I was reading through this thread, and noticed the comments on hit/miss governors. Regarding gas engines (in the UK sense coal-gas, now usually propane but there are some exceptions), on full-size units (which means anything from not-much-bigger-than-a-model, up to 18" bore 28" stroke, at our museum) the governor usually works on a valve which admits gas to the somewhat basic mixing chamber, which replaces a carb on a petrol engine. Larger ones tend to all be rotating-ball governors, which act on a fat shim which rides up and down, pulled by the balls as they rise and fall. This shim sides between a kind of 'pecker' which is operated by the cam which is meant to admit gas, and the gas valve itself. If the shim rides up (too fast) it is not 'there' to bridge the gap when the pecker comes forward, so the valve is not actuated, no gas is admitted, and speed falls. Inlet and exhaust valves just operate as normal all this time. There is often an air-start valve too, for compressed-air starting, on big ones, but modellers probably don't need to know about that!
On the way in the gas pipe goes through a simple conical tap. This is not a throttle - it controls the mixture, and basically determines whether the engine will run or not. If you want to go faster or slower you adjust the governor - which may mean messing with the gas tap too, to get the thing to run. These are single-speed engines, designed to run very slowly and economically producing a tiny amount of power (since such low speed) for their size, but to do so for decades on end (again, low speed means they don't wear out).
Smaller engines, and the ones I have seen this on are by Gardner, have an interesting and different arrangement. There the pecker (my term and a lousy one - I'm sure it has a real name) is flung forward by the cam but is free to rotate on a pin somewhere at its 'back' end. As soon as it starts to move it hits a peg which flings it up in the air, rotating around this pin - and then gravity starts to rotate it back down, all the while it is moving forwards towards the gas valve pushed by the cam. If the engine is going slow enough to need another 'bang', it will also be going slow enough to give time for the pecker to have fallen back into position before the assembly meets the gas valve - and gas is admitted. If the thing is to fast, the pecker is still on its way down when the assembly meets the valve, so the valve is not actuated and no bang occurs.
That's a pretty crappy explanation. It's easier when you stand and look at it!
Oh - the exceptions to running on propane are flame-licker vertical rack-and-pinion engines, where (I am told) the lack of hydrogen in propane as compared to coal-gas means the flame licker ignition does not work properly. So we run the engine on propane but the pilot light for the ignition, on hydrogen. This is a bit of a pain as the inspection regime for hydrogen stuff is a bit onerous here.
Design speed on a medium-sized one is typically only about 140rpm, and ticking over it will only fire every 3 or 4 cycles - on-load it will of course fire every time. Big flywheels will try to even the running speed out (and make running any faster, really dangerous!), but twins were increasingly employed as mill lighting went electric and the pulsing of the lightbulbs with the firing stroke was annoying. That itself was a change from the pulsing of gas lights with the firing stoke, due to the engine taking a big gulp of gas and dropping the pressure in the pipe. That was helped by rubber bags on the gas inlets near the engine, which act as a local reservoir or gas accumulator.
cheers
Mark