Fahrenheit to Celsius: The Backstory Is Wronger Than the Math

A Glassblower from Danzig Decided How Hot You Are

Daniel Fahrenheit didn't start out trying to build a temperature scale. He was an instrument maker — one of the best glassblowers in northern Europe — and his real obsession was the thermometer itself. Before Fahrenheit, thermometers used alcohol or water. They were inconsistent. One instrument maker's "cold" was another's "lukewarm." Fahrenheit switched to mercury, which expanded uniformly with heat across a much wider range, and he figured out how to blow glass tubes with perfectly uniform inner bores so the mercury column would rise at a predictable rate. The thermometer was a breakthrough. But a thermometer without numbers is just a fancy glass straw.

He needed a scale. So he picked three reference points — not because physics dictated them, but because those were the temperatures he could reliably reproduce in his workshop in 1724.

0°F. The coldest thing Fahrenheit could make in his lab: a slush of water, ice, and ammonium chloride (sal ammoniac). This brine mixture stabilizes at about −17.78°C — not the coldest temperature possible, not absolute zero, just the most convenient deep-freeze available to an 18th-century instrument maker with a bucket and some salt.

32°F. The freezing point of pure water. Why 32 and not some rounder number? Because Fahrenheit set 0 at the brine slush, and from there to water's freezing point spanned 32 of his degrees. He didn't choose 32 — it fell out of the degree size he'd already defined from the brine-to-body-temperature span. More on that in a moment.

96°F. Human body temperature. Fahrenheit calibrated this by putting a mercury thermometer under his wife's arm (or, some historians say, a "healthy man's" armpit — the primary sources are ambiguous). He set that reading as 96. From his brine zero to 96 spanned exactly 96 degrees. Each degree was, by construction, 1/96 of the temperature range a human body could produce. It was elegant. It was also wrong.

Fahrenheit later re-anchored his scale so that water froze at exactly 32°F and boiled at exactly 212°F — an interval of 180 degrees. Half a circle. In the angular logic of his era, that felt satisfying. The body-temperature anchor drifted from 96°F to 98.6°F not because Fahrenheit corrected it, but because later scientists recalibrated against the new freezing/boiling anchors. Fahrenheit died in 1736 believing body temperature was 96°F.

The Other Scale: Why Celsius Had It Backwards

Anders Celsius was a Swedish astronomer who proposed his scale in 1742 — eighteen years after Fahrenheit published his. And Celsius did something peculiar: he set 0° as water's boiling point and 100° as water's freezing point. The scale ran backwards. Boiling water was "zero degrees Celsius." Freezing water was "one hundred degrees Celsius." It made a kind of sense to an astronomer — 0° for the hottest thing (boiling), higher numbers for colder things, like counting down elevation from a mountain peak. But it was inconvenient for everyone else.

Celsius died of tuberculosis in 1744, age 42. Two years later, Carl Linnaeus — the Swedish botanist who invented biological taxonomy — commissioned a thermometer with the Celsius scale flipped: 0° at freezing, 100° at boiling. Linnaeus didn't always get credit for the reversal, but the surviving thermometers from his Uppsala greenhouse prove it was his idea. For nearly 200 years the scale was called "centigrade" (from Latin centum = hundred and gradus = step). The name "Celsius" was only officially adopted by the 9th General Conference on Weights and Measures in 1948. Celsius had been dead for 204 years by then.

Today Celsius (or Kelvin, its absolute sibling) is the standard everywhere outside the United States, the Bahamas, Belize, the Cayman Islands, and Palau. Liberia switched to Celsius for weather in the 1990s. Myanmar still uses Fahrenheit informally but adopted Celsius officially in 2013. The holdouts are shrinking, but Fahrenheit is stubborn — it maps better to human comfort. A 0–100°F scale spans roughly the range of outdoor weather most people experience (very cold to very hot), while 0–100°C spans the freezing and boiling of water, a range that only chemists and cooks encounter daily.

The Formula (and Why You Can't Just Multiply)

Most unit conversions are just multiplication. Inches to millimeters: multiply by 25.4. Pounds to kilograms: multiply by 0.45359237. Done. Temperature is different — it's an affine transform, not a proportional one, because the two scales disagree about where zero sits.

Step one: subtract 32. This shifts Fahrenheit's zero — that ammonium chloride brine slush at −17.78°C — down to match Celsius's zero, which is the freezing point of pure water. Now the origins are aligned. If you skip this step and just multiply, you'll be off by 32 Fahrenheit degrees every time.

Step two: multiply by 5/9. A single Celsius degree is bigger than a Fahrenheit degree — 1°C spans 1.8°F. Or said the other way, 1°F spans 5/9 of a Celsius degree. The factor corrects for that difference in degree size. Think of it as converting the unit on the axis after you've lined up the zero. If you've ever done a coordinate transform — changing from one map projection to another — it's the same math: translate the origin, then scale the axis.

The formula going the other direction, Celsius to Fahrenheit, is °F = (°C × 9/5) + 32. Multiply first (scale up the bigger degree), then add 32 (shift the zero point back). Same affine transform, inverted.

Worked Examples

−40°: The One Temperature Where Nobody Needs a Converter

Set °C equal to °F and solve:

x = (x − 32) × 5/9
9x = 5x − 160
4x = −160
x = −40

There is exactly one solution. No other pair of linear temperature scales can have more than one crossing point — it's a consequence of the affine form y = mx + b where m ≠ 1. If m = 1 the lines are parallel and never cross. If b = 0 the scales share a zero and cross only at zero. With m = 5/9 and b = −32 × 5/9 = −160/9, the lines cross precisely once. That crossing happens to land at −40.

This is not just a mathematical curiosity. Jet A-1 aviation fuel has a freezing point specification of −40°C — which also reads −40°F. Pilots crossing the North Atlantic don't need to convert. Weather stations in Yakutsk, Siberia and Snag, Yukon (where it hit −63°C / −81°F in 1947) routinely record temperatures below the crossing point. Wind chill at −40° freezes exposed skin in 2 to 5 minutes depending on wind speed. When you're at −40°, the last thing you want to be doing is arithmetic.

The 150-Year-Old Body Temperature Lie

In 1851, Carl Reinhold August Wunderlich — a German physician at Leipzig University — published what would become the most cited medical measurement in history. He recorded over one million axillary (armpit) temperature readings from roughly 25,000 patients using a 22-centimeter mercury thermometer. The average: 37.0°C. Convert to Fahrenheit: 37 × 9/5 + 32 = 98.6. That number, "98.6°F normal," went into every medical textbook published in the English-speaking world for the next century and a half.

There were two problems, and neither of them was small.

Problem one: the thermometer. Wunderlich's instrument was calibrated against a reference that was itself roughly 1.5°C warmer than the true thermodynamic temperature. Every reading in his million-point dataset was systematically high. His "37°C" was actually about 35.5°C if you correct for the calibration offset. That's not a fever — that's mild hypothermia by modern standards. But nobody knew, because accurate temperature standards didn't exist in 1851 outside a handful of national metrology labs, and Wunderlich wasn't at one of them.

Problem two: the measurement site. Wunderlich measured armpit temperatures. Axillary readings run 0.3°C to 0.6°C cooler than oral readings, and oral readings run about 0.3°C to 0.5°C cooler than core (rectal) temperature. The variation isn't constant — it depends on ambient temperature, how tightly the arm was held against the body, recent physical activity, and about a dozen other variables. Wunderlich didn't control for any of them systematically.

In 1992, Philip Mackowiak and colleagues at the University of Maryland published a study in JAMA that finally re-measured the thing. They took oral temperatures from 148 healthy adults, three times a day for three days, using calibrated modern thermometers. The mean: 36.4°C (97.5°F), not 37°C (98.6°F). The range across all subjects and times was 35.8°C to 37.2°C. A 1992 editorial in the same journal read: "The 98.6°F standard is a 19th-century artifact, not a 20th-century norm."

And it gets more complicated with age. A 2017 study in the British Medical Journal tracked body temperatures across 35,000 patients and found that normal oral temperature drops roughly 0.02°C per year of adult life. A 70-year-old with a 37°C reading might be running a 0.6°C fever by their own baseline. A 90-year-old with "normal" 37°C could have a serious infection. The clinical fever threshold of 38°C (100.4°F) was always a rough guideline, but for elderly patients it's actively misleading.

Wunderlich's million measurements were not worthless — they were the first systematic medical thermometry ever attempted, and they established that body temperature varies with disease, time of day, and individual physiology. But the single number "98.6°F" that survived is the statistical mean of a miscalibrated instrument, measured at the wrong anatomical site, from the population of one German city in the 1850s. Every American who panics at a 99°F reading is reacting to an artifact of an 1851 armpit.

Common °F to °C Conversions

Where Unit Confusion Kills

Medical devices are the highest-stakes application of Fahrenheit-to-Celsius conversion — and the one where errors are most frequent. A patient with a 38°C fever misrecorded as 38°F (which is below freezing) gets triaged for hypothermia. A sterilization autoclave programmed to 250°F (the correct cycle temperature) accidentally set to 250°C melts its contents and destroys the heating elements. A vaccine refrigerator at 4°C accidentally dialed to 4°F (−15.6°C) freezes every vial. The WHO estimates that temperature-related vaccine damage costs $200–300 million annually — a significant fraction from Fahrenheit/Celsius confusion in cold-chain logistics between US-supplied and internationally-supplied equipment.

In aviation, the stakes are different. METAR weather reports worldwide use Celsius for temperature and dew point regardless of the reporting country — a US pilot landing in Chicago reads "28/15" and knows it means 28°C, not 28°F. But the automated terminal information service (ATIS) voice broadcast may give Fahrenheit. A pilot who hears the ATIS saying "temperature eight five" (85°F = 29.4°C) and mentally plugs "85°C" into the aircraft performance computer will get an impossible density altitude and an incorrect takeoff roll calculation. The aircraft will fly — it just might not clear the obstacles at the end of the runway.

In 1999, a UK pharmaceutical facility ran a sterilization cycle at 121°F instead of 121°C. Surviving bacterial spores contaminated a batch of injectable saline. The recall covered 22 hospitals. Nobody died, but the investigation found that the temperature controller had a °F/°C toggle switch accessible from the front panel — no lockout, no confirmation dialog. The operator had bumped it while cleaning.

Related Unit Converters

Frequently Asked Questions