The Maxim chip is ideal, but as others mention, CCCV methods with a couple of 317s will dig you out of a hole fast. I think Linear technology do good charging solutions, but I have no experience with them. As Erik points out fast charging of Lead-acid is tricky when the cells are gel electrolyte, because of the loss of acid (excess hydrogen generation at the plates ??)

Apparently Yuasa's latest fibre filled cells are better
at handling this Erik.

If you are using LA batteries, you might like to look at something like the 'Cyclon' batteries (Hawker ??, Gates ??) , which are neat TOUGH 2V almost 'D' cell size units, with a fanatastic AH capacity ,and a more convenient shape for wrapping a product in !

LA work well for low self-discharge (better than Ni-cad, MUCH better than Ni-mh, not as good as LI-ion)
LA have great storage capacity, but don't like fast charge cycles which Nimh, Ni-cad and Li-ion can survive very well with smart chargers.

Steve