CEMC Banner

2026 Hypatia Contest
Solutions
(Grade 11)

Wednesday, April 1, 2026
(in North America and South America)

Thursday, April 2, 2026
(outside of North American and South America)

©2026 University of Waterloo


    1. Since \(A\), \(B\) and \(C\) are collinear, segments \(AB\) and \(AC\) (and \(BC\)) have the same slope.
      The slope of segment \(AB\) is \(\dfrac{5-2}{2-1} = 3\). The slope of segment \(AC\) is \(\dfrac{c-2}{3-1} = \dfrac{c-2}{2}\).
      Therefore, we have that \(3 = \dfrac{c-2}{2}\). Solving for \(c\) gives us that \(c=8\).

    2. Since segment \(DF\) has slope \(\dfrac{7-0}{0-14} = - \dfrac{1}{2}\) and \(y\)-intercept \(7\), the equation of the line
      through points \(D\) and \(F\) is \(y=-\dfrac{1}{2}x + 7\).
      The coordinates \((x,y)\) of \(E\) must be positive integers and \(E\) must lie on the line \(y=-\dfrac{1}{2}x + 7\). When \(x \geq 14\) we have that \(y \leq 0\), and when \(x < 14\) we have that \(y>0\). Therefore \(x\) and \(y\) are both positive exactly when \(0<x<14\). Also, \(y\) is an integer exactly when \(x\) is even.
      Therefore the number of possible points \(E\) is the number of integers \(0 < x < 14\) such that \(x\) is even. There are \(6\) such integers which give the following \(6\) possibilities for \(E\): \((2,6)\), \((4,5)\), \((6,4)\), \((8,3)\), \((10,2)\), and \((12,1)\).

    3. The slope of segment \(PQ\) is \(\dfrac{n-4-12}{6-15}=\dfrac{n-16}{-9}\).
      The slope of segment \(PR\) is \(\dfrac{n-12}{18-15}=\dfrac{n-12}{3}\).
      Therefore, points \(P\), \(Q\) and \(R\) are collinear exactly when \(\dfrac{n-16}{-9}=\dfrac{n-12}{3}\). This is equivalent to \(3n-48 = -9n+108\), which is equivalent to \(12n=156\).
      Therefore, the value of \(n\) that makes \(P\), \(Q\) and \(R\) collinear is \(n=\dfrac{156}{12}=13\).

    1. Since \(0 < A < B < C < 10\), the largest possible value of \(A\) is \(7\), and when \(A=7\), we must have \(B=8\) and \(C=9\).
      Therefore the largest possible peak number is \(78\, 987\).
      Similarly, the smallest possible peak number is \(12\, 321\).
      Therefore the positive difference is \(78\, 987-12\,321 = 66\, 666\).

    2. Since \(36\,245 < ABCBA < 45\,932\), we have \(A=4\) or \(A=3\).
      If \(A=4\), then the first inequality is true for all possible values of \(B\) and \(C\), so we only need to consider \(4BCB4 < 45\,932\). We must then have \(B=5\), and then \(C\) can be any of \(6\), \(7\), or \(8\), so there are \(3\) peak numbers when \(A=4\). (They are \(45\, 654\), \(45 \, 754\), \(45 \, 854\).)
      If \(A=3\), then the second inequality is true for all possible values of \(B\) and \(C\), so we only need to consider \(36\,245 < 3BCB3\). Then \(B\) can be \(6\), in which case \(7 \leq C \leq 9\), or \(B = 7\), in which case \(8 \leq C \leq 9\), or \(B=8\), in which case we have that \(C=9\). Altogether, there are \(6\) peak numbers when \(A=3\). (They are \(36\,763\), \(36\,863\), \(36\,963\), \(37\,873\), \(37\,973\), \(38\,983\).)
      Putting both cases together, there are a total of \(3+6 = 9\) peak numbers satisfying the desired inequalities.

    3. A number is a multiple of \(15\) exactly when it is a multiple of both \(5\) and \(3\), and a number is a multiple of \(5\) exactly when it has units digit \(0\) or \(5\). Since \(0<A<10\) and \(ABCBA\) is a multiple of \(15\), we must have \(A=5\).
      Furthermore, \(5BCB5\) is a multiple of \(15\) exactly when \(5BCB5\) is a multiple of \(3\).
      A positive integer is a multiple of \(3\) exactly when the sum of its digits is a multiple of \(3\). That is, \(5BCB5\) is a multiple of \(3\) exactly when \(10 + 2B + C\) is a multiple of \(3\).
      Since \(A=5\) and \(A<B<C\), the possible values of \(B\) are \(6\), \(7\) and \(8\).
      When \(B=6\), \(10+12+C\) is a multiple of \(3\) precisely when \(C=8\) (among the possible values \(7\), \(8\), \(9\) of \(C\)).
      When \(B=7\), \(10+14+C\) is a multiple of \(3\) precisely when \(C=9\).
      Finally, when \(B=8\), the only possible value of \(C\) is 9, and \(10+16+9\) is not a multiple of \(3\).
      Therefore, the peak numbers that are a multiple of \(15\) are \(56\,865\) and \(57\,975\).

    1. Since \(DC\) is a diameter of the circle with radius \(36\), we have that \(DC = 36(2) = 72\). Similarly, \(CE= 10(2) = 20\).
      By Fact 1, we have that \(DE = CD - CE = 72-20 = 52\).

    2. Extend segment \(FG\) to \(K\), noting that these three points lie on the same line by Fact 1. Also draw in segment \(JG\). These constructions are shown in the diagram below.

      We have that \(FG = FK - GK = 49-12= 37\). By Fact 2, we have that \(\angle GJF = 90\degree\). By the Pythagorean Theorem \(FG^2 = JG^2 + FJ^2\). That is, \(37^2 = 12^2 + FJ^2\). Therefore \(FJ^2 = 1225\), and so \(FJ=35\) (since \(FJ>0\)).

    3. We will continue to use Facts 1 and 2, but we will no longer cite them each time.
      First draw in line segment \(ZL\) (noting it passes through \(X\)), and then draw in segment \(XM\). These constructions are shown below.

      We have that \(XZ = LZ - LX = 18 - 5 = 13\). Since \(XM=5\), by the Pythagorean Theorem we deduce that \(MZ=12\). With this result, erase line segment \(ZL\) from the above diagram and draw in line segment \(XY\), giving us the diagram below.

      Note that \(18 = ZV = ZY+ YV = ZY+r\) and therefore \(ZY = 18-r\). Therefore \(MY = MZ + ZY = 12 + (18-r) = 30-r\).
      Note that \(XY= XN+NY = 5 + r\). Now apply the Pythagorean Theorem to \(\triangle XMY\). We have the following equivalent equations: \[\begin{align*} XM^2 + MY^2 &= XY^2 \\ 5^2 + (30-r)^2 &= (5+r)^2 \\ 25 + 900 - 60r + r^2 &= 25 + 10r + r^2 \\ 70r &= 900 \\ r &= \dfrac{90}{7}\end{align*}\] which is our final answer.

    1. From \((\ast)\) we have that \(y^2 - z^2 = (m+2)^2 - (m+1)^2\) and \(x^2 - y^2 = (m+1)^2 - m^2\). We then have that \[\begin{align*} k &= (y^2 - z^2 ) - (x^2 - y^2) \\ &= ((m+2)^2 - (m+1)^2) - ( (m+1)^2 - m^2) \\ &= (m^2 + 4m + 4 - m^2 - 2m -1) - ( m^2 + 2m + 1 - m^2) \\ &= (2m+3) - (2m + 1) \\ &= 2\end{align*}\] We showed that \(y^2-z^2=2m+3\) and \(x^2-y^2 = 2m+1\) and since \(m\) is a positive integer it follows that \(z<y<x\), which we will use in part (b).

    2. We begin with a useful fact: the sum of the first \(n\) odd positive integers is equal to \(n^2\). That is, \(1+3+5+\cdots+(2n-1)=n^2\).
      Note: Contestants were not required to justify this, so we save a justification until the end.
      We are given that \(y^2 - z^2\) is the sum of \(p\) consecutive odd integers, the largest of which is \(2y-1\). Therefore, \[\begin{align*} y^2 - z^2 &= \overbrace{(2y-1) + (2y-3) + (2y-5) + \cdots + (2y - (2p-1))}^{\text{p terms}} \\ &= 2yp - (1+3+5 + \cdots + (2p-1)) \\ &= 2yp - p^2\end{align*}\] where we used the useful fact in the final step. Therefore \(z^2 = y^2 - 2yp + p^2 = (y-p)^2\), so \(z=y-p\) or \(-z = y-p\). We will show in fact that \(y-z=p\) (and thus \(z=y-p\)). To see this, note that \(y^2 = 1 + 3 + 5 + \cdots + (2y-1)\) expresses \(y^2\) as a sum of \(y\) odd integers, and \(z^2 = 1 + 3 + 5 + \cdots + (2z-1)\) expresses \(z^2\) as a sum of \(z\) odd integers, and therefore \[\begin{align*} y^2 - z^2 &= 1 + 3 + 5 + \cdots + (2y-1) - (1 + 3 + 5 + \cdots + (2z-1)) \\ &= \text{sum of $y-z$ consecutive odd integers, the largest of which is $2y-1$}\\ & \text{\ \ \ \ (since $y>z$ as shown in part (a))}\end{align*}\] But \(y^2 - z^2\) is also the sum of \(p\) consecutive odd integers, the largest of which is \(2y-1\), and so \(y-z=p\). Similar calculations yield \(x^2 - y^2 = 2xq - q^2\) and \(x-y=q\) or \(x=y+q\). We now finish this question with a series of equivalent equations. \[\begin{align*} y^2 - z^2 &= x^2 - y^2 + 2 & \text{(from part (a))}\\ 2yp - p^2 &= 2xq - q^2 + 2\\ 2yp - p^2 &= 2(y+q)q - q^2 + 2\\ 2yp - p^2 &= 2yq +2q^2 - q^2 + 2 \\ 2y (p-q) &= p^2 + q^2 + 2 \\ y &= \dfrac{p^2 + q^2 + 2}{2(p-q)}\end{align*}\] For the final step, note that \(p-q \neq 0\), since if \(p=q\), the second to last equation would give \(0=p^2 + p^2 + 2\). This is equivalent to \(p^2 = -1\), which is not possible for an integer \(p\).
      We have obtained an expression for \(y\) in terms of \(p\) and \(q\).

    3. We just established that \(y = \dfrac{p^2+q^2+2}{2(p-q)}\). We begin by finding some values of \(p\) and \(q\) for which \(y\) is a positive integer. If \(p-q\) is odd, then one of \(p\), \(q\) is even and the other is odd, and then \(p^2 + q^2 + 2\) is odd. Since the denominator \(2(p-q)\) is even, it follows that \(y\) is not an integer when \(p-q\) is odd. Therefore \(p-q\) is even.
      Suppose \(p-q=2\). Then \(p=q+2\) and \[y = \dfrac{p^2 + q^2+2}{2(p-q)} = \dfrac{(q+2)^2 + q^2+2}{2((q+2)-q)} = \dfrac{q^2 + 4q + 4 + q^2 + 2}{4} = \dfrac{q^2+2q + 3}{2}\] If \(q\) is even, then the numerator is odd, and therefore \(y\) is not an integer. Thus we must consider odd values of \(q\). From (b) we know that \(z=y-p = y-(q+2)\) and \(x=y+q\). Furthermore, we have that \(m^2 + x^2 = (m+1)^2 + y^2\). Solving for \(m\) (the \(m^2\) terms subtract out) yields that \(m= \frac{1}{2}(x^2-y^2 - 1)\). In the table below, for some odd values of \(q\), we determine \(x\), \(y\), \(z\), and \(m\):

      \(q\) \(x\) \(y\) \(z\) \(m\)
      \(1\) \(4\) \(3\) \(0\) \(3\)
      \(3\) \(12\) \(9\) \(4\) \(31\)
      \(5\) \(24\) \(19\) \(12\) \(107\)

      When \(q=1\), we have that \(3^2+4^2 = 4^2 + 3^2 = 5^2 + 0^2\), but we reject \(S=25\) since we require \(z>0\).
      When \(q=3\), we have that \(31^2+12^2 = 32^2 + 9^2 = 33^2 + 4^2 = 1105\).
      When \(q=5\), we have that \(107^2 + 24^2 = 108^2+19^2 =109^2 + 12^2 = 12\, 025\).
      Therefore \(1105\) and \(12\, 025\) are two Sorensen numbers.

      Editorial Comment: It turns out that \(1105\) and \(12 \,025\) are the smallest two Sorensen numbers and every odd \(q \geq 3\) results in a corresponding Sorensen number.

      In part (b), we stated and used a useful fact that we will now restate and prove for completeness.

      Useful Fact: The sum of the first \(n\) odd positive integers is equal to \(n^2\).
      Using the more familiar fact that \[1+2+3+\cdots+n = \dfrac{n(n+1)}{2}\] we have that \[\begin{align*} 1+3+5+\cdots+(2n-1) + n &= (1+1)+(3+1)+(5+1)+\cdots+(2n-1+1 )\\ &= 2 + 4 + 6 + \cdots 2n \\ &= 2 [1+2+3+ \cdots + n] \\ &= 2 \dfrac{n(n+1)}{2} \\ &= n^2 + n \end{align*}\] Subtracting \(n\) from both sides results in the desired identity, \(1+3+5+ \cdots + (2n-1) = n^2\).