The discovery of this dissection is normally attributed to Henry Ernest Dudeney but may have been first discovered by C. W. McElroy.
The dissection is hingeable.
This is not a very elegant solution because of the rather small piece, but it is another example of a TT22 dissection. There are a number of different six piece solutions possible and this raises the question of whether or not a five piece solution exists. There is a range of rectangle shapes that will dissect to a pentagon in just five pieces, but I think it unlikely that anyone will find a five piece solution for the square.
This was one of the first dissection improvements I found, and I am particularly proud of finding it. The previous record that I knew of was a 9 piece dissection found by Lindgren. I managed to improve this in 8 pieces in a number of ways, and this made me sure that there had to be a 7 piece solution. The problem is the plain square strip cannot be overlaid over the usual heptagon strip since the square strip is too wide. So I looked for a narrower heptagon strip. The technique I use allows me to produce a range of heptagon strips, but I chose the one that ensures that an edge of the heptagon coincides with an edge of the square. This saves a piece giving a 7 piece record. I don’t believe that a further improvement exists.
The first of these two dissections was my first solution of this dissection. It suffers from having several short straight lines that don’t show up clearly in diagrams of this size. Click on the diagrams to see an enlargement. The second solution is much more elegant.
Greg Frederickson suggested that I tried dissecting the hendecagon to a square. This is my best solution after many attempts.
Compare this dissection with that for the hendecagon, tridecagon, pentadecagon and heptadecagon. Each of these dissections uses basically the same technique.
This dissection has a small piece at the left hand side of the 23-gon.