Well that went ok. I got two capacitors and put a diode in series with each, reversed one and put them across the wires that were showing a 44Ohm resistance (they're electrolytic and I don't want them to pop in my face). That should keep some energy around in the coils. It does look as if it'll have to be spinning quite fast to generate any useful voltage... but best of all, I didn't electrocute myself.

I'm guessing that the voltage produced against rpm looks like a nice smooth curve, and that the frequency of the AC generated goes up with increasing rpm. Anyway, if it's an 1800 rpm motor, we'd need somewhere around 30 revolutions a second. That sounds a bit fast to me. I don't really want to have to do any gearing, but maybe I'll have to. I suppose I should try and plot some graphs before I start worrying about all that.