I'll second what iguanodon recommended, but recommend that you pick a larger problem. A friend of mine refers to it as the "search for why". It's much easier to learn something new if you have an itch that you have to scratch or a particular problem that you have to solve, and for which the new thing looks like the correct tool

Something I used to do whenever I wanted to learn a new language was write an implementation of John Conway's life. Perl is one of the few where I haven't done that, and I recently saw a post here where it was very impressively done in just a few lines.

I started learning Perl in my late 30's, about your age, because there was some web stuff I wanted to do and Perl looked like the right tool. I spent a bunch of time getting up to speed with Perl and MySQL and web things, having had no prior experience with web things, but quite a bit of scientific and data acquisition/analysis programming. I learned a ton, and then ended up applying a lot of what I learned to a Ruby implementation (which was also new to me) that I eventually got up and running. Since then I'm using both Perl and Ruby for another, bigger project.