The ’73 GT had started to miss a few days earlier, and denial of the symptom worked for a while. But the missing got worse. After a longer and slower trip on the tollway than was comfortable, we knew that we had to face the problem. It was night, and we both thought it was time to at least check for a ‘Christmas tree’ around the distributor and ignition coil. We have tracked high-voltage wiring losses several times with nothing more than darkness as a tool - here is one saving grace for the Prince of Darkness.

The engine must be running for this simple check, and, since darkness is required to see ignition discharges, we were very careful not to risk the club neck tie, long hair, necklace or fingers in the engine bay. ‘Look, but don’t touch’, is the main rule when the engine is running, but it is too dark to seen anything.

But there were no streaks of light nor snapping sounds around the distributor cap or ignition coil or ignition wires. Then, as we stared longer into the deeper dark of the engine bay, we saw perfectly round rings of blue glowing a few inches off the side of the head. The electrical discharges occurred at the very ends of the spark plug boots. The rings of fire were most pronounced around the spark plugs 3 and 4, with a small band around spark plug 1.

We reasoned that the plug wires were fine, otherwise the spark would not have gotten to the spark plugs, where it evidently preferred to travel down the outside of the porcelain rather than the inside core. Had the electrodes all burned off? What exactly could cause this change of a porcelain insulator into a conductor that was better than the 25 thousandths air gap at the electrode?

We unscrewed the plugs, which, by the way, were only two months old, having been fitted just before passing the EPA emissions test on the first try. The tips were white, just like the too-lean problem illustration in a Chilton manual, although the electrodes had not been burned off, and the gaps were still fine. The Champion logo and plug type (RN9YC) numerals were very faint, as though the plugs had been held in a propane torch flame.

We corrected the condition, we hope, by replacing the plugs and enrichening the mixture by four flats on both carburetors. The engine runs without a problem, on idle, accelerating and at road speed.

The original mixture setting was border-line lean to pass the EPA standards. What we think happened was that the mixture went very lean, causing a very hot cylinder combustion condition. Fortunately, the plugs failed before the exhaust valves burned or a piston burned through or the head gasket failed, though we will not know for certain which of these conditions was developing until we remove the head at some future time. We think the plugs failed because they got hot enough to burn some of the every-present oil mists, oil drops and splashes of road salt. The oil and other materials turned into a carbon-salt coating on the porcelain of the plugs, and an insulator became a conductor.

Continuing our guessing game, we think the condition developed when the car, which had been tuned in February on winter gas, was asked to run on a different blend. We generally use one specific station for fuel, but have only a limited choice when out of the area. We use premium fuel, with the hope that the improved anti-knock properties will ultimately pay back the additional cost in terms of a longer time between replacing heads and pistons. But the fuel in the northern areas does change between winter and summer, and we have no way of telling what the level of oxygenation is for the fuel we use.

This is one case where a computer controller for the engine could have saved a problem: an oxygen sensor in the exhaust would have sensed a too-lean condition and increased the mixture in the fuel-air supply. And so the combustion temperature would not have increased. But an oxygen sensor for street use is not what an MG is about, and we should have done our part as MG owners and tinkered a little more and a little sooner.