Using 2 triangles each with base of 8 and height of 3, and 2 trapezoids with heights of 3 on top, 5 on bottom and height of 5, these four figures can create an area with 64 units squared. However, when rearranged as a rectangle with 13 x 5=65, one additional unit squared seemed to have been created. How is this possible?
Asked
Active
Viewed 622 times
3

4This is the [missing square puzzle](https://en.wikipedia.org/wiki/Missing_square_puzzle). – José Carlos Santos Sep 02 '17 at 00:04

so the second figure has triangles slightly bigger than the triangles in figure one? – Goodwin Lu Sep 02 '17 at 00:06

also, unlike the wikipedia article, the hypothenuse is completely and utterly straight: http://prntscr.com/gg199b – Goodwin Lu Sep 02 '17 at 00:10

@GoodwinLu then it doesn't work, the truth is it's not possible it's an illusion. – Sep 02 '17 at 00:14

1According to the figure on the right, $\frac25=\frac38$. Who knew? – G Tony Jacobs Sep 02 '17 at 01:41
3 Answers
13
Here's a slightly less subtle "demonstration" that a rectangle with area 5 can be rearranged into a rectangle with area 6.
grand_chat
 32,010
 1
 28
 54
6
This is a classic illusion based on the Fibonacci number identity $$ 13 \times 5 = 1 + 8 \times 8 . $$
The "diagonal" of the rectangle isn't one. The slopes on each segment don't agree. There's one unit of area between the "diagonals".
Ethan Bolker
 80,490
 6
 95
 173
3
unlike the wikipedia article, the hypothenuse is completely and utterly straight
No, it's not. Consider the bottomleft corner of the rectangle.
 Let $\alpha$ be the angle in the yellow triangle, then $\tan \alpha = 3/8\,$.
 Let $\beta$ be the angle in the green trapezoid, then $\tan \beta = 5/(53)=5/2\,$.
But then $\,\tan \alpha \tan \beta = 15 / 16 \ne 1\,$, so $\,\alpha+\beta \ne 90^\circ\,$ i.e. the two angles do not add up to a right angle. The slopes of the two hypotenuses differ by $\,90^\circ  \arctan 3/8  \arctan 5/2 \simeq 1.25 ^\circ$.
dxiv
 68,892
 6
 58
 112

In fact, not only is the "hypotenuse" not straight, it is bent in _exactly the same way_ as in the Wikipedia article: part of the "hypotenuse" has slope 2/5, and the rest of it has slope 3/8. – David K Sep 02 '17 at 00:57

@DavidK Indeed, and also thanks for the pointer to that cute animation in the duplicate link. Somewhat related: [the perpetual chocolate](https://math.stackexchange.com/a/743074). – dxiv Sep 02 '17 at 01:08