The best original source is probably An Enquiry Concerning Human Understanding, though it covers much more than the point I'm talking about. The point I'm talking about is that, according to Hume, there's actually no reason to believe that the future will continue to be like the past. Without this premise--which must be accepted without evidence--empiricism is broken. A basic treatment of Hume's argument is available from the Stanford Encyclopedia of Philosophy here: http://plato.stanford.edu/entries/hume/#CauIndInfNegPha