I've thought about this myself.

Solar Panels:
Unless the car spends alot more time parked in the sun than actually driving it's probably not worth it. The tray area may be worth about 100W at the most in strong direct sun and cost a bundle. They'd be better on the roof of your house. The solar cars have state of the art panels and an extremely low drag coefficient and area using only on order of 100W to drive at speed. The main problem is solar energy is diffuse (i.e. spread out).

It would cost alot but if you drove the car everyday it would probably pay for its self in say 5-10 years and would be pretty unique. There's a number of places that do kits in the US, a place in Victoria that does conversions (Castlemain I think) and loads of webpages. Here's a couple

http://www.strathsteam.com/page10.html
http://www.aeva.asn.au/

I'd be impressed if you could do a conversion with good to high performance for less than around $10k. Depends what you want to use it for, for big power you need big batteries to supply the current, e.g. for a 144V system (a common dc system) you need 280 odd amps to make 40kW.

Another interesting prospect is salvaging electric motors from hybrid cars, I think Adelaide Uni had a student project along these lines, autospeed also have an article on it.

Best of luck, it's an interesting topic.