Another daft chemist question: I'm looking at two charges spread over two overlapping Gaussians. I want the energy of the interaction, so, I take qq'/r and integrate first over one Gaussian then over the other... except for it all goes up the spout, because the charges are overlapping, and there are a lot of r=0. There's got to be a really easy way round this, which doesn't involve disrupting the continuity of the Gaussians. What is it??