First, if you look at the internal wiring diagram/schematic of the can type coils in our antiques, they do not have any of its leads wired to ground.
Going forward here, in order to ease describing things, I'm going to need to get a little soft in the engineering descriptions, and instead use some everyday type descriptions/examples. Not trying to insult anyone's intelligence here, but it's the only way I can figure out how to get the coil operating process explained in a non-engineering geek environment.
Although an ignition coil isn't designed or used in the same manner as a transformer, they both have primary and secondary windings. There are two terminals for each winding, producing four terminals in a transformer. Put an AC signal into the primary windings of a transformer, and you'll get an AC waveform out of the secondary windings. With an AC signal in the secondary, you'll obviously get one secondary terminal to be positive, and the other secondary terminal will be negative (relative to each other).
Oops, gotta stop and take care of a house issue. Will be back later to finish. In the mean time, remember the part about the secondary winding voltage levels.
(Okay, I'll admit my ignorance here on what is required to edit/add to a previous post, like above.)
Okay, let's use the analogy of a step-up transformer. When you input a "small" voltage into the primary windings, you get a larger voltage induced in the secondary windings. In a 100:1 turns ratio (Secondary to Primary) transformer, if I put 200 volts in the primary winding, I'll get 20,000 volts in the secondary winding (measured across the two terminals of the secondary). If it was possible to quickly put a DC voltmeter across the secondary terminals and read the voltage (at the peak of the waveform) we would read either +
20,000 volts, or we would see -20,000 volts, just depending on which way you hooked the voltmeter.
A "similar enough" thing happens in an ignition coil (except we're using pulsed DC current, instead of AC current). When the points open, the primary current is interrupted/halted. Because the coil is an inductor (and not a wire resistor), when the magnetic field collapses we get a flyback (inductive kick) voltage seen at the primary windings. This flyback voltage can jump up to 100-200 volts. If you take this 200 volts reading, and multiply it by 100 (ie: the coil turns ratio) you get 20,000 volts across the secondary winding. (Remember, one end is going to be +20,000 volts compared to the other, and the other is going to be -20,000 compared to the other.)
Okay, we've got 20,000 volts at the coil secondary to play with to arc the plug gap. How do we hook things up? Well, remember that to get a plug to arc acceptably "easy" we need to make the hotter of the two plug electrodes negative voltage compared to the "cooler" electrode. (Those of you who are familiar with vacuum tube theory and operation will understand.) So, we hook the negative voltage side of the secondary winding to the coil center terminal that goes to the distributor cap, and eventually through the plug wires and then the spark plug terminal and finally to the plug center electrode. Now, what to do with the other (ie
ositive) terminal of the secondary winding? There's no fourth terminal in our old can coils (it's cheaper to make them that way), so the manufacturers just hook the positive terminal to the coil C+ terminal. It's not as "pretty" as hooking this secondary terminal to ground in a fourth coil connection, but the 12 volt DC voltage offset on the secondary windings is insignificant compared to the -20,000 volts on the other end of the windings.
So, during ignition (plug arcing) your fast voltmeter, or oscilloscope will read (using the chassis as a ground/reference point) -20,000 volts at the plug center electrode, and zero volts at the side electrode. Or, if you'd like, you can use the center electrode as the reference, and you get zero volts at the center, and +20,000 volts at the side electrode. Either way, the center electrode is always negative compared to the side electrode, and the side electrode is always positive (the anode) compared to the center electrode (the cathode).
Well, this went longer than I had planned, and as I mentioned before, I've had to bend the physics a little here to make it palatable to read.
It's late, and I'm tired of typing for now. Let me know if something needs additional clarification.