Mystery Spot

Point E E lies within rectangle A B C D , ABCD, as shown below.

If the distances from the vertices to point E E are all distinct integers, what is the least possible value of A E + B E + C E + D E ? AE + BE + CE + DE?


The answer is 20.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

2 solutions

Relevant wiki: British Flag Theorem

According to the British Flag Theorem , if E E is a point within the rectangle A B C D ABCD , then A E 2 + C E 2 = B E 2 + D E 2 AE^2 + CE^2 = BE^2 + DE^2 .

Let n = A E 2 + C E 2 = B E 2 + D E 2 n = AE^2 + CE^2 = BE^2 + DE^2 for some integer n n . We will then attempt to find the least possible natural number n n that can be written as the sum of two squares in two different ways.

Then for some integers a a and b b , a 2 + b 2 = ( a + b i ) ( a b i ) a^2 + b^2 = (a + bi)(a - bi) , which is the product of Gaussian integers or z 2 |z|^2 , and when plotting the Gaussian integers z z over the complex plane, we can draw a Gaussian circle of radius z |z| as the following examples:

The number of lattice points of z z on the circumference will then equal to the number of sums of squares for some natural number n = z 2 n = |z|^2 .

According to Sum of 2 Squares theorem, n n can be written as the sum of two squares if & only if n n only has prime factors p = 4 m + 1 p = 4m + 1 for some integer m m .

Now when expressing n n as prime factorization, n = 2 a 0 p 1 a 1 p i a i q 1 b 1 q j b j n = 2^{a_{0}}\cdot p_{1}^{a_{1}} \cdots p_{i}^{a_{i}} \cdot q_{1}^{b_{1}} \cdots q_{j}^{b_{j}} , where p p is a prime factor in a form of 4 m + 3 4m + 3 and q q is a prime factor in a form of 4 m + 1 4m + 1 .

If any p p has any odd number, the remainder will be 3 3 in modulus 4 4 , making the sum of squares impossible.

On the other hand, considering the powers of q q , the combination for each q q can be chosen from 0 0 to b j b_{j} . The overall combination B = ( b 1 + 1 ) ( b 2 + 1 ) ( b j + 1 ) B = (b_{1} + 1)(b_{2} + 1)\cdots (b_{j} + 1) .

Considering only positive signs over the first quadrant in the complex plane, the number of sum of squares will then equal to B 2 \dfrac{B}{2} as reflection over the line real part = imaginary part.

Altogether, the function for number of sums of squares, r 2 ( n ) r'_{2}(n) can be formulated as:

r 2 ( n ) = { 0 if a i is odd B 2 if a i is even and B is even B ( 1 ) a 0 2 if a i is even and B is odd r'_{2}(n) = \begin{cases} 0 & \quad \text{if} a_{i} \text{is odd} \\ \dfrac{B}{2} & \quad \text{if } a_{i} \text{is even and}B \text{ is even}\\ \dfrac{B - (-1)^{a_{0}}}{2} & \quad \text{if }a_{i} \text{is even and} B \text{ is odd}\\ \end{cases}

Hence, our desired r n = 2 r'_{n} = 2 . For the second scenario, where B B is even, B = 4 B = 4 . In this case, B = ( 1 + 1 ) ( 1 + 1 ) B = (1+1)(1+1) or B = 3 + 1 B = 3+1 . In other words, the number n n can be the product of two different primes or the cube of one prime.

For prime q = 4 m + 1 q = 4m +1 , the list involves: {5, 13, 17, ...}.

For lowest value possible, then n = 5 13 = 65 n = 5\cdot 13 = 65 while n = 5 3 = 125 n = 5^3 =125 , which is not the least number.

Now for the third scenario, where B B is odd, B ( 1 ) a 0 = 4 B - (-1)^{a_{0}} = 4 . Thus, B = 3 = 2 + 1 B = 3 = 2+1 and a 0 = 1 a_{0} = 1 . Then n = 2 q 2 n = 2q^2 , and the least value would be n = 2 5 2 = 50 n = 2\cdot 5^2 = 50 .

Finally, from both scenarios, we can show that these numbers can really be written as two sums of two squares:

50 = 1 2 + 7 2 = 5 2 + 5 2 50 = 1^2 + 7^2 = 5^2 + 5^2

65 = 1 2 + 8 2 = 4 2 + 7 2 65 = 1^2 + 8^2 = 4^2 + 7^2

However, since the question specifically asks for different lengths within the rectangle, the only sum of the least possible four lengths = 1 + 8 + 4 + 7 = 20 1 + 8 + 4 + 7 = \boxed{20} .

Shouldn't you check that a rectangle actually exists for which a point a distance of 1, 4, 7, 8 from its vertices actually can be found. This is not difficult, but nonetheless important.

Mark Hennings - 4 years, 8 months ago

Fermat's theorem is valid only if n is prime, doesn't this detail invalidate the demonstration ?

Samuel Coron - 2 years, 10 months ago

Log in to reply

Ok. Let's generalize as Sum of 2 Squares Theorem then.

Worranat Pakornrat - 2 years, 10 months ago

Log in to reply

I didn't know it, actually, thanks

Samuel Coron - 2 years, 10 months ago
Ritabrata Roy
Jul 31, 2018

It's simple to prove AE^2+CE^2=BE^2+DE^2

However the question is asked to find the distinct a,b,c,d such that a^2+b^2=c^2+d^2

One can use the well-known 15 for 15=4^2-1^2=8^2-7^2

Therefore 8^2+1^2=4^2+7^2 (by trial it is easy to prove that this is the minimum case)

Answer is 20

This jumped in difficulty pretty quickly, wow

Yash Permalla - 11 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...