[ds9]

Another example is the Fahrenheit scale. The eponymous originator calibrated his thermometer by the temperatures where he could obtain brine (0 F) and an inaccurate estimate of human body temperature (100 F). https://en.wikipedia.org/wiki/Fahrenheit

[amalag]

It is still nice because it is human centric, roughly 0 to 100 for temperatures is easily relatable to the human body. 90 is hot, 10 is cold. Celsius starting at the freezing point of water is not as human centric.

[raverbashing]

No, it's still a piece of crap

"Human centric" is a false advantage. And it's not even human centric, the temperature of 100F is wrong, and what does brine have to do again with the human body?

[fancyketchup]

Gah. Can we please stop with this bullshit already? Fahrenheit has nothing to do with body temperature. The human body temperature is ~100F by coincidence, not by design.

The Fahrenheit scale has 180 divisions (nice round number, evenly divisible by loads of integers) between the freezing point of water and the boiling point of water (212F-32F=180, for the math challenged). The 0F point is the equilibrium point of a frigorific mixture of equal parts ammonium chloride, ice, and water. One can easily produce two reference points for the F scale in a way that isn't sensitive to the local pressure using just ice brine (0F) and ice water (32F). The melting temperature of ice and the equilibrium temperature of the brine mixture are constant on the pressure scales and temperature scales available to metrologists before about the 19th century[1][2].

The same can't be said for the Celsius scale, which requires either a triple-point cell or a known atmospheric pressure (so that the boiling temperature of water is well-known).

Edit: So the close out the point, the Fahrenheit is a unit which is much better matched for the calibration tools and computational methods available around the time it was invented.

[1] Maybe a better way to state this sentence is that the temperatures are constant "to within the experimental uncertainty achievable at the time."

[2] It's well known to anyone who has ever cooked that water boils at an appreciably different temperature in Denver than in Los Angeles.

[stephencanon]

0 is uncomfortably cold. 100 is uncomfortably hot. The temperatures that non-STEM people (“humans”) deal with on a day-to-day basis can be represented with satisfactory precision using two digits (OK, three in Australia), no negative numbers, no decimals. It’s a great system for casual temperatures.

Whereas “reasonable” temperatures in C run from what, -18ish to 38ish? Where’s the sense in that?

If you want to defend a unit of temperature, defend Kelvin. Don’t pretend that ˚C is significantly more reasonable than ˚F. They’re both bonkers in their own way (mysticism about the vitality of water is not a real justification).

If (more likely) you just want to make fun of Americans for being rubes, use distances or weights or volumes as your example.

[bcj]

There are a lot of humans that live places where we regularly see negative temperatures. Where I live the temperature range is roughly -40–85ºF.

Choosing the freezing point of water isn't really mystic. It lets you know whether it will rain or snow. It also means that any negative temperature is capable of causing frostbite.

[mpyne]

> It lets you know whether it will rain or snow.

Not really, since it depends upon temperature at the clouds, not at the ground level. Oh, and it also depends on atmosphere pressure, and on time (since a phase change requires latent heat transfer in addition to merely being at the right temperature).

And as baddox mentions, 32 is hardly harder to memorize than 0, especially if you know anything at all about computers.

[baddox]

> Choosing the freezing point of water isn't really mystic. It lets you know whether it will rain or snow.

Fahrenheit also has a chosen freezing point of water: 32 degrees. This is no more difficult to remember than Celsius' zero degree freezing point.

[scott_karana]

"Uncomfortably cold" is quite an understatement, when talking about allegedly "human" temperatures.

Sure, at 100, healthy humans can survive as long as their increased water needs are met. (Sun is another matter, but I'm just talking temperature.)

However, at 0F, permanent tissue damage onset is within 30 minutes, with almost any wind speed. (http://www.nws.noaa.gov/os/windchill/images/windchillchart3....)

[stephencanon]

Humans have developed clothing, which allows us to stay alive in cold weather. There’s very little we can do (short of going into a building with AC) to survive hot weather — 100 is tolerable (just), but heat gets deadly very quickly when you go above it. You don’t want to be outdoors for extended stretches in either condition (the analogy is imperfect; realistically my experience of 10F is subjectively similar to 100F, and 0F is more like 110F).

[scott_karana]

How does that change the fact that 0F is more than "uncomfortably cold" for most people?

I'm not disputing the 0F to 100F (or -18C to 40C) range in terms of being "regularly seen".

EDIT: I suppose the core of the misunderstanding is that I was addressing the "human centric" advantage and relation to brine, where you were just arguing (quite reasonably) the convenient representation of the range?

[scott_karana]

I agree that the smaller degrees, and lack of fractions, is more convenient for casual usage.

However, I'd argue that being water-based is hardly mystical: if water is boiling, or frozen, survival is more difficult. It's also a pragmatic system for any nation that regularly drops below freezing. (Eg, could it snow today? Could there be ice to slide on?)

[saraid216]

> (mysticism about the vitality of water is not a real justification)

Water was also used in the original definition of a gram, which is a fairly good reason to reuse it in another definition.

[Pxtl]

Yeah, you could argue that if it was an accurate 100F for the human body, but you'd still want the bottom level to be something useful - freezing point of water is as good a zero as anything (well, short of absolute zero).

My bigger question is why Metric stuck with Celcius and Kelvin instead of factoring in the Gas Constant (R) into Kelvin so the math would be completely constant-free.

Of course, then you'd have water freezing at 2271 and boiling at 3102, which wouldn't be fun in conversation.

[raverbashing]

> My bigger question is why Metric stuck with Celcius and Kelvin instead of factoring in the Gas Constant (R) into Kelvin so the math would be completely constant-free

That's a really good question. Another possibility would be to fix the calories/joules mismatch in heating water

[marcosdumay]

The Metric system changes are all about preserving the measures people are used to, and not at all about removing constants from calculations.

[eigenvector]

Fahrenheit is not intuitive, you just think so because you grew up with it. People who grew up with Celcius have no problem relating to how cold 30 C or 100 C or -20 C are.

[baddox]

Celsius is no more intuitive for everyday usage. Of all the scientific unit debates, the temperature one is by far the silliest. There are very few temperatures that people need to memorize, and they're trivial to memorize in either temperature scale.

[com2kid]

This I disagree with. Born and raised in America, I still cannot remember what temperature water boils at in Fahrenheit, I just remember it is something over 200 degrees.

Celsius is easy, 100. :)

[amalag]

I never said intuitive, I said human centric. Everyone can relate to a number, but it is not as human centric as Fahrenheit which has 0 to 100 more relative to a human scale.