I couldn't find any textbook references either, but the wiki notes cover it pretty well. Mudpuppy is spot-on with the analogy of an LC circuit (I'm taking out the R, since a strong R would dampen the resonance). Poor or non-existent grounding exacerbates the problem.
Years ago we had a large customer data center that was self supporting during power outage. When they lost the utility connection, some strange combination of the non-linear loads (overhead lighting, computers, and UPS), underground cables, generators, and an ungrounded delta transformer would lead to occasional burning up of the weakest link (a transformer winding). It took a while to figure out that there was some electrical voodoo going on, not just a single bad piece of equipment.
There are many IEEE papers on ferroresonance as well.