My "to-read" pile is so huge that I'm not allowed to buy any more books until Christmas. That is, books for me; buying for other family members is OK. Anyway, I found this in the pile left over from my summer purchases and got through it reasonably quickly. It's the story of mathematician Andrew Wiles and his predecessors in the solution of Fermat's Last Theorem, the proposition that
a^n + b^n = c^nhas no solution for n>2 and a,b,c > 0. To be honest I was a little disappointed in this best-seller. Having a doctorate in HPS I expect a little more social grit with my history of mathematical Great Men (and, surprisingly, a couple of women in this case). As a former Cambridge astrophysics student I like a little more maths with my accounts of what they did - contrast here the Sarah Flannery In Code book. And as a Clare graduate myself, I was frankly mystified that the college where Wiles actually did his doctorate and where his father was Dean is not even named - the only college that is actually name-checked is Emmanuel, where his supervisor was a Fellow. So I'm glad I bought it second hand, rather than new, and it didn't really enlighten me much more about any of my own occasional speculations into number theory.
I guess it's pretty certain that Fermat was wrong. His "wonderful proof" which was too small to fit in the margin cannot possibly have included the work of any of the dozens of later mathematicians drawn on by Wiles to compile his eventual paper of more than 100 pages. So presumably he had found what he thought was a proof, but because he never shared it with anyone in his lifetime, he never discovered that it was unsound.