Some notes on the history of temperature scales

Yesterday’s xkcd led me down a Wikipedia rabbit hole about the history of temperature scales. Some particular things that struck me:

  • It hadn’t occurred to me that standardized temperature scales weren’t invented until around 1700, and thus that before that, there were no standardized numerical representations of temperature.
  • Galileo or someone connected to him invented the thermoscope around 1600; it could tell you whether the temperature was going up or down, but it didn’t have numerical gradations. The general principle dates back to two thousand years earlier.
  • I’m not seeing anything in these Wikipedia articles about the history of temperature measurement outside of Europe, and TSOR isn’t bringing anything up.
  • I’m amused that two early and totally different scales, created about 30 years apart in the early 1700s, coincidentally were created by people with similar names: the Rømer scale and the Réaumur scale.
  • “Rømer’s thermometer was also an improvement in the fluid that he used. He used a mixture of alcohol and water, conveniently available in the form of wine.” I am tickled by the idea of a thermometer that measures temperature using wine.
  • I am also tickled by the word “thermometry.”
  • I am also tickled by the phrase “Vienna Standard Mean Ocean Water.” A lot of tickling going on over here this morning.
  • “Prior to 1743 the [Celsius scale] values were reversed (i.e. the boiling point was 0 degrees and the freezing point was 100 degrees).”
  • I started wondering why the Fahrenheit scale has the values it does—why 32 for freezing, 212 for boiling? Turns out that Fahrenheit adapted Rømer’s scale. Rømer had set the freezing point of brine (I’m simplifying here) to 0, and the freezing point of water to 7.5, and human body temperature to 22.5. So Fahrenheit multiplied Rømer’s numbers by 4 to avoid fractions and get more gradations, which meant the freezing point of water was 30 and human body temperature was 90. There are 60 degrees between those two numbers. But Fahrenheit wanted to be able to easily mark degree lines on a thermometer, so he adjusted the scale to get 64 degrees between the two; 64 is a power of 2, so he could just repeatedly bisect the distance between the two marks. So he set the freezing point of water to 32, and human body temperature to 96. Later, the scale was slightly redefined to set the boiling temperature of water to 212 (exactly 180 degrees higher than the freezing point), which left human body temperature at 98.6.
  • Every time I notice how isolated the US is in avoiding metric units, I’m surprised all over again.

Join the Conversation