I disagree on the book recommendations. If you don't have a knowledge of the calculus through multivariable, reasonable performance in a formal course on differential equations, or probability, those books by Wilmott, Neftci, or Baxter and Rennie are not for you.
There's a reason that the programs that teach this mathematics in a formal setting, either through a phd program or a financial engineering program, require you to have such a comfort level with math through diff eq's and probabiliity.
I think some people "fake" their way through these books, but don't have the mathematical skill set to fully digest the message of what the theories are saying.
Anyone self teaching their way through that Wilmott book, a Student Introduction to ... is absolutely BS-ing themselves, and anyone they tell it came in handy.
Take a semester of calculus based probability and go from there.