About a year ago I read a book on Galois theory. This is the part of maths that answers the question of exactly when a polynomial is solvable by radicals, and also there's some stuff about straight edge and compass constructions flying about. It's basically linear algebra and group theory being used to answer very well known problems in a neat way. These are the best bits of maths for me. I love it when a classical problem is solved.
On the other hand, I took a course on Algebras before Christmas. That was exactly the "answering questions nobody asked" stuff you mentioned! Though it does lead the way to group characters which are one of the neater things I've seen.
Just finished the learning part of my 3rd year. I've done most of the usual undergraduate stuff (linear algebra, analysis, algebra, topology, prob, stats, logic, set theory). Next year I want to look at Godel's incompleteness theorems, algebraic topology, analytic topology, more set theory, representation theory of symmetric groups and some other stuff that I haven't decided. But it'll be pure. Because I'm no good at applied maths.