For what it's worth, you may find the fluctuation theorem (http://en.wikipedia.org/wiki/Fluctuation_theorem http://en.wikipedia.org/wiki/Fluctuation_theorem) in

For what it's worth, you may find the fluctuation theorem (http://en.wikipedia.org/wiki/Fluctuation_theorem http://en.wikipedia.org/wiki/Fluctuation_theorem) interesting. It shows that the second law is incorrect in the same way as Newton's theory of gravity was: both are accurate, in their domain of applicability, but neither had such boundaries well defined. Where Newton's law fails for very heavy systems, the second law fails for very small (though by a far smaller amount).