Thomas Midgley’s Leaded Gasoline Choice Wasn’t About Chemistry

In October 1924, workers at Standard Oil’s tetraethyl lead plant in Bayway, New Jersey, started seeing things that weren’t there. Colleagues nicknamed it the “looney gas building” after men on the line began hallucinating, convulsing, and being carried out in straitjackets. Five or six of them died — accounts from the period disagree on the exact number — and more than forty were hospitalized. The chemist behind the additive, Thomas Midgley Jr., had a plan to convince the public that leaded gasoline was safe: he was going to inhale it, on camera.

The standard version of this story is a straightforward corporate cover-up: industry knew tetraethyl lead was poison, industry lied anyway, regulators caught up fifty years too late. That’s true, but it skips the more interesting failure. General Motors, Standard Oil, and DuPont didn’t choose leaded gasoline because it was the best anti-knock chemical on the table — they chose it because it was patentable, and the safer alternative sitting right next to it, ethanol, was not. The deciding variable in one of the twentieth century’s worst public-health decisions wasn’t combustion science. It was patent law.

Thomas Midgley’s Leaded Gasoline Demonstration

That demonstration happened in February 1925, in front of reporters, and it’s the emotional center of this whole story. Midgley poured liquid tetraethyl lead over his hands, held a beaker of it under his nose, and breathed the vapor for sixty seconds, announcing, “I am not running any danger.” What the reporters in the room didn’t know — what Midgley did know — was that he’d spent several months of 1923 recovering from his own lead poisoning, contracted while developing the same compound. He wasn’t gambling with an unknown; he was performing certainty about something he’d already been proven wrong about.

Midgley had run the numbers before he ever ran the press conference. He calculated that tetraethyl lead — branded and sold as “Ethyl” — could capture roughly 20% of the U.S. gasoline market, worth about $36 million a year. In 1924, General Motors, Standard Oil, and DuPont formalized that bet by founding the Ethyl Corporation as a joint venture, with Midgley as the public face of its safety. Bayway was one of several plants built to meet demand nobody outside the industry had actually asked for.

The industry’s confidence held for fifty years, more or less on schedule with federal regulation finally catching up. What changed in the meantime shows up cleanly in one number: the amount of lead circulating in the average American’s bloodstream.

How Tetraethyl Lead Actually Stops Engine Knock

Here’s the chemistry Midgley was selling, in plain terms. Gasoline engines rely on a spark to ignite a precisely timed fuel-air mixture; knock happens when pockets of that mixture detonate on their own, ahead of the spark — like a crowd starting the concert before the band walks on stage. Tetraethyl lead works by releasing lead-oxide radicals during combustion that scavenge the free radicals responsible for that premature ignition, smoothing the burn so it happens on the engine’s schedule instead of the fuel’s. It’s a genuinely elegant fix, the same kind of hidden physics that quietly runs everything from your car to your data center — which is exactly why nobody thought hard enough about where the lead went afterward.

That combustion lead didn’t just disappear out the tailpipe — it left as fine particulate, got inhaled or ingested, crossed the blood-brain barrier, and stayed there. Lead is neurotoxic at chronic low doses, which is why its damage doesn’t show up as one symptom; it shows up decades later as a population-wide shift in blood-lead surveys, and, more permanently, as an actual stratigraphic layer in polar ice. Arctic lead recorded in Greenland ice cores rose 250- to 300-fold between the early Middle Ages and a peak in the 1960s, then fell sharply once the phaseout began, a finding published in PNAS in 2018. You can date the twentieth century’s gasoline choices in ice, the same way you’d date a tree by its rings.

None of this was a surprise waiting to happen — ethanol was already a known, working anti-knock option. It suppresses knock through an entirely different route: a much higher latent heat of vaporization, meaning it cools the incoming fuel-air charge enough on its own to prevent premature ignition. Kovarik’s reading of the archival record has GM’s own researchers working on ethanol blends well before Ethyl reached the pumps. It just wasn’t chosen.

Why Leaded Gasoline Beat Ethanol at the Pump

This is where historian William Kovarik’s reading of the historical record gets sharp, and where I think most retellings of this story undersell the point. Kovarik argues that tetraethyl lead was chosen over ethanol not because it was chemically superior, but because it was a molecule GM, Standard Oil, and DuPont could patent, brand, and meter into every gallon sold — ethanol is a commodity anyone with a still and some corn could make, and nobody can own that. A patented additive dosed in parts per thousand creates a permanent, controllable revenue stream; a crop-derived fuel additive creates a market with no gatekeeper. Put crudely: they didn’t pick the better chemistry, they picked the ownable chemistry.

Here’s the uncomfortable part, and it’s worth sitting with instead of reaching for a villain. Midgley very likely wasn’t a cynic — by every account, he seems to have believed each reassurance as he gave it, including the one that put lead on his own skin. That’s worse, not better. Malice is at least self-limiting; a confident, sincere expert who has convinced himself he’s right is a failure mode that scales, because nobody around him has a reason to check his work.

The Slow, Measurable Cleanup

It took decades, and one very stubborn geochemist, to even establish how much of that lead was industrial rather than natural. Clair Patterson, trying to date the age of the Earth using lead isotopes, kept finding his own clean-room contaminated by airborne lead, and turned that nuisance into a career-long campaign against leaded gasoline that the industry fought him on for years. The blood-lead collapse charted above is the human half of the cleanup; the air itself tells a parallel story, measured against a legal limit that kept getting stricter as the science caught up. That regulatory chase is worth putting on its own chart, because the gap between “legal” and “actually present” turns out to be enormous.

The EPA’s legal ceiling for airborne lead dropped tenfold between 1978 and 2008, from 1.5 µg/m³ down to 0.15 µg/m³ — but the actual measured air had already been undercutting that ceiling, and kept falling well past it. The national median design value went from 0.40 µg/m³ in 2010 to 0.05 µg/m³ in 2021, an 88% drop in barely a decade. Regulation set the floor; removing lead from gasoline is most of why reality fell through it.

Freon, and a Confidence That Never Wavered

Six years after washing his hands in tetraethyl lead, Midgley delivered his second world-changing invention: dichlorodifluoromethane, marketed as Freon, the chlorofluorocarbon (CFC) refrigerant that made modern refrigeration and air conditioning possible. He introduced it at a public demonstration too — inhaling a lungful and exhaling it to snuff out a candle, proof it was inert and safe. It was, for humans, in the room, that day. It also turned out to be quietly dismantling the ozone layer, a fact nobody in that room could have detected with 1930s instruments.

There’s a pattern here worth naming: an expert stakes his own body on a claim of safety, is believed, and is wrong in a way that takes decades to show up in the data. I’ve written before about a researcher who became an experiment in his own field by accident; Midgley did it on purpose, twice, as marketing. Having spent years around CERN and inside data centers myself, I’d bet the next version of this story isn’t chemical at all — it’s a material or an algorithm someone insists is safe because they designed it, patented it, and have every incentive to believe their own demo.

Midgley died in 1944, strangled by a pulley-and-harness rig he’d built himself to get in and out of bed after contracting polio — killed, in the end, by his own invention, the way his fuel and his refrigerant had already started killing at a distance. The lesson isn’t that inventors are villains. It’s that “patentable” and “safe” are different questions, and every time we let the first one answer the second, we’re running Midgley’s experiment again.