Hacker News Reader

Not books as such, but great lost works: Fermat's proof of his last theorem. He wrote a letter to a friend that said something like "I have a most elegant proof, but it's too big to fit in the margin." Also the second half of Coleridge's Kubla Khan, lost because of the man from Porlock. Douglas Adams made great play of it in Dirk Gently, IIRC.

View on HN

Then look up Wiles' proof of the Taniyama-Shimura conjecture.

The math required is so advanced, and there have been numerous people who taught they solved it but turned out to be wrong.

It's extremely unlikely that Fermat had solved it.

PS. He wrote that note when he was young (and working on special cases, small n and such) and probably realized later on that he didn't have a proof for the general case.

Edit: There was no letter to a friend. It was in the margin of a book he was reading. More like a little reminder to himself to remember to jot down the proof.

"Great Lost Works" is an interesting subcategory. I'd include Aristotle's 2nd book of poetics on comedy, Claudius' Art of playing Dice, Euclid's Book of Fallacies, Philip K. Dick's three lost novels (A Time for George Stavros, Pilgrim on the Hill, and Nicholas and the Higs.), L. Frank Baum's four adult novels (Our Married Life and Johnson, The Mystery of Bonita, and Molly Oodle).

Reply on news.ycombinator.com