
this counts each of the possible arrangements twice.
(To see this, consider for example that placing the first Q in the 4th position and the
second Q in the 7th position is the same sequence as placing the first Q in the 7th position
and the second Q in the 4th position.)
Therefore in this case, the number of ways to place the two Q’s is 12×112
= 66 and so the
number of non-predictive sequences is 66.
Case 2: The sequence begins PQ

Any sequence that begins with PQ is a non-predictive sequence.
Therefore, the total number of non-predictive sequences that begin with PQ is equal to
the number of ways of arranging the remaining 9 P’s and 2 Q’s in 11 positions.
We count these arrangements by considering the number of distinct ways to place the 2
Q’s in the 11 possible positions and filling in the remaining positions with the 9 P’s.
Using the same argument as was used in Case 1, this can done in 11×10 2
= 55 possible
ways.
Case 3: The sequence begins PPQ
There are two possibilities for the 4th letter in the sequence beginning PPQ.
The sequence beginning PPQQ is non-predictive, however, so is the sequence that begins
PPQPQQ.
(Can you verify for yourself that these are the only two ways to begin a non-predictive
sequence with m = 10, n = 2 starting with PPQ?)
The total number of non-predictive sequences that begin PPQQ is equal to the number of
ways of arranging the remaining 8 P’s and 1 Q in 9 positions.
Since the Q can be placed in any one of the 9 positions, and each placement produces a
distinct sequence, there are 9 non-predictive sequences that begin PPQQ.
There is only 1 non-predictive sequence that begins PPQPQQ since the remaining 7 letters
are all P’s.
Therefore in this case, the number of non-predictive sequences is 9 + 1 = 10.
Case 4: The sequence begins PPPQ
Since there are only 2 more Q’s available, this sequence must begin PPPQQQ so that it
is non-predictive.
There is only 1 non-predictive sequence that begins PPPQQQ since the remaining 7 letters
are all P’s.
Thus in this final case, there is only 1 non-predictive sequence.
The number of non-predictive sequences with m = 10 and n = 2 is 66 + 55 + 10 + 1 = 132.
4. (a) Since the edge length of each cube is 1 cm, then the length, width and height of the
rectangular prism are 5 cm, 4 cm and 1 cm, respectively.
The top face of the prism has dimensions 5 cm by 4 cm, and so has area 20 cm2 .
Similarly, the bottom face on the opposite side of the prism has the same area, 20 cm2 .
The front and back faces of the prism each have dimensions 4 cm by 1 cm, and so have
area 4 cm2 .
The right and left faces of the prism each have dimensions 5 cm by 1 cm, and so have
area 5 cm2 .
Therefore, the surface area of the rectangular prism is 2 × (20 + 4 + 5) = 2 × 29 = 58 cm2 .
(b) Suppose that the rectangular prism in question is l cm (cubes) in length and w cm (cubes)
wide, with l ≥ w.
Then the top surface of the rectangular prism has area (l × w) cm2 , the front surface has
area (w × 1) cm2 , and the right side has area (l × 1) cm2 .
Therefore, the surface area of the entire rectangular prism is 2 × (lw + w + l) cm2 .
Since the surface area is 180 cm2 , then 2 × (lw + w + l) = 180 or lw + w + l = 90.
Adding 1 to both sides of this equation, we get lw+w+l+1 = 91, so w(l+1)+1(l+1) = 91.
Factoring the left side of this equation one step further, we have (w + 1)(l + 1) = 91.
Since both l and w are positive integers, then (w + 1)(l + 1) is the product of two positive
integers.
The right side, 91, can be written as the product of two positive integers in exactly two
different ways, 1 × 91 and 7 × 13.
Since both l and w are positive integers, then 2 ≤ (w + 1) ≤ (l + 1) and so neither factor
can be equal to 1.
Therefore (w + 1) = 7 and (l + 1) = 13 (since l ≥ w), and so w = 6 and l = 12.
Since the rectangular prism is 6 cubes wide and 12 cubes in length, then it has 6 × 12 = 72
cubes in total.
(c) As shown in part (b), the surface area of the original prism without the removal of the
internal prism, is 2 × (lw + w + l) cm2 .
To find the surface area of the frame, we must account for the area that is lost and that
which is gained by removing the internal prism.
The area of the original prism that is lost is equal to the front and back rectangular faces
of the internal prism.
Since the width of the original prism is w cm and the internal prism is located k cm from
each side of the original prism, then the width of the internal prism is (w − 2k) cm.
Similarly, the length of the internal prism is (l − 2k) cm.
(Note that w − 2k > 0 and l − 2k > 0, so w > 2k and l > 2k.)
Thus, the area from the original prism that is lost by removing the internal prism is
2 × (w − 2k) × (l − 2k) cm2 .
The area that is gained by removing the internal prism is equal to the top, bottom, left,
and right rectangular surfaces of the internal prism.
Since the width of the internal prism is (w − 2k) cm and its thickness is 1 cm, then the
total area of the top and bottom faces is 2 × (w − 2k) × 1 cm2 .
Similarly, since the length of the internal prism is (l − 2k) cm and its thickness is 1 cm,
then the total area of the right and left faces is 2 × (l − 2k) × 1 cm2 .
Summarizing, the surface area of the original prism without the removal of the internal
prism, is 2 × (lw + w + l) cm2 .
The area of the original prism that is lost by removing the internal prism is,
2 × (w − 2k) × (l − 2k) cm2 .
The area that is added to the area of the original prism by removing the internal prism is,
(2 × (w − 2k) × 1 + 2 × (l − 2k) × 1) cm2 or 2 × (w − 2k + l − 2k) = 2 × (w + l − 4k) cm2 .
That is, the surface area of the frame, in cm2 , is
2 × (lw + w + l) − 2 × (w − 2k) × (l − 2k) + 2 × (w + l − 4k).
Since the surface area of the frame is 532 cm2 , then we equate this with the expression for
the surface area and simplify the resulting equation.
2 × (lw + w + l) − 2 × (w − 2k) × (l − 2k) + 2 × (w + l − 4k) = 532

532
lw + w + l − (w − 2k) × (l − 2k) + (w + l − 4k) = 2
lw + w + l − (lw − 2kw − 2kl + 4k 2 ) + (w + l − 4k) = 266
2w + 2l + 2kw + 2kl − 4k 2 − 4k = 266
w + l + kw + kl − 2k 2 − 2k = 266
2
kw + kl − 2k 2 + w + l − 2k = 133
k(w + l − 2k) + 1(w + l − 2k) = 133
(w + l − 2k)(k + 1) = 133
Recall that w > 2k and l > 2k, so then (w + l − 2k) is a positive integer since w, l and k
are positive integers.
Therefore, (w + l − 2k)(k + 1) is the product of two positive integers.
Expressed as the product of two positive integers, 133 can only be written as 1 × 133 or
7 × 19.
Note that (k + 1) > 1 , and (w + l − 2k) > 2k + 2k − 2k = 2k > 1, since k is a positive
integer.
That is, (k + 1) 6= 1 and (w + l − 2k) 6= 1, and so either k + 1 = 7 or k + 1 = 19.
If k + 1 = 7, then k = 6 and w + l − 2k = 19 or w + l − 12 = 19, and so w + l = 31.
Since w > 2k = 12, we require all possible values for w and l such that w ≥ 13, l ≥ w,
and w + l = 31.
Written as ordered pairs (w, l) the only possibilities are (13, 18), (14, 17) and (15, 16).
If k + 1 = 19, then k = 18 and w + l − 2k = 7 or w + l − 36 = 19, and so w + l = 55.
Since w > 2k = 36, then l < 55 − 36 = 29.
This is not possible since l ≥ w.
Therefore, the only possible values for w and l such that the frame has surface area 532 cm2
are 13 and 18, or 14 and 17, or 15 and 16.
www.cemc.uwaterloo.ca


Solutions

1. (a) In 4P T Q, ∠P T Q = 90◦ .
2 2 2 2
Using the
√ Pythagorean Theorem, P Q = 32 + 24 or P Q = 1024 + 576 = 1600 and so
P Q = 1600 = 40, since P Q > 0.
(b) In 4QT R, ∠QT R = 90◦ .
the Pythagorean Theorem, 512 = T R2 + 242 or T R2 = 2601 − 576 = 2025 and so
Using √
T R = 2025 = 45, since T R > 0.
Since P T = 32 and T R = 45, then P R = P T + T R = 32 + 45 = 77.
In 4P QR, QT is perpendicular to base P R and so 4P QR has area
1
2
× (P R) × (QT ) = 21 × 77 × 24 = 924.
(c) From part (b), the length of P R is 77.
QS 12 12
Since QS : P R = 12 : 11, then = or QS = 77 × = 84.
77 11 11
So then T S = QS − QT = 84 − 24 = 60. √ √
Using the Pythagorean Theorem in 4P T S, P S = 322 + 602 or P S = 4624 = 68, since
P S > 0. √ √
Using the Pythagorean Theorem in 4RT S, RS = 452 + 602 or RS = 5625 = 75, since
RS > 0.
Thus, quadrilateral P QRS has perimeter 40 + 51 + 75 + 68 or 234.
2. (a) Expanding, (a + b)2 = a2 + 2ab + b2 = (a2 + b2 ) + 2ab.
Since a2 + b2 = 24 and ab = 6, then (a + b)2 = 24 + 2(6) = 36.
(b) Expanding, (x + y)2 = x2 + 2xy + y 2 = (x2 + y 2 ) + 2xy.
Since (x + y)2 = 13 and x2 + y 2 = 7, then 13 = 7 + 2xy or 2xy = 6, and so xy = 3.
(c) Expanding, (j + k)2 = j 2 + 2jk + k 2 = (j 2 + k 2 ) + 2jk.
Since j + k = 6 and j 2 + k 2 = 52, then 62 = 52 + 2jk or 2jk = −16, and so jk = −8.
(d) Expanding, (m2 + n2 )2 = m4 + 2m2 n2 + n4 = (m4 + n4 ) + 2m2 n2 .
Since m2 +n2 = 12 and m4 +n4 = 136, then 122 = 136+2m2 n2 or 2m2 n2 = 8 or m2 n2 = 4,
and so mn = ±2.
3. (a) Since ∠M ON = 90◦ , the product of the slopes of N O and OM is −1.
n2 − 0
The slope of N O is = n, since n 6= 0 (points N and O are distinct).
n−0
1
−0 1
The slope of OM is 41 = .
2
−0 2
1
Thus, n × = −1 or n = −2.
2
(b) Since ∠ABO = 90◦ , the product of the slopes of BA and BO is −1.
b2 − 4 (b − 2)(b + 2)
The slope of BA is = = b + 2, since b 6= 2 (A and B are distinct).
b−2 b−2
b2 − 0
The slope of BO is = b, since b 6= 0 (B and O are distinct).
b−0
Thus, (b + 2) × b = −1 or b2 + 2b + 1 = 0.
Factoring, (b + 1)(b + 1) = 0 and so b = −1.
(c) Since ∠P QR = 90◦ , the product of the slopes of P Q and RQ is −1.
p2 − q 2 (p − q)(p + q)
The slope of P Q is = = p + q, since p 6= q (P and Q are distinct).
p−q p−q
r2 − q 2 (r − q)(r + q)
The slope of RQ is = = r + q, since r 6= q (R and Q are distinct).
r−q r−q
Thus, (p + q) × (r + q) = −1.
Since p, q and r are integers, then p + q and r + q are integers.
In order that (p + q) × (r + q) = −1, either p + q = 1 and r + q = −1 or p + q = −1 and
r+q = 1 (these are the only possibilities for integers p, q, r for which (p+q)×(r+q) = −1).
In the first case, we add the two equations to get p + q + r + q = 1 + (−1) or 2q + p + r = 0.
In the second case, adding the two equations gives p + q + r + q = −1 + 1 or 2q + p + r = 0.
In either case, 2q + p + r = 0, as required.
4. (a) Since p is an odd prime integer, then p > 2.
Since the only prime divisors of 2p2 are 2 and p, then the positive divisors of 2p2 are
1, 2, p, 2p, p2 , and 2p2 .
So then, S(2p2 ) = 1 + 2 + p + 2p + p2 + 2p2 = 3p2 + 3p + 3.
Since S(2p2 ) = 2613, then 3p2 + 3p + 3 = 2613 or 3p2 + 3p − 2610 = 0 or p2 + p − 870 = 0.
Factoring, (p + 30)(p − 29) = 0, and so p = 29 (p 6= −30 since p is an odd prime).
(b) Suppose m = 2p and n = 9q for some prime numbers p, q > 3.
The positive divisors of 2p, thus m, are 1, 2, p, and 2p (since p > 3).
Therefore, S(m) = 1 + 2 + p + 2p = 3p + 3.
The positive divisors of 9q, thus n, are 1, 3, q, 3q, 9, and 9q (since q > 3).
Therefore, S(n) = 1 + 3 + 9 + q + 3q + 9q = 13q + 13.
Since S(m) = S(n), then 3p + 3 = 13q + 13 or 3p − 13q = 10.
Also, m and n are consecutive integers and so either m − n = 1 or n − m = 1.
If m − n = 1, then 2p − 9q = 1.
We solve the following system of two equations and two unknowns.
2p − 9q = 1 (1)
3p − 13q = 10 (2)
Multiplying equation (1) by 3 and equation (2) by 2 we get,
6p − 27q = 3 (3)
6p − 26q = 20 (4)
Subtracting equation (3) from equation (4), we get q = 17.
Substituting q = 17 into equation (1), 2p − 9(17) = 1 or 2p = 154, and so p = 77.
However, p must be a prime and thus p 6= 77.
There is no solution when m − n = 1.
If n − m = 1, then 9q − 2p = 1.
We solve the following system of two equations and two unknowns.
9q − 2p = 1 (5)
3p − 13q = 10 (6)
Multiplying equation (5) by 3 and equation (6) by 2 we get,
27q − 6p = 3 (7)
6p − 26q = 20 (8)
Adding equation (7) and equation (8), we get q = 23.
Substituting q = 23 into equation (6), 3p − 13(23) = 10 or 3p = 309, and so p = 103.
Since q = 23 and p = 103 are prime integers greater than 3, then m = 2(103) = 206 and
n = 9(23) = 207 are the only pair of consecutive integers satisfying the given properties.
(c) Since the only prime divisors of p3 q are p and q, then the positive divisors of p3 q, are
1, p, q, pq, p2 , p2 q, p3 , and p3 q (since p and q are distinct primes).
Therefore, S(p3 q) = p3 q + p3 + p2 q + p2 + pq + p + q + 1.
Simplifying,
S(p3 q) = p3 q + p3 + p2 q + p2 + pq + p + q + 1
= (p3 q + p2 q + pq + q) + (p3 + p2 + p + 1)
= q(p3 + p2 + p + 1) + (p3 + p2 + p + 1)
= (q + 1)(p3 + p2 + p + 1)
= (q + 1)(p2 (p + 1) + (p + 1))
= (q + 1)(p + 1)(p2 + 1)
We are to determine the number of pairs of distinct primes p and q, each less than 30,
such that (q + 1)(p + 1)(p2 + 1) is not divisible by 24.
There are 10 primes less than 30. These are 2, 3, 5, 7, 11, 13, 17, 19, 23 and 29.
Therefore, the total number of possible pairs (p, q), where p 6= q, is 10 × 9 = 90.
We will count the number of pairs (p, q) for which (q + 1)(p + 1)(p2 + 1) is divisible by 24
and then subtract this total from 90.
If p or q equals 23, then 24 divides (q + 1)(p + 1)(p2 + 1).
There are 9 ordered pairs of the form (23, q) and 9 of the form (p, 23).
Thus, we count 18 pairs and since we have exhausted all possibilities using 23, we remove
it from our list of 10 primes above.
Since 24 = 23 × 3, we can determine values of q for a given value of p by recognizing that
each of these prime factors (three 2s and one 3) must occur in the prime factorization of
(q + 1)(p + 1)(p2 + 1).
For example if p = 2, then (q + 1)(p + 1)(p2 + 1) = (q + 1)(3)(5).
Therefore, for (q + 1)(p + 1)(p2 + 1) to be a multiple of 24, q + 1 must be a multiple of 8
(since we are missing 23 ).
Thus when p = 2, the only possible value of q is 7 (we get this by trying the other 8 values
in the list of primes).
We organize all possibilities for p (and the resulting values of q) in the table below.
p (p + 1)(p2 + 1) q + 1 must be q Number of

a multiple of (distinct from p) ordered pairs
2 (3)(5) 23 = 8 q=7 1
3 (4)(10) = 23 × 5 3 q = 2, 5, 11, 17, 29 5
5 (6)(26) = 22 × 3 × 13 2 q = 3, 7, 11, 13, 17, 19, 29 7
7 (8)(50) = 23 × 50 3 q = 2, 5, 11, 17, 29 5
11 (12)(122) = 23 × 3 × 61 any q will work q = 2, 3, 5, 7, 13, 17, 19, 29 8
13 (14)(170) = 22 × 595 2×3=6 q = 5, 11, 17, 29 4
17 (18)(290) = 22 × 3 × 435 2 q = 3, 5, 7, 11, 13, 19, 29 7
19 (20)(362) = 23 × 905 3 q = 2, 5, 11, 17, 29 5
29 (30)(842) = 22 × 3 × 2105 2 q = 3, 5, 7, 11, 13, 17, 19 7
The total number of pairs (p, q) for which 24 divides S(p3 q) is
18 + 1 + 5 + 7 + 5 + 8 + 4 + 7 + 5 + 7 = 67.
Thus, the total number of pairs of distinct prime integers p and q, each less than 30, such
that S(p3 q) is not divisible by 24, is 90 − 67 = 23.
The CENTRE for EDUCATION in
MATHEMATICS and COMPUTING

Solutions
©2011 Centre for Education in Mathematics and Computing

1. (a) Since D is the midpoint of AB, it has coordinates ( 12 (0 + 0), 12 (0 + 6)) = (0, 3).
3−0
The line passing through C and D has slope 0−8 or − 83 .
The y-intercept of this line is the y-coordinate of point D, or 3.
Therefore, the equation of the line passing through points C and D is y = − 83 x + 3.
(b) Since E is the midpoint of BC, it has coordinates ( 12 (8 + 0), 21 (0 + 0)) = (4, 0).
Next, we find the equation of the line passing through the points A and E.
This line has slope 6−0
0−4
or − 46 or − 23 .
The y-intercept of this line is the y-coordinate of point A, or 6.
Therefore, the equation of the line passing through points A and E is y = − 32 x + 6.
Point F is the intersection point of the lines with equation y = − 38 x + 3 and y = − 23 x + 6.
To find the coordinates of point F we solve the system of equations by equating y:
− 38 x + 3 = − 23 x + 6
8(− 83 x + 3) = 8(− 32 x + 6)
−3x + 24 = −12x + 48
9x = 24
So the x-coordinate of point F is 24 9
or x = 38 .
Substituting x = 83 into y = − 32 x + 6, we find that y = − 23 × 83 + 6 or y = 2.
The coordinates of point F are ( 83 , 2).
(c) Triangle DBC has base BC of length 8, and height BD of length 3.
Therefore, the area of 4DBC is 12 × 8 × 3 or 12.
(d) The area of quadrilateral DBEF is the area of 4DBC minus the area of 4F EC.
Triangle F EC has base EC.
Note that EC = BC − BE = 8 − 4 or 4.
The height of 4F EC is equal to the vertical distance from point F to the x-axis.
That is, the height of 4F EC is equal to the y-coordinate of point F , or 2.
Therefore, the area of 4F EC is 12 × 4 × 2 or 4.
Thus, the area of quadrilateral DBEF is 12 − 4 or 8.
2. (a) Since the tens digit to be used is 3, we must consider the number of possibilities for the
ones digit.
Given that the digits 0 and 9 cannot be used, there are 8 choices remaining for the ones
digit.
However if the ones digit is a 3, then the number formed, 33, is a multiple of 11 which is
not permitted.
Since 33 is the only multiple of 11 with tens digit 3, each of the other 7 choices for the
ones digit is possible.
There are 7 numbers in S whose tens digit is a 3.
In fact, the argument above can be repeated to show that there are 7 numbers in S for
each of the possible tens digits 1 through to 8.
This will be useful information for part (d) to follow.
(b) Since the ones digit to be used is 8, we must consider the number of possibilities for the
tens digit.
Given that the digits 0 and 9 cannot be used, there are 8 choices remaining for the tens
digit.
However if the tens digit is 8, then the number formed, 88, is a multiple of 11 which is not
permitted.
Since 88 is the only multiple of 11 with ones digit 8, each of the other 7 choices for the
tens digit is possible.
There are 7 numbers in S whose ones digit is 8.
In fact, the argument above can be repeated to show that there are 7 numbers in S for
each of the possible ones digits 1 through to 8.
This will also be useful information for part (d) to follow.
(c) Solution 1
Ignoring the second restriction that no number in S be a multiple of 11, there are 8 choices
for the ones digit and 8 choices for the tens digit.
For each of the 8 choices for the tens digit, there are 8 choices for the units digit ignoring
multiples of 11.
Thus, there would be 8 × 8 or 64 numbers in S, ignoring the second restriction.
Included in these 64 numbers are the numbers 11, 22, 33, 44, 55, 66, 77, 88, or 8 multiples
of 11.
These are the only multiples of 11 in our 64 possibilities.
Removing these from the number of possibilities, there are 64 − 8 or 56 numbers in S.
Solution 2
Given that the digits 0 and 9 cannot be used, there are 8 choices for the tens digit.
For each choice of a tens digit, choosing the ones digit to be equal to that tens digit gives
the only number that is a multiple of 11.
That is, for each possible choice of a tens digit, the ones digit cannot equal the tens digit.
Since the ones digit cannot equal 0, 9 or the tens digit, there are 7 possible choices of a
ones digit for each choice of a tens digit.
Thus, there are 8 × 7 = 56 numbers in S.
(d) Our work in part (a) shows that for each of the possible tens digits, 1 through 8, there are
7 numbers in S that have that tens digit.
That is, of the 56 numbers in S there are 7 whose tens digit is 1, 7 whose tens digit is 2,
and so on to include 7 whose tens digit is 8.
Similarly, our work in part (b) shows that for each of the possible ones digits, 1 through
8, there are 7 numbers in S that have that ones digit.
That is, of the 56 numbers in S there are 7 whose ones digit is 1, 7 whose ones digit is 2,
and so on to include 7 whose ones digit is 8.
We may determine the sum of the 56 numbers in S by considering the sum of their tens
digits separately from the sum of their ones digits.
First, consider the sum of the ones digits of all of the numbers in S.
Each of the numbers 1 through 8 appear 7 times in the ones digit.
The sum of the numbers from 1 to 8 is 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 = (8)(9)2
= 36.
Since each of these occur 7 times, then the sum of the ones digits for all numbers in S is
7 × 36 or 252.
Next, consider the sum of the tens digits of all of the numbers in S.
Again, each of the numbers 1 through 8 appear 7 times in the tens digit.
The sum of the numbers from 1 to 8 is 36.
Since each of these occur 7 times, the sum of the tens digits for all numbers in S is 7 × 36
or 252.
Since these are tens digits, they add 10 × 252 or 2520 to the total sum.
Thus, the sum of all of the numbers in S is the combined sum of all 56 ones digits, 252,
and all 56 tens digits, which add 2520 to the sum, for a total of 2772.
3. (a) Solution 1
Since 3x = 5y, then y = 53 x.
In the given Trenti-triple, the value of x is 50.
Thus, y = 35 (50) or y = 30.
Since 3x = 2z, then z = 23 x.
Since x = 50, then z = 32 (50) or z = 75.
The Trenti-triple is (50, 30, 75).
Solution 2
Let 3x = 5y = 2z = k.
Since x, y, z are positive integers, then k is a positive integer that is divisible by 3, 5 and 2.
Any positive integer that is divisible by 3, 5 and 2 must be divisible by their least common
multiple, which is 3 × 5 × 2 or 30.
Since k is divisible by 30, then k = 30m for some positive integer m.
That is, 3x = 5y = 2z = 30m and so x = 10m, y = 6m and z = 15m.
Since x = 50, then 50 = 10m or m = 5.
Therefore, y = 6 × 5 or y = 30 and z = 15 × 5 or z = 75.
The Trenti-triple is (50, 30, 75).
(b) Solution 1
Since 3x = 5y, then x = 35 y.
Since x is a positive integer, then 35 y is a positive integer and so y is divisible by 3, since
5 is not.
Since 5y = 2z, then z = 52 y.
Since z is a positive integer, then 52 y is a positive integer and so y is divisible by 2, since
5 is not.
Thus, y is divisible by both 2 and 3 and so y is divisible by the least common multiple of
2 and 3.
Therefore, in every Trenti-triple, y is divisible by 6.
Solution 2
From our work in part (a) Solution 2, it follows that since y = 6m for some positive integer
m, then y is divisible by 6 for every Trenti-triple.
(c) Solution 1
From part (b) Solution 1, we know that y is divisible by 6 for every Trenti-triple.
We can similarly show that x is divisible by 10 for every Trenti-triple.
Since 3x = 5y, then y = 35 x.
Since y is a positive integer, then 35 x is a positive integer and so x is divisible by 5, since
3 is not.
Since 3x = 2z, then z = 32 x.
Since z is a positive integer, then 32 x is a positive integer and so x is divisible by 2, since
3 is not.
Thus, x is divisible by both 5 and 2 and so x is divisible by the least common multiple of
5 and 2.
Therefore, in every Trenti-triple, x is divisible by 10.
We can similarly show that z is divisible by 15 for every Trenti-triple.
Since 3x = 2z, then x = 32 z.
Since x is a positive integer, then 23 z is a positive integer and so z is divisible by 3, since
2 is not.
Since 5y = 2z, then y = 25 z.
Since y is a positive integer, then 25 z is a positive integer and so z is divisible by 5, since

2 is not.
Thus, z is divisible by both 3 and 5 and so z is divisible by the least common multiple of
3 and 5.
Therefore, in every Trenti-triple, z is divisible by 15.
Since in every Trenti-triple, y is divisible by 6, x is divisible by 10, and z is divisible by
15, then their product xyz is divisible by 6 × 10 × 15 or 900.
Solution 2
From our work in part (a) Solution 2, we have that x = 10m, y = 6m and z = 15m for
some positive integer m.
Therefore, the product xyz is (10m)(6m)(15m) or 900m3 , and thus is divisible by 900 for
every Trenti-triple.
4. (a) F (8) = 6 since
8 = 1+1+1+1+1+1+1+1
= 1+1+1+1+1+3
= 1+1+3+3
= 1+1+1+5
= 3+5
= 1+7
(b) Let us first begin by defining each of the ways that a positive integer n can be written as
the sum of positive odd integers as a representation of n.
There are F (n) representations of n.
To each possible representation of n, we may add a 1 to create a representation of n + 1.
For example, 3 + 3 + 1 is a representation of 7; adding a 1, 3 + 3 + 1 + 1, creates a repre-
sentation of 8.
Since every representation of n can be used to create a representation of n + 1 in this way,
then there are at least as many representations of n+1 as there are of n, so F (n+1) ≥ F (n).
Next, we will show that in fact F (n + 1) ≥ F (n) + 1 by finding one additional represen-
tation of n + 1 not described above.
We will do this by considering the cases when n + 1 is odd and when n + 1 is even.
Case 1: n + 1 is odd
Since n + 1 is odd, then n + 1 is a representation of itself.
Since this representation of n + 1 does not include a 1 (n > 3), then it must be a new
representation not created by adding a 1 to a representation of n as described above.
Therefore, if n + 1 is odd then F (n + 1) ≥ F (n) + 1 and so F (n + 1) > F (n).
Case 2: n + 1 is even
Since n + 1 is even, then n is odd.
Since n is odd and n > 3, then n ≥ 5.
Since n is odd and n ≥ 5, then n − 2 is odd and n − 2 ≥ 3.
Since n + 1 = (n − 2) + 3 and n − 2 ≥ 3, then (n − 2) + 3 is a representation of n + 1 that
does not include a 1.
Since this representation of n+1 does not include a 1, then it must be a new representation
not created by adding a 1 to a representation of n as described above.
Therefore, if n + 1 is even then F (n + 1) ≥ F (n) + 1 and so F (n + 1) > F (n).
Thus for all integers n > 3, F (n + 1) > F (n).
(c) Let an be the representation of n as the sum of n 1s.

As an example from part (a), a8 = 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1.
Let bn be the representation of n as (n − 1) + 1 if n is even and as (n − 2) + 1 + 1 if n is
odd.
Therefore, b8 = 7 + 1 since 8 is even.
Let Sn be the list of the remaining representations of n.
From part (a), list S8 consists of the following representations:
8 = 1+1+1+1+1+3
= 1+1+3+3
= 1+1+1+5
= 3+5
Since each of an and bn are single representations of n, there are F (n) − 2 representations
in Sn .
Note that when n = 4, Sn has no representations for n.
Consider the representations an + Sn of 2n.
These representations of 2n are n 1s added to each of the representations of Sn .
Again using our work from part (a) as an example, the representations of 16 given by
a8 + S8 are:
a8 + S8 = (1 + 1 + 1 + 1 + 1 + 1 + 1 + 1) + (1 + 1 + 1 + 1 + 1 + 3)
= (1 + 1 + 1 + 1 + 1 + 1 + 1 + 1) + (1 + 1 + 3 + 3)
= (1 + 1 + 1 + 1 + 1 + 1 + 1 + 1) + (1 + 1 + 1 + 5)
= (1 + 1 + 1 + 1 + 1 + 1 + 1 + 1) + (3 + 5)
In general, consider the following representations of 2n:
• an + Sn (there are F (n) − 2 representations here and when n = 4 there are none)
• bn + Sn (there are F (n) − 2 representations here and when n = 4 there are none)
• an + an
• an + b n
• bn + bn
• (2n − 1) + 1
• (2n − 3) + 3
There are 2 × [F (n) − 2] + 5 = 2F (n) + 1 representations in this list.
If these are all distinct, then F (2n) ≥ 2F (n) + 1 > 2F (n), as required.
Since n > 3, then n − 3 > 0 or 2n − 3 > n and thus 2n − 1 > n also.
Since both 2n − 3 and 2n − 1 are greater than n, then there can be no overlap between
the last two lists of representations and the first five lists of representations in the above
list.
There is no overlap between any of the third, fourth or fifth lists of representations by the
definitions of an and bn .
Similarly, there can be no overlap between the first two lists of representations and the
third, fourth and fifth lists of representations by the definitions of an , bn and Sn .
This leaves us to consider the possibility of overlap between the first two lists of represen-
tations only.
Suppose that there is a representation of 2n that is included in both an +Sn and in bn +Sn .
Since this representation is included in an + Sn , then part of it looks like an , so the repre-
sentation includes n 1s.
Since this representation is included in bn + Sn , then part of it looks like bn , so the repre-
sentation includes either n−1 or n−2 depending on whether n is even or odd. Because the
representation already includes some 1s (from the an portion), we cannot automatically
include the 1 or 1 + 1 from bn .
So this representation includes n 1s and either n − 1 or n − 2.
These parts add to either 2n − 1 or 2n − 2.
The only way to complete the representation is then either with a 1 or with a 1 + 1.
But this means that the representation then is n 1s plus (n − 1) + 1 if n is even or is n 1s
plus (n − 2) + 1 + 1 if n is odd.
This means that this representation must be an + bn .
Since neither an or bn is in Sn , then this representation cannot actually be in
an + Sn or in bn + Sn , so there cannot be any overlap between these two collections
of representations.
Therefore, F (2n) ≥ 2F (n) + 1 > 2F (n), as required.
Canadian
Mathematics
Competition
An activity of the Centre for Education
in Mathematics and Computing,
University of Waterloo, Waterloo, Ontario

Solutions
c 2010 Centre for Education in Mathematics and Computing

1. (a) The cost to fly is $0.10 per kilometre plus a $100 booking fee.
To fly 3250 km from A to B, the cost is 3250 × 0.10 + 100 = 325 + 100 = $425.
(b) Since 4ABC is a right-angled triangle, then we may use the Pythagorean Theorem.
Thus, AB 2 = BC 2 + CA2 , and so BC 2 = AB 2 − CA2 = 32502 − 30002 = 1 562 500, and
BC = 1250 km (since BC > 0).
Piravena travels a distance of 3250 + 1250 + 3000 = 7500 km for her complete trip.
(c) To fly from B to C, the cost is 1250 × 0.10 + 100 = $225.
To bus from B to C, the cost is 1250 × 0.15 = $187.50.
Since Piravena chooses the least expensive way to travel, she will bus from B to C.
To fly from C to A, the cost is 3000 × 0.10 + 100 = $400.
To bus from C to A, the cost is 3000 × 0.15 = $450.
Since Piravena chooses the least expensive way to travel, she will fly from C to A.
To check, the total cost of the trip would be $425 + $187.50 + $400 = $1012.50 as required.
2. (a) Substituting x = 6, then f (x) − f (x − 1) = 4x − 9 becomes f (6) − f (5) = 4 × 6 − 9.

Since f (5) = 18, then f (6) − 18 = 24 − 9 or f (6) − 18 = 15 and f (6) = 33.
(b) Substituting x = 5, then f (x) − f (x − 1) = 4x − 9 becomes f (5) − f (4) = 4 × 5 − 9.
Since f (5) = 18, then 18 − f (4) = 20 − 9 or 18 − f (4) = 11 and f (4) = 7.
Substituting x = 4, then f (x) − f (x − 1) = 4x − 9 becomes f (4) − f (3) = 4 × 4 − 9.
Since f (4) = 7, then 7 − f (3) = 16 − 9 or 7 − f (3) = 7 and f (3) = 0.
(c) Since f (5) = 18, then 2(52 ) + 5p + q = 18, or 50 + 5p + q = 18 and so 5p + q = −32.
Since f (3) = 0, then 2(32 ) + 3p + q = 0, or 18 + 3p + q = 0 and so 3p + q = −18.
We solve the system of equations:
5p + q = −32
3p + q = −18
Subtracting the second equation from the first gives 2p = −14 or p = −7.
Substituting p = −7 into the first equation gives 5(−7) + q = −32, or −35 + q = −32 and
q = 3.
Therefore if f (x) = 2x2 + px + q, then p = −7 and q = 3.
3. (a) Since 4ABE is equilateral, then ∠ABE = 60◦ .

Therefore, ∠P BC = ∠ABC − ∠ABE = 90◦ − 60◦ = 30◦ .
Since AB = BC, then 4ABC is a right isosceles triangle and ∠BAC = ∠BCA = 45◦ .
Then, ∠BCP = ∠BCA = 45◦ and
∠BP C = 180◦ − ∠P BC − ∠BCP = 180◦ − 30◦ − 45◦ = 105◦ .
(b) Solution 1
In 4P BQ, ∠P BQ = 30◦ and ∠BQP = 90◦ , thus ∠BP Q = 60◦ . √
Therefore, 4P BQ is a 30◦ - 60◦ - 90◦ triangle
√ with P Q : P B : BQ=1 : 2 : 3.
PQ
Since BQ = √13 , then BQ
x
= √13 and BQ = 3x.
Solution 2
In 4P QC, ∠QCP = 45◦ and ∠P QC = 90◦ , thus ∠CP Q = 45◦ .
Therefore, 4P QC is isosceles and QC = P Q = x.
Since BC = 4, then BQ = BC − QC = 4 − x.
(c) Solution 1
In 4P QC, ∠QCP = 45◦ and ∠P QC = 90◦ , thus ∠CP Q = 45◦ .
Therefore, 4P QC is isosceles and QC =√P Q = x. √ 4
Since BC = 4, then BC = BQ + QC = 3x + x = 4 or x( 3 + 1) = 4 and x = √3+1 .
√ √ √ √
4
Rationalizing the denominator gives x = √3+1 × √3−1
3−1
= 4( 3−1
3−1)
= 4( 23−1) = 2( 3 − 1).
Solution 2
In 4P BQ, ∠P BQ = 30◦ and ∠BQP = 90◦ , thus ∠BP Q = 60◦ . √
Therefore, 4P BQ is a 30◦ -60◦ -90◦ √
triangle with P Q√: P B : BQ=1 : 2 : √ 3.
PQ 1 x 1
Since BQ = 3 , then 4−x = 3 or 3x = 4 − x or 3x + x = 4, and x( 3 + 1) = 4 so
√ √
4
x = √3+1 .
√ √
4( 3−1)
√
4( 3−1)
√
Rationalizing the denominator gives x = √3+14
× √3−1
3−1
= 3−1
= 2
= 2( 3 − 1).
(d) Solution 1
We adopt the notation |4XY Z| to represent the area of
triangle XY Z. A R S B
Then, |4AP E| = |4ABE| − |4ABP |.
Since 4ABE is equilateral, BE = EA = AB = 4 and
the altitude from E to AB bisects side AB at R as shown.
Thus, AR = RB = 2 and by the Pythagorean √ Theorem√
ER2 = BE 2 − RB 2 = 42 − 22 = 12 or ER = 12 = 2 3, x
since ER > 0. P Q
1
Therefore, the area of 4ABE is 2 (AB)(ER),
√ √
or 21 (4)(2 3) = 4 3. E
In 4ABP , construct the altitude from P to S on AB.
Then P S ⊥ AB and QB ⊥ AB, so P S k QB. D C
Also, SB ⊥ QB and P Q ⊥ QB, so SB k P Q.
Thus, SBQP is a rectangle and P S = QB. √ √
From (b) and (c), QB = 4 − x = 4 − 2( 3 − 1) √ = 6 − 2 3. √ √
Therefore, |4ABP | = 21 (AB)(P S) = 12 (4)(6 − 2 3) = 2(6 − 2 3) = 12 − 4 3.
√ √ √ √ √
Then, |4AP E| = 4 3 − (12 − 4 3) = 4 3 − 12 + 4 3 = 8 3 − 12.
Solution 2
We adopt the notation |4XY Z| to represent the area of triangle XY Z.
Then, |4AP E| = |4ABE| − |4ABP |.
However, |4ABP | = |4ABC| − |4BP C|.
Thus, |4AP E| = |4ABE| − (|4ABC| − |4BP C|) = |4ABE| + |4BP C| − |4ABC|.
Since 4ABE is equilateral, BE = EA = AB = 4 and the altitude from E to AB bisects
side AB at R as shown. √ √ √
Thus, 4ERB is a 30◦ -60◦ -90◦ triangle with ER
√ : RB√= 3 : 1 or ER = (RB) 3 = 2 3.
Therefore, |4ABE| = 12 (AB)(ER) = 12 (4)(2 3) = 4 3.
√ √
Since P Q is an altitude of 4BP C, |4BP C| = 21 (BC)(P Q) = 21 (4)(2 3 − 2) = 4 3 − 4.
In triangle ABC, ∠ABC = 90◦ .
Thus, |4ABC| = 12 (AB)(BC) = 12 (4)(4) = 8.
√ √ √
Thus, |4AP E| = |4ABE| + |4BP C| − |4ABC| = (4 3) + (4 3 − 4) − 8 = 8 3 − 12.
4. (a) We solve by factoring,

x4 − 6x2 + 8 = 0
(x2 − 4)(x2 − 2) = 0
√
Therefore, x2 = 4 or x2 = 2, and so x = ±2 or x = ± 2.
√ √
The real values of x satisfying x4 − 6x2 + 8 = 0 are x = −2, 2, − 2, and 2.
(b) We want the smallest positive integer N for which,
x4 + 2010x2 + N = (x2 + rx + s)(x2 + tx + u)

x4 + 2010x2 + N = x4 + tx3 + ux2 + rx3 + rtx2 + rux + sx2 + stx + su
x4 + 2010x2 + N = x4 + tx3 + rx3 + ux2 + rtx2 + sx2 + rux + stx + su
x4 + 2010x2 + N = x4 + (t + r)x3 + (u + rt + s)x2 + (ru + st)x + su
Equating the coefficients from the left and right sides of this equation we have,
t + r = 0, u + rt + s = 2010, ru + st = 0, and su = N .
From the first equation we have t = −r.
If we substitute t = −r into the third equation, then ru − rs = 0 or r(u − s) = 0.
Since r 6= 0, then u − s = 0 or u = s.
Thus, from the fourth equation we have N = su = u2 .
That is, to minimize N we need to minimize u2 .
If we substitute t = −r and s = u into the second equation, then u + rt + s = 2010
2010 + r2
becomes u + r(−r) + u = 2010 or 2u − r2 = 2010 and so u = .
2
2
Thus, u > 0. So to minimize u , we minimize u or equivalently, we minimize r.
Since u and r are integers and r 6= 0, u is minimized when r = ±2 (r must be even) or
2014
u= = 1007.
2
Therefore, the smallest positive integer N for which x4 + 2010x2 + N can be factored as
(x2 + rx + s)(x2 + tx + u) with r, s, t, u integers and r 6= 0 is N = u2 = 10072 = 1014049.
(c) Replacing the coefficient 2010 with M in part (b) and again equating coefficients, we have
the similar four equations t + r = 0, u + rt + s = M , ru + st = 0, and su = N .
Thus we have,
N −M = su − (u + rt + s)
37 = u2 − (2u − r2 )
37 = u2 − 2u + r2
37 + 1 = u2 − 2u + 1 + r2
38 = (u − 1)2 + r2
p
and so r = ± 38 − (u − 1)2 .
In the table below we attempt to find integer solutions for u and r:
u (u − 1)2 r
√
1 0 ± 38
√
0 or 2 1 ± 37
√
−1 or 3 4 ± 34
√
−2 or 4 9 ± 29
√
−3 or 5 16 ± 22
√
−4 or 6 25 ± 13
√
−5 or 7 36 ± 2
We see that for all choices of u above, r is not an integer.
For any other integer choice of u not listed, (u − 1)2 > 38 and then 38 − (u − 1)2 < 0, so
there are no real solutions for r.
Thus, when u is an integer, r cannot be, so u and r cannot both be integers. Therefore,
x4 + M x2 + N cannot be factored as in (b) for any integers M and N with N − M = 37.
Note: Alternatively, we could have stated that (u − 1)2 + r2 represents the sum of two
perfect squares. Since no pair of perfect squares (from the list 0, 1, 4, 9, 16, 25, 36) sums to
38, then (u − 1)2 + r2 6= 38 for any integers u and r.
Canadian
Mathematics
Competition

Solutions

1. Throughout this problem, we will need to know the total number of students in the class.
According to the chart, the total number is 3 + 2 + 1 + 2 + 4 + 2 + 2 + 3 + 1 = 20.
(a) There are 20 students in total.

Of these students, 2 have both green eyes and brown hair.
2
Therefore, the percentage who have both green eyes and brown hair is 20
× 100% = 10%.
(b) There are 20 students in total.
There are five entries in the table that appear in the either the “green eyes” row or the
“brown hair” column. Note that the entry that is in both the “green eyes” row and the
“brown hair” column is only counted once.
The sum of these five entries is 2 + 4 + 2 + 3 + 2 = 13.
Therefore, the percentage of the students with either green eyes or brown hair is
13
20
× 100% = 65%.
(c) There are 2 + 4 + 2 = 8 students in total with green eyes.
Of these students, 2 have red hair.
2
Therefore, the percentage of those with green eyes who have red hair is 8
× 100% = 25%.
(d) Initially, there are 20 students in the class, of whom 1 + 2 + 1 = 4 have red hair.
Suppose that x students with red hair join the class. There will then be 20 + x students
in total, of whom 4 + x have red hair.
4+x 36
We want = .
20 + x 100
4+x 9
Reducing the fraction on the right-hand side, this equation becomes = .
20 + x 25
Cross-multiplying, we obtain 25(4 + x) = 9(20 + x) which gives 100 + 25x = 180 + 9x and
so 16x = 80 or x = 5.
Therefore, 5 students with red hair must join the class.
4+x 9
(We could have seen that x = 5 by inspection from the equation = but this
20 + x 25
would not immediately guarantee us that this was the only solution to this equation.)
2. (a) Solution 1
Suppose that the middle term of the sequence is x.
If the common difference between terms in the sequence is d, then the first term is x − d
and the third term is x + d.
Since the sum of the terms is 180, then (x − d) + x + (x + d) = 180 or 3x = 180 and so
x = 60.
Therefore, the middle term in the sequence is 60.
Solution 2
Suppose that the first term in the sequence is a and the common difference between terms
in the sequence is d.
Thus, the second term is a + d and the third term is (a + d) + d = a + 2d.
The second term is the middle term, so we need to determine the value of a + d.
Since the sum of the three terms is 180, then a + (a + d) + (a + 2d) = 180 or 3a + 3d = 180
which gives 3(a + d) = 180.
Thus, a + d = 60 and so the middle term is 60.
(b) Solution 1
We show that the middle term (which is the third term) equals 36.
Suppose that the third term of the sequence is x.

If the common difference between terms in the sequence is d, then the second term is x−d,
the first term is x − 2d, the fourth term is x + d, and the fifth term is x + 2d.
Since the sum of the terms is 180, then
(x − 2d) + (x − d) + x + (x + d) + (x + 2d) = 180

5x = 180
x = 36
Therefore, the middle term in the sequence is 36.

(Note that if d = 0 then all of the terms in the sequence are equal to 36, so it is possible
for more than one term to equal 36.)
Solution 2
Suppose that the first term in the sequence is a and the common difference between terms
Thus, the second term is a + d, the third term is (a + d) + d = a + 2d, the fourth term is
a + 3d, and the fifth term is a + 4d.
The third term is the middle term, so we need to determine the value of a + 2d.
Since the sum of the five terms is 180, then
a + (a + d) + (a + 2d) + (a + 3d) + (a + 4d) = 180

5a + 10d = 180
5(a + 2d) = 180
a + 2d = 36
Thus, a + 2d = 36 and so the middle term equals 36.

(Note that if d = 0 then all of the terms in the sequence are equal to 36, so it is possible
for more than one term to equal 36.)
(c) Suppose that the first term in the sequence is a and the common difference between terms
Thus, the second term is a + d, the third term is a + 2d, the fourth term is a + 3d, the
fifth term is a + 4d, and the sixth term is a + 5d.
We need to determine the sum of the first and sixth terms, which equals
a + (a + 5d) = 2a + 5d.
Since the sum of the six terms is 180, then
a + (a + d) + (a + 2d) + (a + 3d) + (a + 4d) + (a + 5d) = 180

6a + 15d = 180
3(2a + 5d) = 180
2a + 5d = 60
Thus, 2a + 5d = 60 and so the sum of the first and sixth terms is 60.
3. (a) The line through B that cuts the area of 4ABC in half is the median – that is, the line
through B and the midpoint M of AC.
This line cuts the area of the triangle in half, because if we consider AC as its base, then
the height of each of 4AM B and 4CM B is equal to the distance of point B from the
line through A and C. These two triangles also have equal bases because AM = M C, so
their areas must be equal.
The midpoint, M , of AC has coordinates 21 (0 + 8), 21 (8 + 0) = (4, 4).

4−0
The slope of the line through B(2, 0) and M (4, 4) is = 2.
4−2
Since this line passes through B(2, 0), it has equation y − 0 = 2(x − 2) or y = 2x − 4.
(b) Since line segment RS is vertical and S lies on BC, which is horizontal, then 4RSC is
right-angled at S.
y
A
x
O B S C
0−8
Also, R lies on line segment AC, which has slope = −1.
8 −◦ 0
Since AC has a slope of −1, it makes an angle of 45 with the x-axis. In particular, the
angle between RC and SC is 45◦ .
Since 4RSC is right-angled at S and has a 45◦ angle at C, then the third-angle must be
180◦ − 90◦ − 45◦ = 45◦ , which means that the triangle is right-angled and isosceles.
Suppose that RS = SC = x.
Since 4RSC is right-angled, then the area of 4RSC in terms of x is 12 x2 .
But we know that the area of 4RSC is 12.5, so 21 x2 = 12.5 or x2 = 25.
Since x > 0, then x = 5.
This tells us that point S is 5 units to the left of C, so has coordinates (8 − 5, 0) = (3, 0).
Also, point R is 5 units above S, so has coordinates (3, 0 + 5) = (3, 5).
(c) Solution 1
Since line segment BC is horizontal and the line segment through T and U is also hori-
zontal, then BC and T U are parallel.
Therefore ∠AT U = ∠ABC.
y
A
T U
x
O B C
Since 4AT U and 4ABC also share a common angle at A, then 4AT U is similar to
4ABC.
Since 4AT U and 4ABC are similar, then the ratio of their areas equals the square of
the ratio of their heights.
Considering BC as the base of 4ABC, we see that its area is 21 (8 − 2)(8) = 24. Note that
its height is 8 when considered in this direction.
Suppose that the height of 4AT U considered from T U is h.
2
13.5 h h2 27 64(27)
Then = or = or h2 = = 36.
24 8 64 48 48
Since h > 0, then h = 6.
Therefore, the line segment T U is 6 units lower than the point A(0, 8), and so has equation
y = 8 − 6 or y = 2.
Solution 2
Suppose that the equation of the horizontal line is y = t.
We find the coordinates of points T and U first.
To do this, we need to find the equation of the line through A and B and the equation of
the line through A and C.
0−8
The line through A and B has slope = −4 and passes through (0, 8), so has equation
2−0
y = −4x + 8.
0−8
The line through A and C has slope = −1 and passes through (0, 8), so has equation
8−0
y = −x + 8.
The point T is the point on the line y = −4x + 8 with y-coordinate t.
To find the x-coordinate, we solve t = −4x + 8 to get 4x = 8 − t or x = 14 (8 − t).
The point U is the point on the line y = −x + 8 with y-coordinate t.
To find the x-coordinate, we solve t = −x + 8 to get x = 8 − t.
Therefore, T has coordinates ( 14 (8 − t), t), U has coordinates (8 − t, t), and A is at (0, 8).
To find the area, we remember that T U is horizontal and has length
(8 − t) − 14 (8 − t) = 43 (8 − t), and the distancefrom T U to A is 8 − t.
Therefore, the area in terms of t is 21 34 (8 − t) (8 − t) = 38 (8 − t)2 .
Since we know that the area equals 13.5, then 83 (8 − t)2 = 13.5 or (8 − t)2 = 83 (13.5) = 36.
Note that t < 8 because line segment T U is below A, so 8 − t > 0.
Therefore, 8 − t = 6 and so t = 8 − 6 = 2.
Thus, the equation of the horizontal line through T and U is y = 2.
4. (a) Since 4ABC is equilateral with side length 12 and X and Y are the midpoints of CA and
CB, respectively, then CX = CY = 21 (12) = 6.
Since the height of the prism is 16 and Z is the midpoint of CD, then CZ = 12 (16) = 8.
Since faces ACDE and BCDF are rectangles, then ∠ACD = ∠BCD = 90◦ .
Thus, 4XCZ and 4Y CZ are right-angled √ at C. √ √
By the Pythagorean Theorem, XZ = CX 2 + CZ 2 = 62 + 82 = 100 = 10.
√ √ √
Similarly, Y Z = CY 2 + CZ 2 = 62 + 82 = 100 = 10.
Lastly, consider 4CXY .
We know that CX = CY = 6 and that ∠XCY = 60◦ , because 4ABC is equilateral.
Thus, 4CXY is isosceles with ∠CXY = ∠CY X.
These angles must each be equal to 21 (180◦ − ∠XCY ) = 12 (180◦ − 60◦ ) = 60◦ .
But this means that 4CXY is equilateral, and so XY = CX = CY = 6.
Therefore, XY = 6 and XZ = Y Z = 10.
(b) To determine the surface area of solid CXY Z, we must determine the area of each of the
four triangular faces.
Areas of 4CZX and 4CZY
Each of these triangles is right-angled and has legs of lengths 6 and 8.
Therefore, the area of each is 12 (6)(8) = 24.
Area of 4CXY
This triangle is equilateral with side length 6.
We draw the altitude from C to M on XY . Since 4CXY is equilateral, then M is the
midpoint of XY .
C
6 6
X 3 M 3 Y
Each of 4CM X and 4CM Y is thus a 30◦ -60◦ -90◦ triangle, since each already has a 60◦
angle and a 90◦ angle. √ √ √
Using the ratios from this special triangle, CM = 23 (CX) = 23 (6) = 3 3.
√ √
Since XY = 6, then the area of 4CXY is 21 (6)(3 3) = 9 3.
Area of 4XY Z
Here, XY = 6 and XZ = Y Z = 10.
Again, we drop an altitude from Z to XY .
Since 4XY Z is isosceles, then this altitude meets XY at its midpoint, M .
Z
10 10
X 3 M 3 Y
Note that XM = M Y = 12 (XY ) = 3.

√ √ √
By the Pythagorean Theorem, ZM = ZX 2 − XM √
2 =
√ 102 − 32 = 91.
Since XY = 6, then the area of 4XY Z is 12 (6)( 91) = 3 91.
√ √ √ √
Therefore, the total surface area of solid CXY Z is 24+24+9 3+3 91 = 48+9 3+3 91.
(c) Step 1: Examination of 4M DN
We know that DM = 4, DN = 2, and ∠M DN = 60◦ (because 4EDF is equilateral).
Since DM : DN = 2 : 1 and the contained angle is 60◦ , then 4M DN must be a
30◦ -60◦ -90◦ triangle.
Therefore, M N is perpendicular to DF . √ √
Using the ratios in the special triangle, M N = 3DN = 2 3.
We could have instead calculated the length of M N using the cosine law to determine
this.
Step 2: Calculation of CP
We know that QC = 8 and ∠QCP = 60◦ .
Since M N is perpendicular to DF , this tells us that the plane M N P Q is perpendicular
to the plane BCDF .
Since QP is parallel to M N (they lie in the same plane M N P Q and in parallel planes
ACB and DEF ), then QP is perpendicular to CB.
Therefore, 4QCP is right-angled at P and contains a 60◦ angle, making it also a
30◦ -60◦ -90◦ triangle. √ √

Thus, CP = 12 (CQ) = 12 (8) = 4 and QP = 3CP = 4 3.
Step 3: Construction
Extend CD downwards.
Next, extend QM until it intersects the extension of CD at R. (Note here that the line
through QM will intersect the line through CD since they are two non-parallel lines lying
in the same plane.)
12
A B
Q
P
C
16
E F
M
N
D
Consider 4RDM and 4RCQ.

The two triangles share a common angle at R and each is right-angled (4RDM at D and
4RCQ at C), so the two triangles are similar.
Since QC = 8 and M D = 4, then their ratio of similarity is 2 : 1.
This means that RC = 2RD, ie. D is the midpoint of RC.
Since CD = 16, then DR = 16.
Similarly, since CP : DN = 2 : 1, then when P N is extended to meet the extension of
CD, it will do so at the same point R.
12
A B
Q
P
C
16
E F
M
N
D
Step 4: Calculation of volume of QP CDM N

The volume of QP CDM N equals the difference between the volume of the triangular
based pyramid RCQP and the volume of the triangular based pyramid RDM N .
(Another name for a triangular based pyramid is a tetrahedron.)
The volume of a tetrahedron equals one-third √ times √
the area of the base time the height.
The area of 4CP Q is 12 (CP )(QP ) = 21 (4)(4 3) = 8 3.
√ √
The area of 4DN M is 21 (DN )(M N ) = 21 (2)(2 3) = 2 3.
The length of RD is 16 and the length of RC √ is 32. 1 √ √ √ √
Therefore, the volume of QP CDM N is 3 (8 3)(32) − 3 (2 3)(16) = 2563 3 − 323 3 = 2243 3 .
1
Canadian
Mathematics
Competition

Solutions

1. (a) By definition, 3∇2 = 2(3) + 22 + 3(2) = 6 + 4 + 6 = 16.

(b) We have
x∇(−1) = 8
2
2x + (−1) + x(−1) = 8
x+1 = 8
x = 7
so x = 7.
(c) We have
4∇y = 20
2
2(4) + y + 4y = 20
y 2 + 4y + 8 = 20
y 2 + 4y − 12 = 0
(y + 6)(y − 2) = 0
so y = −6 or y = 2.
(d) We have
(w − 2)∇w = 14
2
2(w − 2) + w + (w − 2)w = 14
2w − 4 + w2 + w2 − 2w = 14
2w2 − 4 = 14
2w2 = 18
w2 = 9
so w = 3 or w = −3.
8−0 8
2. (a) The slope of the line through A(7, 8) and B(9, 0) is = = −4.
7−9 −2
Therefore, the line has equation y = −4x + b for some b.
Since B(9, 0) lies on this line, then 0 = −4(9) + b so b = 36.
Thus, the equation of the line is y = −4x + 36.
(b) We want to determine the point of intersection between the lines having equations y =
−4x + 36 and y = 2x − 10.
Equating values of y, we obtain −4x + 36 = 2x − 10 or 46 = 6x so x = 233
.
We substitute this value of x into the equation y = 2x − 10 to determine the value of y,
obtaining y = 2( 23
3
) − 10 = 46
3
− 30
3
= 16
3
.
23 16
Thus, the coordinates of P are ( 3 , 3 ).
(c) Solution 1
The x-coordinate of A is 7 and the x-coordinate of B is 9.
The average of these x-coordinates is 12 (7 + 9) = 8.
Since the x-coordinate of P is 23
3
< 8, then the x-coordinate of P is closer to that of A
than that of B.
Since the points P , A and B lie on a straight line, then P is closer to A than to B.
Solution 2
The y-coordinate of A is 8 and the y-coordinate of B is 0.
The average of these y-coordinates is 21 (8 + 0) = 4.
Since the y-coordinate of P is 16
3
> 4, then the y-coordinate of P is closer to that of A
than that of B.
Since the points P , A and B lie on a straight line, then P is closer to A than to B.
Solution 3
The coordinates of A are (7, 8), of B are (9, 0), and of P are ( 23 , 16 ), then
3 3
q q q
23 2 16 2
PA = (7 − 3
) + (8 − 3
) = (− 23 )2 + ( 38 )2 = 68
9
and q q q
23 2 16 2
PB = (9 − 3
) + (0 − 3
) = ( 43 )2 + (− 16
3
)2 = 272
9
Thus, P B > P A, so P is closer to A than to B.
3. (a) Solution 1
Trapezoid ABCD has bases AD = 6 and BC = 30, and height AB = 20. (AB is a height
since it is perpendicular to BC.)
Therefore, the area of ABCD is 21 (6 + 30)(20) = 360.
Solution 2
Join B to D.
Since AB is perpendicular to BC and AD is parallel to BC, then AB is perpendicular
to AD.
Thus, 4DAB is right-angled at A, and so has area 12 (6)(20) = 60.
Also, 4BDC can be considered as having base BC = 30 and height equal to the length
of BA (that is, height equal to 20), and so has area 12 (30)(20) = 300.
The area of trapezoid ABCD is the sum of the areas of 4DAB and 4BDC, so equals
60 + 300 = 360.
Solution 3
Since AB is perpendicular to BC and AD is parallel to BC, then AB is perpendicular
to AD.
Drop a perpendicular from D to F on BC.
Then ADF B is a rectangle that is 6 by 20, and so has area 6(20) = 120.
Also, F C = BC − BF = 30 − AD = 30 − 6 = 24.
Thus, 4DF C is right-angled at F , has height DF = AB = 20 and base F C = 24. Hence,
the area of 4DF C is 12 (20)(24) = 240.
The area of trapezoid ABCD is the sum of the areas of rectangle ADF B and 4DF C, so
equals 120 + 240 or 360.
(b) First, we note that if K is on AB, then 4KBC and quadrilateral KADC cover the entire
area of trapezoid ABCD.
A 6 D
B C
30
Thus, if the areas of 4KBC and quadrilateral KADC are equal, then each equals one-half
of the area of trapezoid ABCD, or 12 (360) = 180.
Suppose that BK = h.
Then 4KBC has base BC = 30 and height BK = h, and so 12 (30)h = 180, or h = 12.
Therefore, BK = 12.
(c) Solution 1
As in (b), we want the area of 4M BC to equal 180. Drop a
perpendicular from M to N on BC. A 6 D
Again as in (b), since the base BC of 4M BC has length 30,
then the height M N of 4M BC needs to be 12 for its area to M
be 180. 20
Drop a perpendicular from D to F on BC.
Since DF is perpendicular to BC, then ADF B is a rectangle,
so BF = 6, which gives F C = BC − BF = 30 − 6 = 24. B N C
F
Also, DF = AB = 20.
√ √ √ √
By the Pythagorean Theorem, DC = 202 + 242 = 400 + 576 = 976 = 4 61.
Lastly, we must calculate the length of M C.
Method 1
DF 20 5
We know that sin(∠DCF ) = = √ =√ .
DC 4 61 61
MN 12 12
√
Since M N = 12, then M C = = √ , or M C = 5
61.
sin(∠DCF ) 5/ 61
Method 2
We know that 4DF C is similar to 4M N C, since each is right-angled and they have a
common angle at C.
MC DC √ √
Therefore, = so M C = 1220
(4 61) = 12
5
61.
MN DF
Method 3
NC FC
Since 4DF C is similar to 4M N C, then = so N C = 12(24)
20
= 72
5
.
MN DF
By the Pythagorean Theorem,
√ q 2 12 √ √
M C = M N 2 + N C 2 = 122 + 72 5
= 5 62 + 52 = 12 5
61
Solution 2
As in (b), we want the area of 4M BC to equal 180.
A 6 D
M
20
B N C
Drop a perpendicular from M to N on BC.

Again as in (b), since the base BC of 4M BC has length 30, then the height M N of
4M BC needs to be 12 for its area to be 180.
We place the diagram on a coordinate grid, with B at the origin, A on the positive y-axis,
and C on the positive x-axis.
Thus, the coordinates of B are (0, 0), the coordinates of A are (0, 20), the coordinates of
D are (6, 20), and the coordinates of C are (30, 0).
Since the length of M N is 12, then the coordinates of M are (s, 12) for some real number s.
0 − 20 0 − 12
But M lies on DC, so the slope of DC equals the slope of M C, or = so
30 − 6 30 − s
−20(30 − s) = 24(−12) or 20s − 600 = −288 or 20s = 312 or s = 78 5
.
Using the coordinates of M and of C, we have
q 2 q
72 2
√ √
MC = 30 − 78
5
+ (0 − 12) 2 =
5
+ 122 = 12
5
62 + 52 = 125
61
4. (a) Taking all possible products of pairs,
2(3) + 2x + 2(2x) + 3x + 3(2x) + x(2x) = −7

6 + 15x + 2x2 = −7
2x2 + 15x + 13 = 0
(2x + 13)(x + 1) = 0
so x = −1 or x = − 13
2
.
(b) Since each term equals 1, −1 or 2, then the pairs of the terms whose products is 1 are
those coming from 1 × 1 and (−1) × (−1).
We know that there are m terms equal to 1. How many pairs can these terms form?
To form a pair, there are m choices for the first entry and then m − 1 choices for the first
entry (all but the first term chosen). This gives m(m − 1) pairs.
But we have counted each pair twice here (as we have counted both a and b as well as b
and a, while we only want to count one of these), so we divide by 2 to obtain 12 m(m − 1)
pairs.
m m(m − 1)
(Alternatively, we could have said that there were = pairs.)
2 2
We also know that there are n terms equal to −1.
In a similar way, these will form 21 n(n − 1) pairs.
Therefore, in total, there are 12 m(m − 1) + 21 n(n − 1) pairs of distinct terms whose product
is 1.
(c) Suppose that the sequence contains m terms equal to 2 and so n = 100 − m terms equal
to −1.
The terms equal to 2 form 21 m(m − 1) pairs, each contributing 2 × 2 = 4 to the peizi-sum.
The terms equal to −1 form 21 n(n − 1) = 12 (100 − m)(99 − m) pairs each equal to
(−1) × (−1) = 1.
Since there are m terms equal to 2 and 100 − m terms equal to −1, then there are
m(100 − m) pairs formed by choosing one 2 and one −1, and so m(100 − m) pairs of
terms, each contributing 2 × (−1) = −2 to the peizi-sum.
These are all of the possible types of pairs, so the peizi-sum, S, is
S = 4 12 m(m − 1) + 1 21 (100 − m)(99 − m) + (−2) (m(100 − m))

or
S = 2m2 − 2m + 50(99) − 199
2
m + 12 m2 − 200m + 2m2
or
S = 29 m2 − 603
2
m + 4950
The equation S = 92 m2 − 603
2
m + 4950 is a quadratic equation in m (forming a parabola
opening upwards), and so is minimized at its vertex, which occurs at
− 603 67
m=− 2
9
= 2
= 33 12
2 2
However, this value of m is not an integer, so is not the value of m that solves our problem.
The parabola formed by this equation is symmetric about its vertex and increases in either
direction from its vertex, so the minimum value at an integer value of m occurs at both
m = 33 and m = 34 (since these are the closest integers to 33 21 and are the same distance
from 33 12 ).
We can substitute either value to determine the minimum possible peizi-sum, which is
9
2
(33)2 − 6032
(33) + 4950 = −99.
Canadian
Mathematics
Competition

Solutions
c 2007 Waterloo Mathematics Foundation

1. (a) The possible routes are:

A→B→C→D→A A→B→D→C→A
A→C→B→D→A A→C→D→B→A
A→D→B→C→A A→D→C→B→A
(b) We list each route and its length:
A→B→C→D → A: Length AB + BC + CD + DA = 80 + 120 + 90 + 40 = 330 km
A→B→D→C → A: Length AB + BD + DC + CA = 80 + 60 + 90 + 105 = 335 km
A→C→B→D → A: Length AC + CB + BD + DA = 105 + 120 + 60 + 40 = 325 km
A→C→D→B → A: Length AC + CD + DB + BA = 105 + 90 + 60 + 80 = 335 km
A→D→B→C → A: Length AD + DB + BC + CA = 40 + 60 + 120 + 105 = 325 km
A→D→C→B → A: Length AD + DC + CB + BA = 40 + 90 + 120 + 80 = 330 km
The two routes of shortest length are A → C → B → D → A and A → D → B → C → A,
which are each of length 325 km.
The two routes of longest length are A → B → D → C → A and A → C → D → B → A,
which are each of length 335 km.
(c) Solution 1
We can list the possible routes:
A→B→C→E→D→A A→B→D→E→C→A
A→C→B→E→D→A A→C→D→E→B→A
A→D→B→E→C→A A→D→C→E→B→A
Therefore, there are 6 possible routes.
(Note that in fact each route from (a) gives a route here in (c) by adding an E between
the third and fourth stops on the original route.)
Solution 2
Consider a route A → x → y → E → z → A.
There are 3 possibilities for x (B, C or D).
For each of these possibilities, there are 2 possibilities for y.
After x and y are chosen, there is only 1 possibility of z.
So there are 3 × 2 = 6 possible routes.
(d) From the first piece of information, AD + DC + CE + EB + BA = 600 km so 40 + 90 +
CE + EB + 80 = 600 km or CE + EB = 390 km.
From the second piece of information, AC + CD + DE + EB + BA = 700 km so
105 + 90 + 225 + EB + 80 = 700 km or EB = 200 km.
Since EB = 200 km and CE + EB = 390 km, then CE = 190 km, so the distance from
C to E is 190 km.
2. (a) Here is a sequence of moves that works:

Move # P Q R S Comment
9 9 1 5
1 8 8 4 4 3 added to R
2 7 7 7 3 3 added to R
3 6 6 6 6 3 added to S
There are other sequences of moves that will work.
(b) i. In total, there are 31 + 27 + 27 + 7 = 92 marbles, so if there is an equal number in
each pail, there must be 23 in each pail.
Here is a sequence of moves that works:

Move # P Q R S Comment
31 27 27 7
1 30 26 26 10 3 added to S
2 29 25 25 13 3 added to S
3 28 24 24 16 3 added to S
4 27 23 23 19 3 added to S
5 26 22 22 22 3 added to S
6 25 21 21 25 3 added to S
7 24 24 20 24 3 added to Q
8 23 23 23 23 3 added to R
There are other sequences of moves that will work.
ii. Initially, pail P contains 31 marbles.
We want pail P to contain 23 marbles, so we must decrease the number of marbles in
pail P by 8.
In any legal move, the number of marbles in pail P decreases by at most 1 (that is, it
decreases by 1 or increases by 3).
Therefore, we need at least 8 legal moves in which the number of marbles in pail P
decreases (and potentially some where the number of marbles in pail P increases).
Thus, it takes at least 8 legal moves to obtain the same number of marbles in each
pail.
(Note that in part (i), we showed that we could do this in 8 legal moves, so 8 is the
minimum number of moves needed.)
(c) Solution 1
Starting with 10, 8, 11, and 7 marbles in the pails, there are 10 + 8 + 11 + 7 = 36 marbles
in total.
To have an equal number of marbles in each pail, we would need 36 ÷ 4 = 9 marbles in
each pail.
On any legal move, the number of marbles in any pail decreases by 1 or increases by 3.
If the pail contains an even number n of marbles before a legal move, then it will contain
either n−1 or n+3 marbles after the legal move, so will contain an odd number of marbles.
Similarly, if the pail contains an odd number of marbles before a legal move, then it will
contain an even number of marbles after the legal move.
But we start with two pails containing an even number of marbles and two pails containing
an odd number of marbles.
After the first legal move, the pails originally containing an even number of marbles will
contain an odd number of marbles and the pails originally containing an odd number of
marbles will contain an even number of marbles.
This gets us back to the the same situation – two pails with an even number and two pails
with an odd number of marbles.
Therefore, after any move, this situation will not change.
Therefore, it is impossible to ever have 9 marbles in each pail, as there will always be two
pails containing an even number of marbles.
Solution 2
Starting with 10, 8, 11, and 7 marbles in the pails, there are 10 + 8 + 11 + 7 = 36 marbles
in total.
To have an equal number of marbles in each pail, we would need 36 ÷ 4 = 9 marbles in
each pail.
On any legal move, the number of marbles in any pail decreases by 1 or increases by 3.
Assume that it is possible to end up with 9 marbles in each pail.

We show that this cannot happen by proving the following fact:
If it is possible to end up with 9 marbles in each pail, then after each move, the
difference between the number of marbles in any two pails must be a multiple
of 4.
Assume that this fact was true after a certain move. (We know that it is true at the end,
since the difference between the numbers in any pair of pails is 0.)
Suppose that there were a, b, c and d marbles in the pails.
Pick two of the four pails (say, the pails with a and b marbles). Before this move (that
is, after the previous move), either these two pails each had 1 more marble in each (so
a + 1 and b + 1 marbles which preserves the difference) or one pail had 1 more marble and
the other had 3 fewer marbles (so a + 1 and b − 3 or a − 3 and b + 1 which changes the
difference by 4).
Therefore, before this move, the differences between the number of marbles in the pails
are all multiples of 4.
This tells us that, to end up with 9 marbles in each pail, the difference between the num-
bers of marbles in any pair of pails is always a multiple of 4.
But this is not true with our initial condition of 10, 8, 11 and 7 marbles (since, for example,
11 − 10 is not a multiple of 4).
Therefore, it is impossible to end up with an equal number of marbles in each pail.
3. (a) If f (x) = 0, then x2 − 4x − 21 = 0.

Factoring the left side, we obtain (x − 7)(x + 3) = 0, so x = 7 or x = −3.
(We could obtain the same values of x by using the quadratic formula.)
(b) Solution 1
Completing the square in the original function,
f (x) = x2 − 4x − 21 = x2 − 4x + 4 − 4 − 21 = (x − 2)2 − 25
so the axis of symmetry of the parabola y = f (x) is the vertical line x = 2. (The axis of
symmetry could also have been found using the average of the roots from (a).)
If f (s) = f (t), then s and t are symmetrically located around the axis of symmetry.
y
s t x
x=2
In other words, the average value of s and t is the x-coordinate of the axis of symmetry,
so 12 (s + t) = 2 or s + t = 4.
(Note that this agrees with our answer from part (a), but that we needed to proceed
formally here to make sure that there were no other answers.)
Solution 2
Rearranging,
s2 − 4s − 21 = t2 − 4t − 21
s2 − t2 − 4s + 4t = 0
(s + t)(s − t) − 4(s − t) = 0
(s + t − 4)(s − t) = 0
Therefore, s + t − 4 = 0 or s − t = 0.
Since we are told that s and t are different real numbers, then s − t 6= 0.
Therefore, s + t − 4 = 0 or s + t = 4.
(c) Solution 1
Proceeding algebraically in a similar way to part (b), Solution 2,
(a2 − 4a − 21) − (b2 − 4b − 21) = 4

a2 − b2 − 4a + 4b = 4
(a + b − 4)(a − b) = 4
Since a and b are integers, then a + b − 4 and a − b are integers as well. In particular, they
are integers whose product is 4.
We make a table to check the possibilities:
a + b − 4 a − b 2a − 4 a b
9 7
4 1 5 2 2
2 2 4 4 2
9 1
1 4 5 2 2
−4 −1 −5 − 12 12
−2 −2 −4 0 2
−1 −4 −5 − 12 72
Therefore, the one pairs of positive integer values of a and b that works is (a, b) = (4, 2).
(Note that we could have cut down our work in this table by noticing that if a + b − 4 = x
and a − b = y, then 2a = x + y, so x + y (that is, the sum of the values of a + b − 4 and
a − b) must be even, which eliminates all but two of the rows in the table.)
Solution 2
As in part (b), the axis of symmetry of the parabola y = f (x) is x = 2.
Since the parabola has leading coefficient +1, then it is the same shape as the parabola
y = x2 .
In the parabola y = x2 (and so in the parabola y = f (x)), the lattice points moving to
the right from the axis of symmetry are (0, 0), (1, 1), (2, 4), (3, 9), (4, 16), and so on. The
vertical distances moving from one point to the next are 1, 3, 5, 7, and so on.
A similar pattern is true when we move successive units to the left from the axis of sym-
metry.
Starting from the left, the sequence of successive vertical differences is thus
. . . , −7, −5, −3, −1, 1, 3, 5, 7, . . .

For f (a) − f (b) = 4 with a and b integers, we must find a sequence of consecutive differ-
ences that add to 4 or −4 (depending on whether a or b is further to the left).
We can only get 4 or −4 by using (−3) + (−1) or 1 + 3. The relative positions of these are
starting at the axis of symmetry and moving two units to the right, or starting two units
to the left of the axis of symmetry and moving two units to the right.
Since the axis of symmetry for the given parabola is x = 2, then the only solution is
(a, b) = (4, 2), since a and b must both be positive.
4. (a) Join P Q, P R, P S, RQ, and RS.

Since the circles with centre Q, R and S are all tangent to BC, then QR and RS are each
parallel to BC (as the centres Q, R and S are each 1 unit above BC).
This tells us that QS passes through R.
When the centres of tangent circles are joined, the line segments formed pass through the
associated point of tangency, and so have lengths equal to the sum of the radii of those
circles.
Therefore, QR = RS = P R = P S = 1 + 1 = 2.
Q R S
Since P R = P S = RS, then 4P RS is equilateral, so ∠P SR = ∠P RS = 60◦ .

Since ∠P RS = 60◦ and QRS is a straight line, then ∠QRP = 180◦ − 60◦ = 120◦ .
Since QR = RP , then 4QRP is isosceles, so ∠P QR = 21 (180◦ − 120◦ ) = 30◦ .
Since ∠P QS = 30◦ and ∠P SQ = 60◦ , then ∠QP S = 180◦ − 30◦ − 60◦ = 90◦ , so 4P QS
is a 30◦ -60◦ -90◦ triangle.
(b) In (a), we saw that QS is parallel to BC.
Similarly, since P and S are each one unit from AC, then P S is parallel to AC.
Also, since P and Q are each one unit from AB, then P Q is parallel to AB.
Therefore, the sides of 4P QS are parallel to the corresponding sides of 4ABC.
Thus, the angles of 4ABC are equal to the corresponding angles of 4P QS, so 4ABC is
a 30◦ -60◦ -90◦ triangle.
This means that if we can determine one of the side lengths of 4ABC, we can then
determine the lengths of the other two sides using the side ratios in a 30◦ -60◦ -90◦ triangle.
Consider side AC.
Since the circle with centre P is tangent to sides AB and AC, then the line through A
and P bisects ∠BAC. Thus, ∠P AC = 45◦ .
Similarly, the line through C and S bisects ∠ACB. Thus, ∠SCA = 30◦ .
We extract trapezoid AP SC from the diagram, obtaining
S 2 P
1 1
30 45
C Z X A
or
A X Z C
45 30
1 1
P 2 S
depending on your perspective. Drop perpendiculars from P and S to X and Z on side

AC.
Since P S is parallel to AC and P X and SZ are perpendicular to AC, then P XZS is a
rectangle, so XZ = P S = 2.
Since 4AXP is right-angled at X, has P X = 1 (the radius of the circle), and ∠P AX =
45◦ , then AX = P X = 1.
Since 4CZS√is right-angled
√ at Z, has SZ = 1 (the radius of the circle), and ∠SCZ = 30◦ ,
then CZ = 3SZ = √ 3 (since √ 4SZC is also a 30◦ -60◦ -90◦ triangle).
Thus, AC = 1 + 2 + 3 = 3 + 3.
Since 4ABC√is a 30◦ -60◦ -90√ ◦
triangle,√with ∠ACB
√ =√60◦ and ∠CAB = 90◦ , then BC =
2AC = 6 + 2 3, and AB = 3AC = 3(3 + 3) √ = 3 3 + 3. √ √
Therefore, the side lengths of 4ABC are AC = 3+ 3, AB = 3 3+3, and BC = 6+2 3.
(c) After the described transformation, we obtain the following diagram.
A
Q R
S
B C
Drop perpendiculars from Q, R and S to D, E and F respectively on BC. Since the

circles with centres Q, R and S are tangent to BC, then D, E and F are the points of
tangency of these circles to BC.
Thus, QD = SF = 1 and RE = r.
Join QR, RS, P S, P Q, and P R.
Since we are connecting centres of tangent circles, then P Q = P S = 2
and QR = RS = P R = 1 + r.
Join QS.
By symmetry, P RE is a straight line (that is, P E passes through R).
Since QS is parallel to BC as in parts (a) and (b), then QS is perpendicular to P R,
meeting at Y .
Q Y S
R
D E F
Since QD = 1, then Y E = 1. Since RE = r, then Y R = 1 − r.

Since QR = 1 + r, Y R = 1 − r and 4QY R is right-angled at Y , then, by the Pythagorean
Theorem,
QY 2 = QR2 − Y R2 = (1 + r)2 − (1 − r)2 = (1 + 2r + r2 ) − (1 − 2r + r2 ) = 4r
Since P R = 1 + r and Y R = 1 − r, then P Y = P R − Y R = 2r.

Since 4P Y Q is right-angled at Y , then
P Y 2 + Y Q2 = P Q2
(2r)2 + 4r = 22
4r2 + 4r = 4
r2 + r − 1 = 0
p √
−1 ± 12 − 4(1)(−1) −1 ± 5
By the quadratic formula, r = = .
√ 2 2
−1 + 5
Since r > 0, then r = (which is the reciprocal of the famous “golden ratio”).
2
Canadian
Mathematics
Competition

Solutions

1. (a) Solution 1
The first odd positive integer is 1. The second odd positive integer is 3, which is 2 larger
than the first. The third odd positive integer is 5, which is 2 larger than the second.
Therefore, the 25th odd positive integer will be 24 × 2 = 48 larger than the first odd
positive integer, since we must add 2 to get to each successive odd number.
Thus, the 25th odd positive integer is 1 + 48 = 49.
There are 1 + 2 + 3 + 4 + 5 + 6 = 21 integers in the first six rows of the pattern, so 49
must appear in the 7th row.
Solution 2
The first odd positive integer is 1, which is 1 less than the first even positive integer,
namely 2.
The second odd positive integer is 3, which is 1 less than the second even positive integer,
namely 4.
This pattern continues, with the 25th odd positive integer being 1 less than the 25th even
positive integer, which is 25 × 2 = 50.
Therefore, the 25th odd positive integer is 49.
There are 1 + 2 + 3 + 4 + 5 + 6 = 21 integers in the first six rows of the pattern, so 49
must appear in the 7th row.
(b) In the triangular pattern, the first row contains 1 number, the second row 2 numbers, the
third row 3 numbers, and so on.
Thus, the first twenty rows contain 1 + 2 + 3 + · · · + 19 + 20 numbers in total. This total
20(21) n(n + 1)
equals = 210 (using the fact that 1 + 2 + · · · + (n − 1) + n = ).
2 2
Therefore, the 19th integer in the 21st row is the 210 + 19 = 229th odd positive integer.
Using either of the methods of part (a), this integer is 1 + 228(2) = 229(2) − 1 = 457.
(c) To get from 1 to 1001, we must add 2 a total of 500 times, so 1001 is the 501st odd positive
integer.
From part (b), we know that the first 20 rows contain 210 integers, so the row number is
larger than 20.
How many integers do the first 30 rows contain? They contain 1 + 2 + · · · + 29 + 30 =
30(31)
= 465 integers.
2
This tells us that the first 31 rows contain 465 + 31 = 496 integers.
Since 1001 is the 501st odd positive integer, it must be the 5th integer in the 32nd row.
2. (a) We find CE by first finding BE.

Since AE = 24 and ∠AEB = 60◦ , then BE = 24 cos(60◦ ) = 24 21 = 12.

(We could have also used the fact that 4ABE is a 30◦ -60◦ -90◦ triangle and the appropri-
ate ratios.)
Since BE = 12 and ∠BEC = 60◦ , then CE = 12 cos(60◦ ) = 12 21 = 6.

(b) Using the same strategy as in (a),
◦
√
3
√
AB = 24 sin(60 ) = 24 2
= 12 3
√ √
BC = 12 sin(60◦ ) = 12 23 =6 3
√ √
CD = 6 sin(60◦ ) = 6 23 = 3 3
ED = 6 cos(60◦ ) = 6 1

2
=3
The perimeter of quadrilateral ABCD is√equal √

to AB √
+ BC + CD + DA and√
DA = DE + EA, so the perimeter is 12 3 + 6 3 + 3 3 + 3 + 24 = 27 + 21 3.
(c) The area of quadrilateral ABCD is equal to the sum of the areas of triangles ABE, BCE
and CDE.
Thus,
1
Area = 2
(BE)(BA) + 21 (CE)(BC) + 21 (DE)(DC)
1
√ √ √
= 2
(12)(12 3) + 12 (6)(6 3) + 12 (3)(3 3)
√ √ √
= 72 3 + 18 3 + 92 3
189
√
= 2
3
−1 − 7
3. (a) The line through points B and C has slope = −1.
7 − (−1)
Since (7, −1) lies on the line, the line has equation y − (−1) = −1(x − 7) or y = −x + 6.
(b) Solution 1
Suppose P has coordinates (x, y). Since P lies on `, then y = −x + 6, so P has coordinates
(x, −x + 6). p
Since A has coordinates (10, −10), then P Ap= (x − 10)2 + (−x + 16)2 .
Since O has coordinates (0, 0), then P O = x2 + (−x + 6)2 .
For P A = P O, we must have P A2 = P O2 or
(x − 10)2 + (−x + 16)2 = x2 + (−x + 6)2

x2 − 20x + 100 + x2 − 32x + 256 = x2 + x2 − 12x + 36
320 = 40x
x = 8
so P has coordinates (8, −2).

y
x
O
B
A (10, 10)
Solution 2
For P to be equidistant from A and O, P must lie on the perpendicular bisector of AO.
Since A has coordinates (10, −10) and O has coordinates (0, 0), then AO has slope −1, so
the perpendicular bisector has slope 1 and passes through the midpoint, (5, −5), of AO.
Therefore, the perpendicular bisector has equation y − (−5) = x − 5 or y = x − 10.
Thus, P must be the point of intersection of the lines y = x − 10 and the line ` which has
equation y = −x + 6, so −x + 6 = x − 10 or 2x = 16 or x = 8.
Therefore, P has coordinates (8, −2).
(c) Solution 1
Since Q lies on the line `, then its coordinates are of the form (q, −q + 6), as in (b).
For ∠OQA = 90◦ , we need the slopes of OQ and QA to be negative reciprocals (that is,
the product of the slopes is −1).
−q + 6
The slope of OQ is .
q
−q + 6 − (−10) −q + 16
The slope of QA is = .
q − 10 q − 10
Therefore, we must solve
−q + 6 −q + 16
· = −1
q q − 10
(−q + 6)(−q + 16) = −q(q − 10)
q 2 − 22q + 96 = −q 2 + 10q
2q 2 − 32q + 96 = 0
q 2 − 16q + 48 = 0
(q − 4)(q − 12) = 0
Therefore, q = 4 or q = 12, so Q has coordinates (4, 2) or (12, −6).

y
x
O
B
A (10, 10)
Solution 2
For ∠OQA = 90◦ , then Q must lie on the circle with diameter OA.
The centre of the circle with diameter OA is the midpoint
p M of OA, √ or (5, −5).
2 2
The radius of the circle with diameter OA is OM or 5 + (−5) = 50.
Therefore, this circle has equation (x − 5)2 + (y + 5)2 = 50.
The points Q that we are seeking are the points on both the circle and the line `, ie. the
points of intersection. Since y = −x + 6,
(x − 5)2 + (−x + 6 + 5)2 = 50

x2 − 10x + 25 + x2 − 22x + 121 = 50
2x2 − 32x + 96 = 0
x2 − 16x + 48 = 0
(x − 4)(x − 12) = 0
Therefore, x = 4 or x = 12, so Q has coordinates (4, 2) or (12, −6).
4. (a) Suppose p is a prime number.

The only positive divisors of p are 1 and p, so σ(p) = 1 + p.
Thus,
1+p 1
I(p) = = + 1 ≤ 12 + 1 = 3
2
p p
since p ≥ 2.
(b) Solution 1
Suppose that p is an odd prime number and k is a positive integer. Note that p ≥ 3.
The positive divisors of pk are 1, p, p2 , . . ., pk−1 , pk , so
1 + p + p2 + · · · + pk 1 1(pk+1 − 1) pk+1 − 1

k
I(p ) = = =
pk pk p−1 pk+1 − pk
Thus,
I(pk ) < 2
pk+1 − 1
⇐⇒ < 2
pk+1 − pk
⇐⇒ pk+1 − 1 < 2(pk+1 − pk )
⇐⇒ 0 < pk+1 − 2pk + 1
⇐⇒ 0 < pk (p − 2) + 1
which is true since p ≥ 3.
Solution 2
Suppose that p is an odd prime number and k is a positive integer. Note that p ≥ 3.
The positive divisors of pk are 1, p, p2 , . . ., pk−1 , pk , so
pk + pk−1 + · · · + p + 1
I(pk ) =
pk
1 1 1
= 1 + + · · · + k−1 + k
p p p
k+1 !
1
1 1−
p
= (geometric series)
1
1−
p
1
<
1
1−
p
p
=
p−1
1
= 1+
p−1
1
≤ 1+
2
< 2
1 + p + p2
(c) Since p is a prime number, the positive divisors of p2 are 1, p and p2 , so I(p2 ) = .
p2
1+q
Since q is a prime number, I(q) = , as in (a).
q
Since p and q are prime numbers, the positive divisors of p2 q are 1, p, p2 , q, pq, and p2 q,
so
1 + p + p2 + q + pq + p2 q
I(p2 q) =
p2 q
(1 + p + p2 ) + q(1 + p + p2 )
=
p2 q
(1 + p + p2 )(1 + q)
=
p2 q
1 + p + p2 1 + q
= ·
p2 q
2
= I(p )I(q)
as required.
(d) We start by listing a number of facts:

• From part (b), if n is a prime number, then I(n) < 2. Thus, to obtain an n with
I(n) > 2, we must combine multiple primes.
• If n = p1 e1 p2 e2 · · · pm em where the pi ’s are distinct primes and the ei ’s are positive inte-
gers, then by extending the ideas of (c), we can see that I(n) = I(p1 e1 )I(p2 e2 ) · · · I(pm em ).
• If p and q are primes with p < q, then I(pk ) > I(q k ).
pk + pk−1 + · · · + p + 1 1 1 1
(This is because I(pk ) = k
= 1 + + · · · + k−1 + k and
p p p p
1 1 1 1 1 1 1
I(q k ) = 1 + + · · · + k−1 + k . If p < q, then > , 2 > 2 , etc., so I(pk ) > I(q k ).)
q q q p q p q
• The previous point tells us that smaller primes are more efficient than larger primes
at increasing I(n). In other words, given an n, we can find an m with I(m) > I(n)
by replacing some of the prime factors of n with smaller primes to obtain m. (For
example, if n = 52 73 11, then m = 52 33 7 would gives I(m) > I(n).)
p
• From (b), I(pk ) < so in particular I(3a ) < 32 and I(5b ) < 45 . Therefore,
p−1
I(3a 5b ) = I(3a )I(5b ) < 23 · 54 = 158
< 2.
• The previous two points tell us that no odd integer n with at most two prime factors
can have I(n) > 2. Therefore, to get I(n) > 2, n must have at least three prime
factors.
Let us consider integers of the form n = 3a 5b 7c and try to find one with I(n) > 2:
a b c n I(n)
1 1 1 105 I(n) = I(3)I(5)I(7) = 34 · 65 · 87 = 64
35
<2
2 13 6 8 208
2 1 1 315 I(n) = I(3 )I(5)I(7) = 9 · 5 · 7 = 105 < 2
3 1 1 945 I(n) = I(33 )I(5)I(7) = 40 · 6 · 8 = 128
27 5 7 63
>2
So I(945) > 2.
Why is n = 945 the smallest odd integer with I(n) > 2?

• We notice that any positive odd integer with at least 4 prime factors is at least
3(5)(7)(11) or 1155, which is larger than 945, so we can restrict to looking at integers
with at most 3 prime factors.
• By our opening remarks, we can restrict our search even further, looking at only those
odd integers with 3, 5 and 7 as possible prime factors, since we can decrease the primes
and make I(n) larger at the same time.
• Furthermore, we only need to consider those integers n = 3e1 5e2 7e3 with e1 ≥ e2 ≥ e3 ,
since otherwise we could reassign the exponents in this order and obtain a smaller
integer. (For example, 32 53 7 is larger than 33 52 7.)
• Since n must have at least three prime factors and we can assume that these prime
factors are 3, 5 and 7 and that the exponents have the property from the previous
point, then there are no integers smaller than 945 left to check other than the ones in
the table above (none of which have I(n) > 2).
Therefore, n = 945 is the smallest odd positive integer with I(n) > 2.
Canadian
Mathematics
Competition

Solutions

1. (a) By definition, 2 3 = 22 − 4(3) = 4 − 12 = −8.

(b) By definition, k 2 = k 2 − 4(2) = k 2 − 8.
By definition, 2 k = 22 − 4(k) = 4 − 4k.
So we want to solve
k 2 − 8 = 4 − 4k
k 2 + 4k − 12 = 0
(k + 6)(k − 2) = 0
so k = −6 or k = 2.
Checking, (−6) 2 = (−6)2 − 4(2) = 28, 2 (−6) = 22 − 4(−6) = 28, so k = −6 works.
Also, if k = 2, then 2 2 = 2 2 so k = 2 works as well.
(c) Since 3 x = y, then 32 − 4x = y or 9 − 4x = y.
Since 2 y = 8x, then 22 − 4y = 8x or 4 − 4y = 8x.
We now have a system of two equations in two unknowns.
Since 4 − 4y = 8x and y = 9 − 4x, then
4 − 4(9 − 4x) = 8x
4 − 36 + 16x = 8x
8x = 32
x = 4
Since x = 4, then y = 9 − 4(4) = −7.

(We could have solved this system of equations in several different ways instead.)
Checking, 3 x = 3 4 = 32 − 4(4) = −7 = y and 2 y = 2 (−7) = 22 − 4(−7) = 32 = 8x,
so x = 4, y = −7 is indeed the solution.
2. (a) Since 3, then 1, then 4 toothpicks have been removed from the initial pile of 11 toothpicks,
there are now 3 toothpicks remaining.
Since players have removed 1, 3 and 4 toothpicks on turns already, then Chris can only
remove 2 or 5 toothpicks now on his turn, because of rules 2 and 3.
Since there are only 3 toothpicks remaining, Chris must remove 2 toothpicks.
This leaves 1 toothpick in the pile, and the only possible move that Gwen can now make
is to remove 5 toothpicks, which is impossible.
Therefore, Gwen cannot make her turn.
Since Chris was the last player able to move, then Chris wins.
(b) After Gwen has removed 5 toothpicks, there are 5 remaining and Chris can remove 1, 2,
3, or 4 on his turn.
If Chris removes 1, there are 4 remaining and Gwen can remove all of them (since no one
has yet removed 4 toothpicks on a turn). This empties the pile, so Gwen wins.
If Chris removes 4, there is 1 remaining and Gwen can remove all of them (since no one
has yet removed 1 toothpick on a turn). This empties the pile, so Gwen wins.
Therefore, no matter what Chris removes, Gwen can always win the game.
(c) After Gwen has removed 2 toothpicks, there are 7 toothpicks remaining, and Chris can
take 1, 3, 4, or 5 on his turn.
If Chris removes 5 toothpicks, there are 2 remaining and Gwen can remove 1, 3 or 4. Thus,
Gwen must remove 1, leaving 1 toothpick and Chris can remove 3 or 4. He is unable to
make his turn, so Gwen wins. So Chris should not remove 5 toothpicks.
If Chris removes 3 or 4 toothpicks, there are 4 or 3 toothpicks remaning, and Gwen

can remove all of them (since in either case that number of toothpicks hasn’t yet been
removed on a turn), so Gwen wins. So Chris should not remove 3 or 4 toothpicks. (If
Gwen removed 1 toothpick instead of 4 or 3 toothpicks, she would still be guaranteed to
win, since Chris would be unable to go again. Why?)
If Chris removes 1 toothpick, there are 6 remaining and Gwen can remove 3, 4 or 5.
If Gwen now removes 5 toothpicks, there is 1 remaining, and Chris is unable to make
his move, since he can now only remove 3 or 4 toothtpicks. So Gwen wins. Similarly, if
Gwen had removed 4 toothpicks, there would be 2 remaining and Chris cannot remove 1
or 2 since these numbers have already been used, so Gwen wins. If Gwen had removed
3 toothpicks, there would be 3 remaining and Chris cannot remove 1, 2 or 3, since these
numbers have already been used, so Gwen wins.
Thus, regardless of what Chris does on his turn, Gwen will win.
3. (a) Solution 1
Drop a perpendicular from B to X on AC.
Since 4ABC is equilateral, then AB = CB, so X will be the midpoint of AC, so AX = 2.
B
60
A 2 X C
√ √ √ √
By the Pythagorean Theorem, BX = AB 2 − AX 2 =√ 42 − 2√ 2 = 12 = 2 3.
Therefore, the area of 4ABC is 21 (AC)(BX) = 12 (4)(2 3) = 4 3.
Solution 2
◦ ◦ ◦ ◦
Since ∠BAX = √ is perpendicular to AX, then 4BAX is a 30 -60 -90 trian-
√ 60 and BX
gle, so BX = 3AX = 2 3. √ √
Solution 3
√ √
Since ∠BAX = 60◦ , then BX = BA sin(60◦ ) = 4 = 2 3. 2
3
√ √
Solution 4
The area of 4ABC is given by the formula
√ !
1 1 3 √
(AB)(AC) sin(∠BAC) = (4)(4) sin(60◦ ) = 8 =4 3
2 2 2
(b) Since AP = BQ = CR = 1, then AR = BP = CQ = 3, since the side length of 4ABC

is 4.
B
60 1
Q
3
3
P
1
60 60
A 3 R 1 C
Since AP = BQ = CR = 1, P B = QC = RA = 3 and ∠RAP = ∠BP Q = ∠QCR = 60◦ ,

then 4RAP , 4P BQ and 4QCR are all congruent (by side-angle-side). Therefore, the
areas of 4P BQ, 4RAP and 4QCR will all be equal.
Finding the area of any of these three triangles will give us the area of all three, so we
determine the area of 4RAP , because this is easiest to visualize.
Method #1
Drop a perpendicular from P to Y on AR.
P
A 60 R
Y 3
Then, the area of 4RAP is equal to 12 (AR)(P Y ) = 12 (3)(P Y ) = 32 (P Y ), so we need to

find the length of P Y . √ √
◦ ◦
Since ∠RAP = 60 , then P Y = AP sin(60 ) = 1 23 = 23 .
√
3 3
Therefore, the area of 4RAP is 23 (P Y ) = 4
.
Method #2
√
The area of 4RAP is 12 (RA)(AP ) sin(∠RAP ) = 12 (3)(1) sin(60◦ ) = 3 3
4
.
√
3 3
Using either method, we obtain that the area of 4P BQ is 4
.
Lastly, we must determine the area of 4P QR.

Method #1
To do this, we can subtract the combined areas of 4P BQ, 4RAP and 4QCR from the
area of the large triangle, 4ABC.
But the areas of these three triangles are equal (as stated above), and we found the area
of 4ABC in part (a).
√ √ √ √ √
Therefore, the area of 4P QR is 4 3 − 3 3 4 3 = 164 3 − 9 4 3 = 7 4 3 .
Method #2
Since 4RAP , 4P BQ and 4QCR are all congruent, then P Q = QR = RP , so 4P QR
is equilateral.
So, if we can calculate the side length of 4P QR, then we can use a similar method to any
of the methods from (a) to calculate the area of 4P QR.
Using the cosine law in 4RAP , we can find P R:
P R2 = P A2 + AR2 − 2(P A)(AR) cos(∠P AR)

P R2 = 12 + 32 − 2(1)(3) cos(60◦ )

2 1
P R = 10 − 6
2
2
PR = 7
√
so P R = 7. √
7 3
We can then use any of the methods from (a) to determine that the area of 4P QR is 4
.
4. (a) Solution 1
Since the middle number has to be the largest of the three numbers in the triple, then the
only possibilities for b are 3, 4 and 5.
If b = 3, then a and c can only be 1 and 2 or 2 and 1, a total of 2 possible triples.
If b = 4, then a and c can be 1 and 2, 1 and 3, 2 and 3, or their reverses, a total of
6 possible triples.
If b = 5, then a and c can be 1 and 2, 1 and 3, 1 and 4, 2 and 3, 2 and 4, 3 and 4, or their
reverses, a total of 12 possible triples.
Thus, in total there are 20 possible triples.
Solution 2
n
This solution uses the combinatorial notation and n!.
r
three different numbers from the set {1, 2, 3, 4, 5}.
First, we choose
5
There are = 10 ways of doing this.
3
From these three numbers, to form a triple (a, b, c) with a < b and b > c, the middle num-
ber b must be the largest, so there is no choice as to what to put in the middle position.
With the two remaining numbers we can put them in the first and last position in either
order (ie. two possibilities).
Therefore, each choice of 3 different numbers gives us two possible triples, so the total
number of possible triples is 10 × 2 = 20.
(To be totally rigorous, we should also note that we can obtain every such triple in this way,
and that we don’t get any overlap, since we’re choosing three different numbers always,
and we can’t have equal triples coming from two different choices of three numbers.)
(b) Solution 1
Each arrangement of {1, 2, 3, 4, 5, 6} has one number in each of six positions.
If 254 occurs as a block in the arrangement, then the arrangement must be of one of the
forms 254xyz, x254yz xy254z, or xyz254, where x, y and z are 1, 3 and 6 in some order.
With each of these 4 forms, there are 6 ways in which the numbers 1, 3 and 6 can fill the
three remaining places – either 1,3,6 or 1,6,3 or 3,1,6 or 3,6,1 or 6,1,3 or 6,3,1.
Therefore, there are 4 × 6 = 24 arrangements containing 254 consecutively in that order.
Solution 2
We treat 254 as a single block and call it B, say.
Then the arrangements which we are counting correspond to the arrangements of {1, 2, 3, B}.
There are 4! = 24 arrangements of the 4 element set {1, 2, 3, B} (since there are four pos-
sibilities for the first element of the arrangement, and for each of these there are three
possibilities for the second element, and so on).
Therefore, there are 24 arrangements containing 254 consecutively in that order.
(c) Solution 1
To determine the average number of local peaks in all of the arrangements, we count the
total number of local peaks in all 40 320 arrangements and then divide by this total num-
ber of arrangements.
To count the total number of local peaks, instead of looking at the arrangements and
counting the number of local peaks in the arrangements, we look at the possible local
peaks and count the number of arrangements in which each occurs.
In an arrangement of {1, 2, 3, 4, 5, 6, 7, 8}, a local peak is a sequence of three numbers abc
inside the arrangement where a < b and b > c.
How many such sequences of three numbers are there?
This is an extension of part (a). Using either technique from (a), we can determine that
the total number of such sequences is 112.
Now fix one of these 112 sequences abc. In how many of the 40 320 arrangements does this
sequence occur as a block?
This is an extension of part (b). Using either technique from (b), we can determine that
the total number of such arrangements is 6! = 720.
Thus, each of the 112 possible local peak sequences occurs in 720 arrangements, so there
are a total of 112 × 720 = 80 640 local peaks in all possible arrangements.
(We have indeed counted all such local peaks, since every local peak occurs as a sub-
sequence of three numbers in this way.)
80640
Therefore, the average number of local peaks in the 40 320 arrangements is = 2.
40320
Solution 2
In an arrangement of the set {1, 2, 3, 4, 5, 6, 7, 8}, a local peak involves three consecutive
positions, so can occur in one of six places – in positions 1 to 3, 2 to 4, 3 to 5, 4 to 6, 5 to
7, or 6 to 8.
Let’s focus on one of these places, say positions 1 to 3. Our arguments will apply equally
to all such places.
What fraction of all of the arrangements will have local peak in this position?
Choose three numbers a, b, c from the set {1, 2, 3, 4, 5, 6, 7, 8}, say with a < b < c.
There are six possible ways to arrange these three numbers: abc, acb, bac, bca, cab, cba.
Of these six possible ways, two will give a local peak: acb and bca (from the condition
that a < b < c). So 13 of the possible ways to arrange a, b, c give a local peak.
Consider all of the arrangements of {1, 2, 3, 4, 5, 6, 7, 8} whose first three numbers are a, b, c
in some order.
Since the same total number of arrangements begin with abc as begin with acb as begin
with bac as begin with any of the six ways of ordering a, b and c, then exactly 31 of the
arrangements of {1, 2, 3, 4, 5, 6, 7, 8} starting with a, b, c in some order have a local peak
across positions 1 to 3.
Since the number of arrangements with any fixed set of three numbers in positions 1 to 3
is the same, then we can extend our argument to say that exactly 13 of all arrangements
of {1, 2, 3, 4, 5, 6, 7, 8} have a local peak across positions 1 to 3.
This argument applies to any of the 6 possible places in which a local peak can occur.
Therefore, the average number of local peaks in all of the arrangements of {1, 2, 3, 4, 5, 6, 7, 8}
is 6 × 13 = 2.
Canadian
Mathematics
Competition
An activity of The Centre for Education
in Ma thematics and Computing,
University of W aterloo, Wa terloo, Ontario
2004 Sol utions

H ypatia Contest (Grade 11)
© 2004 Waterloo Mathematics Foundation

2004 Hypatia Solutions 2
1. (a) Solution 1
Factoring the given equation,
x 2 + 5x + 6 = 0
( x + 2)( x + 3) = 0
so the roots are x = −2 and x = −3 .
Solution 2
Using the quadratic formula,
−5 ± 52 − 4(1)(6)−5 ± 1
x= =
2(1) 2
−5 + 1 −5 − 1
so x = = −2 or x = = −3 .
2 2
(b) When we increase the roots (–2 and –3) from (a) by 7, we get 5 and 4.
2
A quadratic equation that has 5 and 4 as roots is ( x − 5)(x − 4) = 0 or x − 9x + 20 = 0 .
(c) First, we need to find the roots of this equation.

Since x − 4 is a factor, then x = 4 is a root.
2
So we must find the roots of 3x − x − 2 = 0 . We can do this either by factoring the left
side or by using the quadratic formula.
The easiest way is by factoring. We get 3x 2 − x − 2 = (3x + 2)( x − 1) .
Therefore, the second and third roots are x = − 23 and x = 1 .
When we add 1 to each of the three roots, we obtain 5, 13 and 2, so an equation having
2
( )
these three roots is ( x − 5)(3x − 1)( x − 2 ) = 0 or ( x − 5) 3x − 7x + 2 = 0 or
3 2
3x − 22x + 37x − 10 = 0 . (Of course, there are many other equations that have these
three numbers as roots.)
2. (a) In the diagram, L is the top of the lamp-post, O is the base of

the lamp-post, A is the top of Alan’s head, F is the point on
the ground where Alan is standing, and S is the tip of Alan’s
shadow.
We know that LO and AF are perpendicular to SO, and LAS is
a straight line.
Therefore, we see that triangle LOS is similar to triangle AFS,
since they have a common angle and each has a right angle.
LO AF
Therefore, = , or
SO SF
8 2
=
2+x x
8x = 4 + 2x
6x = 4
2
x= 3
2
Therefore, Alan’s shadow has length 3 m.
(b) In the new diagram, L and O are as before, H is the top of Bobbie’s head, P is the point
on the ground where Bobbie is standing, and T is the tip of Bobbie’s shadow.
As in (a), we have that triangle LOT is similar to triangle HPT, and so, if d is the distance
from the lamp to where Bobbie is standing (ie. the length of OP), then
LO HP
=
TO TP
8 1.5
=
d +3 3
1.5d + 4.5 = 24
1.5d = 19.5
d = 13
Therefore, Bobbie is standing 13 m from the lamp-post.
1 1
3. (a) Since triangle OMN is right-angled, its area is 2 (OM )(ON ) = 2 (8)(6) = 24 .
For the areas of triangles POM, PON and PMN all to be
equal, they each must equal 8, ie. one-third of the total
area.
Consider triangle POM. Its base has length 6 (the length
of OM) and its height has length b (the distance from P to
OM), so its area is 12 (6)(b) = 3b . For the area of triangle
POM to be 8, we must have b = 83 .
Consider now triangle PON. Its base has length 8 (the length of ON) and its height has
length a (the distance from P to ON), so its area is 12 (8)( a) = 4a . For the area of triangle
PON to be 8, we must have a = 2 .
So if P has coordinates (2, 83 ) , then each of triangles POM and PON has area 8, and so
triangle PMN must also have area 8, since the area of the whole triangle OMN is 24.
(b) First, we calculate the area of quadrilateral OMLK.

OMLK is a trapezoid with OM parallel to KL, and so has area equal to the average of the
bases times the height, ie. 12 (10 + 6)t = 8t .
So for the areas of the triangles QOM, QML, QLK, and QKO to be all equal, each of these
four areas must be equal to 2t, since the sum of these four areas is the area of the whole
quadrilateral.
Consider triangle QOM. Its base has
length 6 (the length of OM) and its height
has length d (the distance from Q to OM),
so its area is 12 (6)(d ) = 3d . For the area of
triangle QOM to be 2t, we must have
d = 23 t .
Consider next triangle QLK. Its base has
length 10 (the length of LK) and its height
has length 13 t (the distance from Q to LK,
since Q has y-coordinate 23 t ). Thus, the
area of triangle QLK is 12 (10)( 13 t ) = 53 t ,
which is not
5 1
equal to 2t. (If we tried to set 3 t = 2t , we would then get 3t=0 or t = 0 , which is not
possible since we are told that t > 0 .)
So it is impossible for the areas of both triangles QOM and QLK to be equal to 2t.
Therefore, there is no point Q so that the areas of all four triangles area equal.
4. (a) We solve this problem by considering all of the possible cases. We will use the notation
(G, Y, R) to denote the number of green (G), yellow (Y) and red (R) balls remaining. For
example, the initial position is (1,1,2) .
If the green and yellow are chosen at the beginning, we then get (0,0,3) , so all of the
remaining balls are red.
If the green and a red ball are chosen at the beginning, we then get (0,2,1) , so the next
choice must be one yellow and one red, leaving (1,1,0) , and so the final two balls are
chosen, leaving (0,0,1) , and so the final ball is red.
Similarly, the yellow and a green ball are chosen at the beginning, we then get (2,0,1) , so
the next choice must be one green and one red, leaving (1,1,0) , and so the final two balls
are chosen, leaving (0,0,1) , and so the final ball is red.
Thus, in all cases, the colour of the remaining ball or balls is always red.
(b) Here, we could again proceed by cases, but the number of cases would quickly get very
large, so we should look for a better approach.
€ €
We notice that when a “move” is made (that is, two balls are removed and one is
replaced), the parity of all three colours of balls changes. This is because the number of
balls of each colour is being increased by 1 or decreased by 1, and so changes from odd
to even or even to odd.
Therefore, since all three parities change together, the number of green balls and the
number of red balls must always be both even or both odd, and the number of yellow
balls is of the opposite parity. (Green and red start out both odd, after one move they will
be both even, and so on; yellow starts out even, after one move it will be odd, and so on.)
When the process finally finishes, two of the numbers of balls will be 0, and so both be
even. The only two colours which could have the same parity are green and red, so at the
end there are 0 green and 0 red balls, so the colour of the remaining ball or balls is always
yellow.
(Note that since the number of balls decreases on each turn, we are sure to actually end
the game after at most 11 repetitions of the process.)
(c) Solution 1
In this version of the game, the total number of balls does not change, since the number
of balls removed on each turn (2) is equal to the number of balls replaced on each turn
(2).
Therefore, if the process were to end eventually with only one colour of ball remaining,
then there would be 12 of that colour and 0 of each of the other two colours.
Consider G − Y the difference between the number of balls that are green and the number
of balls that are yellow. If the game got to 12 of one colour and 0 of the other two, then
G − Y would be equal to 12, 0 or –12.
We know that G − Y is initially equal to –1. We will show that no matter what happens
on each step of the process, G − Y will always change by 0, 3 or –3 (ie. a multiple of 3).
Once we have shown this, we can conclude that G − Y can never be equal to 12, 0 or
–12, since we cannot add and subtract 3’s to –1 to get a multiple of 3. This will show that
it is impossible for all of the remaining balls to be the same colour.
Let’s suppose that at some point, the number of green balls is g and the number of yellow
balls is y. After the next turn:
i) if a green and a yellow are replaced by two reds, there are g − 1 green balls and
y − 1 yellow balls, so the difference is (g − 1) − ( y − 1) = g − y , the same as its
previous value
ii) if a green and a red are replaced by two yellows, there are g − 1 green balls and
y + 2 yellow balls, so the difference is (g − 1) − ( y + 2) = g − y − 3
iii) if a yellow and a red are replaced by two greens, there are g + 2 green balls and
y − 1 yellow balls, so the difference is (g + 2) − ( y − 1) = g − y + 3
So the difference always changes by 0, –3 or 3, so starting from –1, G − Y can never
become 0, so it is impossible for all of the balls to be the same colour.
Solution 2
Since all of the balls were eventually all of the same colour, then there would be 12 of
one colour and 0 of each of the two remaining colours (since the number of balls does not
change).
Suppose we were able to reach this state. In this solution, we will list the number of balls
remaining by decreasing size (that is, we won’t worry about which number goes with
which colour). We will show by working backwards from the end, that it is impossible to
get to 5,4,3.
When the process is reversed, we subtract 2 from one of the colours and add 1 to each of
the others.
Thus, 12,0,0 can only come from 10,1,1.
10,1,1 can come from only 8,2,2.
8,2,2 can come from 6,3,3 or 9,3,0.
9,3,0 can come from 7,4,1 or 10,1,1.
7,4,1 can come from 5,5,2 or 8,2,2.
5,5,2 can come from 6,6,0 or 6,3,3.
6,3,3 can come from 4,4,4 or 7,4,1.
This creates a loop of possibilities: 12,0,0; 10,1,1; 9,3,0; 8,2,2; 7,4,1; 6,6,0; 6,3,3; 5,5,2;
4,4,4. It is impossible starting at 12,0,0 to get a position not in this list.
Thus, starting with 5,4,3 it is impossible to get to 12,0,0.

Canadian
Mathematics
Competition
An activity of The Centre for Education
2003 Solutions
Hypatia Contest (Grade 11)
© 2003 Waterloo Mathematics Foundation

1. (a) Let N be the number of tiles in Quentin’s possession.

Since he has 92 extra tiles after forming an n by n square, then
N = n 2 + 92
Since he is 100 tiles short of forming an ( n + 2) by ( n + 2) square, then
N = ( n + 2) − 100
2
To solve for N, the easiest approach is to solve for n first. Since we have two expressions
for N, then
(n + 2) 2 − 100 = n 2 + 92
n 2 + 4 n + 4 − 100 = n 2 + 92
4 n − 96 = 92
n = 47
and thus
N = ( 47) + 92 = 2209 + 92 = 2301
2
Therefore, Quentin has 2301 tiles.
(b) Let B be the total number of blocks in Rufus’ pile.

When Quentin tries to make a cube with 8 blocks along an edge, he is 24 blocks short,
and so he must have taken 8 3 − 24 = 488 blocks.
Let r be the edge length of the cube that Rufus makes with his portion of the blocks.
Since the edge length of Rufus’ original cube is r, he uses r 3 blocks.
When Quentin and Rufus combine their blocks, they can make a cube of edge length
(r + 2) , and so have (r + 2) 3 blocks in total.
Thus,
(r + 2) 3 = r 3 + 488
(r + 2) 2 (r + 2) = r 3 + 488
(r 2 + 4 r + 4)(r + 2) = r 3 + 488
r 3 + 4 r 2 + 4 r + 2 r 2 + 8 r + 8 = r 3 + 488
6 r 2 + 12 r − 480 = 0
r 2 + 2 r − 80 = 0
(r + 10)(r − 8) = 0
Since r is positive, then r = 8 .
Therefore, they have ( r + 2) = 10 3 = 1000 blocks in total.
3
Extension
Let N be the total number of tiles that Quentin has, let x be the side length of the first
square that Quentin tries to make, and let y be the side length of the second square that
Quentin tries to make.
As in (a), since Quentin has 92 extra tiles when he tries to make the first square and is
100 tiles short when he tries to make the second square, then
N = x 2 + 92
and
N = y 2 − 100
Again equating the values of N, we get
x 2 + 92 = y 2 – 106
y 2 – x 2 = 192
y 2 − 100 = x 2 + 92
y 2 − x 2 = 192
( y − x )( y + x ) = 192
So we want to find the number of solutions to this equation where x and y are both
integers with y > x . Thus, y − x and y + x are integers whose product is 192. Note also
that y + x is bigger than y − x . Now, 192 = 2(96) = 2(16)(6) = 2 6 31 .
The divisors of 192 are
1, 2, 3, 4, 6, 8, 12, 16, 24, 32, 48, 64, 96, 192
We should verify that each possibility gives an allowable value for N:
y−x y+x y x N
1 192
2 96 49 47 2301
3 64
4 48 26 22 584
6 32 19 13 269
8 24 16 8 164
12 16 14 2 104
[ ]
(Notice that y = 12 ( y − x ) + ( y + x ) so these equations are not too hard to solve!)
In the cases where y − x is 1 and 3, y is not an integer, which is not allowed.
Therefore, there are 5 possible values for the number of tiles that Quentin has.
Comment
It is also possible to set up the equations in the following fashion
n − 92 = x 2
n + 100 = ( x + p)
2
Subtracting the first equation from the second, we have

192 = ( x + p) − x 2 = 2 xp + p 2 = p(2 x + p)
2
The analysis continues in the same fashion as the first solution in that we must write
192 = 3 ⋅ 2 6 and consider divisors p and 2x + p , where p is the smallest of the two
divisors.
2. (a) Solution 1
First we ask the question: When does a player have a winning move? Since to win a
player must remove the last coin, then a player has a winning move when he or she is
choosing from a position with coins in only one pile (and a second empty pile).
So for Yolanda to win, she wants to ensure that she will always be passed an empty pile
and a non-empty pile at some point. How can she force Xavier to pass her an empty pile?
Xavier can only be forced to empty a pile if he receives two piles both of which have 1
coin (otherwise, he could reduce, but not empty, one of the piles).
So if Yolanda chooses from piles with 1 and 3 coins or 1 and 2 coins, then she can pass
back piles with 1 and 1 coins, and be sure to win.
Thus, Xavier does not want to initially remove 2 coins from one pile, otherwise Yolanda
can follow her strategy above. Also, Xavier does not want to remove 3 coins from one
pile, or Yolanda can immediately win by removing the other 3 coins.
So Xavier should start by removing 1 coin, and passing piles with 2 and 3 coins to
Yolanda. She does not want to pass an empty pile or a pile with 1 coin in it to Xavier (or
he can use her strategy from above), so she removes 1 coin from the larger pile, and
passes back 2 and 2 coins. Xavier is then forced to empty one pile, or reduce one pile to
1 coin, and so Yolanda can then guarantee that she wins.
Therefore, in all cases, Yolanda can guarantee that she wins.
Solution 2
Yolanda will always win the game if she can guarantee that at some point when it is her
turn to choose that she is selecting coins from just one pile. If she is selecting coins from
just one pile, she will win the game by removing all of the coins from that pile.
She can guarantee that this will happen by duplicating Xavier’s move only in the other
pile. Thus, if Xavier takes 1, 2 or 3 coins, then Yolanda will take the same number of
coins from the other pile. This strategy, on Yolanda’s part, will mean that Xavier will
always empty one pile first, and thus guarantee that Yolanda will win.
Solution 3
If Yolanda can ensure that she passes two equal piles to Xavier, then Xavier can never
win, because he can never empty the last pile (he’ll always have two non-empty piles).
So if Xavier reduces to 2 and 3 coins, Yolanda passes back 2 and 2 coins.
If Xavier reduces to 1 and 3 coins, Yolanda passes back 1 and 1 coins.
If Xavier reduces to 0 and 3 coins, Yolanda can immediately win by removing the last 3
coins.
From 1 and 1 coins, Xavier must reduce to 1 and 0 coins, and so Yolanda wins.
From 2 and 2 coins, Xavier must reduce to 1 and 2 coins (allowing Yolanda to pass back
1 and 1 coins) or to 0 and 2 coins, allowing Yolanda to win immediately.
Therefore, Yolanda can always win by following an “equalizing” strategy.
(b) In part (a), we saw that Yolanda always won the game if she could guarantee that Xavier
was choosing when there were two piles with an equal number of coins in each pile.
Starting with piles of 1, 2 and 3 coins, Yolanda can always win, because she can always
after her first turn give two equal piles (and an empty third pile) back to Xavier. We see
this by examining the possibilities:
Xavier’s first move Yolanda’s first move
0,2,3 0,2,2
1,1,3 1,1,0
1,0,3 1,0,1
1,2,2 0,2,2
1,2,1 1,0,1
1,2,0 1,1,0
In any of these cases, Yolanda can be sure to win by following her “equalizing” strategy
from part (a).
So Yolanda’s strategy is to create two equal piles (and a third empty pile) after her first
turn, and so force Xavier to lose, using her strategy from (a).
Extension
In part (b), we saw that if Xavier chooses first from three piles with 1, 2 and 3 coins, then
Yolanda can always win.
In part (a), we saw that if Xavier chooses first from two piles with an equal number of
coins, then Yolanda can again always win.
So on his first move, Xavier does not want to create two equal piles (eg. 2,4,4 or 2,2,5
etc.), otherwise Yolanda would remove the third unequal pile and Xavier would then be
choosing on his second turn from two equal piles.
Similarly, Xavier does not want to create a situation where Yolanda can reduce
immediately to 1,2,3, otherwise Yolanda will win by following the strategy from (b).
So we consider the possible first moves for Xavier:

Xavier’s first move Yolanda’s first move Winner
2,4,4 0,4,4 Yolanda
2,4,3 2,1,3 Yolanda
2,4,2 2,0,2 Yolanda
2,4,1 2,3,1 Yolanda
2,4,0 2,2,0 Yolanda
2,3,5 2,3,1 Yolanda
2,2,5 2,2,0 Yolanda
2,1,5 2,1,3 Yolanda
2,0,5 2,0,2 Yolanda
1,4,5 ?? ??
0,4,5 0,4,4 Yolanda
So if Xavier makes any move other than to 1,4,5, Yolanda will win by following the
correct strategy.
What if Xavier moves to 1,4,5? There are then 10 possible moves for Yolanda.
As above, if Yolanda makes her first move to 0,4,5 or 1,1,5 or 1,0,5 or 1,4,4 or 1,4,1 or
1,4,0, then Xavier can reduce to two equal piles.
If Yolanda reduces to 1,3,5 or 1,2,5 or 1,4,3 or 1,4,2, then Xavier can reduce to some
ordering of 1,2,3, and so Xavier can win.
Therefore, Xavier can win by reducing first to 1,4,5, and then to either two equal piles or
some ordering of 1,2,3, and then following Yolanda’s strategy from (a) or (b).
3. We slice the two solids with a vertical plane through the vertex of the cone and highest
point of the sphere. The cross-sections we get then are an isosceles triangle with a height
of 10 cm and a base of 10 cm, and a circle of diameter 10 cm.
We want to find the height, h, that produces equal cross-sectional areas. Since the
formula for the area of a circle is πr 2 and we want to find h, we need to find a
relationship between h and r for the cone and for the sphere.
Case 1 – Cone
From V, the vertex of the cone, we draw in the principal axis of the cone as shown. It is
clear by symmetry that triangle VPQ is similar to triangle VRS.
Thus, V
10 10 − h
=
5 r 10 – h
r
2 r = 10 − h S
r = 2 (10 − h )
1 R h
P Q
Case 2 – Sphere
From O, the centre of the sphere, we draw a M P
line through the centre of the cross-sectional O 5–h
5
circle. This line is perpendicular to the circle. 5–h 5 O
We have OP = 5 (the radius of the circle), M P 5
h
DO = 5 , DM = h , and MP = r .
We need an expression for r – the answer D D
depends on whether h is bigger or smaller
than 5.
If h is bigger than or equal to 5, then OM = h − 5 . If h is smaller than 5, then OM = 5 − h .
Thus, by Pythagoras,
( )
r = OP 2 − OM 2 = 5 2 − (5 − h ) = 25 − h 2 − 10 h + 25 = 10 h − h 2
2
(Notice that even if OM = h − 5 , then we get the same answer, since ( h − 5) = (5 − h ) .)

2 2
So, we compare the areas of the cross-sectional circles to solve for h:

[ (10 − h)] = π  10h − h 2 

2 2
π 1
2  
(10 − h) 2 = 4(10h − h 2 )
100 − 20 h + h 2 = 40 h − 4 h 2
5 h 2 − 60 h + 100 = 0
h 2 − 12 h + 20 = 0
(h − 10)(h − 2) = 0
and therefore, h = 10 or h = 2 .
(Note that h = 10 gives a horizontal plane that just passes through the vertex of the cone
and is tangent to the top of the sphere.)
Therefore, the height is 10 cm or 2 cm.
Extension
To avoid fractions, we start by letting d = 2 R , so the cone has radius R and height R.
As above, we slice the two solids with a vertical plane.
R− h r
In the cone, at a height of h, we have = or r = R − h as in the previous set-up.
R R
In the sphere, at a height of h we have
( )
r = OP 2 − OM 2 = R 2 − ( R − h ) = R 2 − h 2 − 2 Rh + R 2 = 2 hR − h 2
2
Note that h ≤ R since the cone has height R.

Therefore, the sum of the areas of the two circular V
cross-sections is
2
π  2 hR − h 2  + π ( R − h )
2 R–h O
r R
S R–h
( ) (
= π 2 hR − h 2 + π R 2 − 2 hR + h 2 ) h
M
h
P
= πR 2 P Q D
which is a constant, since R is a constant.
4. (a) Solution
The visible area of P (2,−6) is the area of ∆ABP . y
∆ABP has base AB of length 4, and its height is the
distance from AB to the point P, which is 10, since AB is D C
parallel to the x-axis.
Thus, the area of ∆ABP is 12 bh = 12 ( 4 )(10) = 20 square
A 4
units, ie. the visible area of P is 20 square units. B
10
x
(b) Solution
The visible area of Q(11, 0) is the sum of the areas of y
∆QBC and ∆QBA .
∆QBC has base BC of length 4, and its height is the
distance from Q to the line through B and C, which is 6. D C
Thus, the area of ∆QBC is 12 bh = 12 ( 4 )(6) = 12 square
4
units.
4 B
∆QBA has base BA of length 4, and its height is the A
distance from Q to the line through B and A, which is
4
also 4. Thus, the area of ∆QBA is 12 bh = 12 ( 4 )( 4 ) = 8
x
square units. 6 Q
So the visible area of Q is the sum of these two areas, or
20 square units.
(c) Solution
From any point P, there are either 2 or 3 visible y
vertices of the square.
We first consider those points P for which
there are 2 visible vertices.
Geometrically, these will be points which lie
(ii)
“directly opposite” an edge of the square.
That is, they will be point P which
i) lie below the square, with x-coordinate (iii) (iv)
between 1 and 5, inclusive
ii) lie above the square, with x-coordinate
between 1 and 5, inclusive
iii) lie to the left of the square, with y- x
coordinate between 4 and 8, inclusive (i)
iv) lie to the right of the square, with y-
coordinate between 4 and 8, inclusive
In each of these four cases, the visible area will
be a single triangle whose base is a side of the
square (that is, of length 4). For the visible
area of one of these points P to be 20, the
height of this triangle must be 10.
Thus, for those points in case (i), all points P y

which lie 10 units below the square, and have
x-coordinates between 1 and 5 will be in the
20/20 set. In other words, the points P lying
on the line segment joining (1,−6) to (5,−6)
will lie in the 20/20 set. (Notice that this
includes the point (2,−6) from (a).) These are
all of the points in this region that lie in the
20/20 set. So in this region, the 20/20 set has
length 4.
x
(1, – 6) (5, – 6)
The other three regions will give the same result, by symmetry, so we have four line
segments, each of length 4, in the 20/20 set thus far.
There are four more regions to consider – that is, the regions that do not lie directly
opposite an edge of the square (for example, the region of points P ( x, y ) where x ≥ 5 and
y ≤ 4 ). By symmetry each of these regions should give us the same result.
So suppose that P ( x, y ) in the region x ≥ 5 and y ≤ 4 has a visible area of 20. Then the
sum of the areas of ∆PBC and ∆PBA is 20.
∆PBC has a base of length 4 and a height of y
x − 5.
1
∆PBA has a base of length 4 and a height of
4 − y.
1 4
4
(x, y)
1 (5, – 6)
Therefore, for P to be in the 20/20 set, we have

1
2
(4)( x − 5) + 12 (4)(4 − y ) = 20
x − 5 + 4 − y = 10
y = x − 11
ie. P lies on the straight line of slope 1, y = x − 11. (Notice that this includes the point
(11, 0) from (b).) So in this region, the points in the 20/20 set are the points on the line
segment joining (5,−6) to (15, 4 ) , which has length 10 2 . (It is worth noting that the
endpoints of this line segment are the endpoints of the line segments from regions (i) and
(iv) above.)
We can argue by symmetry that the other sides in the required octagon will be as shown
in the diagram.
So the 20/20 is a polygon (in fact, an octagon) with four sides of length 4 and four sides
of length 10 2 . Therefore, the perimeter of the 20/20 set is 16 + 40 2 .
10 2 10 2
1 1
10 2 10 2
Extension
We approach this in an analogous manner to the 2-dimensional case.
Case 1: 4 visible vertices

Where will a point P lie so that it has exactly 4 visible vertices?
In this case, P will lie directly opposite one of the 6 faces of the cube.
The visible volume will be formed by one square-based pyramid.
The base of this pyramid will be a unit square, so since the volume of a pyramid is one-
third times the area of the base times the height, then for a volume of 20, the height of the
pyramid must be 60.
So the points P above one of the faces which have a visible volume of 20 are all those
points 60 units above the faces. In other words, the points form a square (again, of side
length 1) which is 60 units above the face. So the 20/20 set has six square faces each of
area 1.

Where will a point P lie so that it has exactly 6 visible vertices?
In this case, the 6 visible vertices are the 6 vertices of two faces which share a common
edge, and P lies in the region which is between the outer edges of these two faces, but not
directly above either face. We say that P lies above the edge. Since the cube has 12
edges, there will be 12 such regions.
Consider points P in one of these regions. The visible volume here will be formed by
two square-based pyramids, each of which has a base which is a unit square. So the
visible volume will be one-third times the area of the base (which is 1) times the sum of
the two heights. Since the visible volume should be 20, the sum of the two heights will
be 60.
The set of points above the edge which give a combined height of 60 for the two
pyramids will form a rectangle which joins the edges of the two squares over the two
adjacent faces. This rectangle will thus have one edge of length 1 and the other edge of
length 60 2 (since the long edge will form the hypotenuse of a right-angled triangle
with two legs each of length 60).
(Why does this give us a portion of a plane? If we consider points P which lie in a cross-
section which is perpendicular to the two adjacent faces, then we are asking for P to have
sum of the heights to two perpendicular line segments to be 60, which gives a straight
line as in (b). Sliding this cross-section across the faces gives us a segment of a plane.)
Thus, we have twelve rectangular faces, each of area 60 2 .

Where will a point P have 7 visible vertices?
P will have 7 visible vertices if it lies in a position that has not yet been considered. That
is, P will in essence be above one of the 8 vertices of the cube. (This seems sensible
since the 20/20 set currently has 8 holes in it!)
This time, the visible volume is made up of three square-based pyramids, so we want
points P so the sum of the heights of these three pyramids is 60, as before.
This region will again form part of a plane, and it should connect to the three rectangles
already formed over the three edges that meet at the vertex in question (since at the edge
of each of these rectangles, we can consider having a third pyramid of height 0).
So this segment of the 20/20 set is an equilateral triangle of side length 60 2 .
To find the area of this triangle, we join the top vertex to the midpoint of the base, so
each half of the triangle is a 30-60-90 right-angled with a leg of length 30 2 opposite
( )
the 30o angle. Thus, the height of the triangle is 3 30 2 = 30 6 , and so the area of
the whole equilateral triangle is 1
2 (60 2 )(30 6 ) = 900 12 = 1800 3 .
60 2
60 2
60 2
60 2
1
60 2
1 1
A partial sketch of the desired region.

Therefore, the total surface area of the 20/20 set is
( ) ( )
6(1) + 12 60 2 + 8 1800 3 = 6 + 720 2 + 14400 3

Hypatia Combined Solutions

Uploaded by

Copyright:

Available Formats

Hypatia Combined Solutions

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hypatia Combined Solutions

Uploaded by

Copyright:

Available Formats

The CENTRE for EDUCATION

in MATHEMATICS and COMPUTING

2023 Hypatia Contest

Wednesday, April 5, 2023

Thursday, April 6, 2023

©2023 University of Waterloo

1. (a) Jasmin’s total score was 3 · 2 + 4 · 5 = 6 + 20 or 26 points.

3111m + 1311n + 17 970 = 10 100n + 1010m + 7

Consider the next two Venn diagrams shown below.

P(B) P(A and B)

whose product is divisible by 5.

32023 + 52023 − 32023 + 62023 − 32023 = 52023 + 62023 − 32023

Thus, the probability, p, that the product is not divisible by 6 is

52023 + 62023 − 32023

2022 Hypatia Contest

Tuesday, April 12, 2022

Wednesday, April 13, 2022

©2022 University of Waterloo

OM = 3x (since OM > 0). C D

(a2 − 12a + 39) − (b2 − 12b + 39) = b−a

Since a 6= b, then a − b 6= 0 and so a + b − 11 = 0 or b = 11 − a.

4. (a) Written as a product of its prime factors, 240 = 24 31 51 and so

f (240) = (1 + 4)(1 + 1)(1 + 1) = (5)(2)(2) = 20

(b) Suppose f (N ) = 6 and N is refactorable. Then 6 divides N .

2 215 315 < 231 37 < 263 33 < 2127 31

3 215 33 53 < 215 37 51 < 231 33 51 < 263 31 51

4 215 33 51 71 < 231 31 51 71

2021 Hypatia Contest

©2021 University of Waterloo

1. (a) The total cost to rent a car is $180.00.

2. (a) Trapezoid ABCD is drawn, as shown. y

3. (a) Since an = 2n for n ≥ 1, then a5 = 25 = 32.

= 13 (a1 + a2 + a3 + · · · + an ) − 31 ((−1) + (−1)2 + (−1)3 + · · · + (−1)n )

Next, we consider each of the two expressions within parentheses, separately.

and simplifying, we get 

4. (a) In 4XY Z, x = 20, y = 21, and ∠XZY = 90◦ , Y

(c) Since A = 3P , we get 12 xy = 3(x + y + z), and so xy = 6(x + y + z).

did, we could have chosen to rewrite it as

x(y − 12) = 12y − 72

(1, 8k 2 ), (2, 4k 2 ), (4, 2k 2 ), (8, k 2 ), (k, 8k), (2k, 4k)

and the negative factor pairs of 8k 2 are

If for example x − 4k = −1 and y − 4k = −8k 2 , then y = 4k − 8k 2 which is less than zero

2xy = 5102 − 1020z (since x2 + y 2 = z 2 )

From the first factor pair, we get x − 4k = 1 and y − 4k = 8k 2 , and so

2020 Hypatia Contest

Wednesday, April 15, 2020

Thursday, April 16, 2020

©2020 University of Waterloo

1. (a) The cost of 12 bags of avocados is $5.00 × 12 = $60.00.

$4.50 × 8 + $10.00 × 4 = $36.00 + $40.00 = $76.00

$4.50 × 8 + $10.00 × 4 = $36.00 + $40.00 = $76.00

In this case, the chef has 5 × 6 = 30 avocados and 4 × 15 = 60 mangoes.

3. (a) For n ≥ 3 and k ≥ 0, the value of T (n, k) is constant

T(4, 0) = 2 T(4, 1) = 4 T(4, 2) = 6 T(4, 3) = 8

xi+1 = (xi )2 + 1 = (3k)2 + 1 = 3(3k 2 ) + 1

which is 1 more than a multiple of 3.

7k + 2 49k 2 + 28k + 5 = 7(7k 2 + 4k) + 5

From this table, we make the following observations:

2, 7, 8, 12, 13, 22, 23, 27, 28, 33

2019 Hypatia Contest