I remember there is a proof in Michael Artin's book, by considering some "concrete" rotation of some polyhedron. But I thought that proof is not nice.
Tried to google the proof on web. Found the good website Groupprops. And then found the page for S5 and A5.
Yes, there is a proof by considering the conjugacy classes of A5. And by induction, one can prove that
A question: is there an intuitive way to see that in A5, the conjugacy class containing (1 2 3 4 5) and the conjugacy class containing (1 3 5 2 4) are different?