Well, I learned something tonight. I took a brand new out of the box GP. Put my meter on voltage, and measured across the terminal and shell at room temperature. The reading was 0. I then put the tip of the GP on my kitchen cooktop, and turned it on. Damned if the voltage reading didn't increase in proportion to temperature. It makes sense, now that I think about it, since both the steel shell of the GP and the nichrome element are exposed to the same high temperatures. Steel and nichrome have different Seebeck coefficients, so a potential is created as the part is heated. As for "new" vs "old", it shouldn't matter, since the current arises directly from the properties of the materials. For the same reason, it shouldn't be necessary to use one reference plug moved from cylinder to cylinder, as long as all the plugs are from the same manufacturer and batch. The one hitch is that carbon deposits would affect the readings.