Is there a standard for temperature similar to those for length, mass, etc.?

I’m piggybacking here to ask sort-of related questions about temperature.

I do a lot of canning. Standard instructions for making jam with added powdered pectin are:

  1. Heat fruit and pectin to a full rolling boil.

  2. Add sugar all at once.

  3. Heat until mixture reaches a full, rolling boil – one that cannot be stirred down. Boil for exactly one minute.

I’ve long wondered about 2 things: Why add the sugar all at once, rather than stirring it in? (I assume it has to do with temperature.) What’s the temperature difference between jam at a boil, and jam that is at a full rolling boil that cannot be stirred down?

All the rolling boil business is to assure the jam gets the proper jell.

(I take that “exactly one minute” instruction with a grain of salt. There have been many times when I noted the time when I thought it was at a full boiling boil, decided it wasn’t quite, and started the timer again or just added ten seconds to the time. Other times I’ve thought, “Gee, maybe that wasn’t quite a full rolling boil when I started the timer. Oh, well, close enough.” The jam has always come out okay.)

I am guessing that if you add it gradually, the sugar you put in first will be more dissolved than the sugar you put in last. You probably want a degree of consistency.

It’s because those units had absolutely no logical connection between them. They were different standard measures used in different contexts by different people in different places and for different reasons. So you have barrels of liquid. The barrels are all different sizes. Then somebody decides that we need a standard sized hogshead barrel. But the volume of that barrel isn’t defined by how many gallons it holds, it’s defined by the standard barrel, nobody cares how many gallons it holds.

Then someone bothers to ask, “how many gallons is this hogshead anyway?” And they measure it and it turns out to be fairly close to 63 gallons. That’s wine hogsheads, which are different than beer hogsheads. And then people get really finicky and adjust the standard hogshead to be exactly 63 gallons. But there’s no logical connection between hogsheads and gallons, any more than there’s a reason a pound is 453.6 grams, or that a US gallon is 231 cubic inches. It just turns out that a gallon is 231 cubic inches.

These units were never meant to be converted to one another. You’d never measure the length of a road in feet, or the height of a building in inches. Each trade would have customary units that they’d do all their measuring in, and only convert out to other types of unit when absolutely necessary. So there were literally dozens of standard measures for volume, all used by different trades for different purposes, and nobody cared that the standard for volume for grain wasn’t the same as the standard for volume for wine which wasn’t the standard for volume for beer.

Yes, of course that’s how it happened, but that’s a bug, not a feature. Because eventually, you really will need to measure the length of a road in inches. You start with the road measured in miles. But then someone comes along who wants to put up houses along the road: He knows how many feet wide the house is; how many can he fit in that length of road? And each house has walls made from studs every 18 inches; how many 2x4s does the builder need?

Me too, but in the U.K… and later in life I moved to the U.S.

One of the weirdest things is weather temperatures.

I grew up with warm weather being the 70s or 80s, and I never got the same quick intuitive feel for warm temperatures expressed in Centigrade. As a youngster in the U.K. I would mentally convert new-fangled warm Centigrade temperatures to Fahrenheit. So, at the top of the thermometer it felt comfortable to revert back to Fahrenheit when I came to the U.S.

But Centigrade was always more natural for low temperatures, because of the relationship to water freezing. For a while after I moved to the U.S. I would convert cold temperatures the other way (Fahrenheit to Centigrade) to figure out how cold it was, although that improved with time.

Today I just discovered (doing some googling to see when the U.K. officially switched) that I’m not alone in this.

Related to the original question, how do we interpolate intermediate temperatures accurately? Presumably with a simple mercury thermometer we rely upon accurate construction and assume linear apparent expansion. Are there better ways?

To the OP, Chronos may come in and correct me, but I don’t think it would be possible even in theory to achieve the same sort of accuracy with temperature than you could with other measurements such as time, distance and weight. temperature by its nature is random molecular motion, so there is always going to be some element of random error that can’t be eliminated.

here’s an interesting video on trying to change the definition on a kilogram, so it no longer relies on the vagaries of a physical object.

I talked about this in Post #21. I’ll expand here.

Between 13.8 K (-259.35 °C) and 1234.93 K (961.78 °C) the standard interpolation instrument for measuring temperature is the standard platinum resistance thermometer (SPRT). In Post #21 I incorrectly stated the Type S thermocouple was a standard interpolation instrument above 961.78 °C. That was the case before 1990. Today we use an optical pyrometer as the standard interpolation instrument above 961.78 °C.

For temperatures less than -259.35 °C we use the helium vapor pressure thermometer as a standard interpolation instrument.

More info here.

Every measurement has random error that cannot be eliminated. This includes time, mass, and distance.

I went to a lecture a couple of years ago by one of the lead scientists at the UK’s National Physical Laboratory. They were working on Acoustic Thermometry which aims to measure temperature through variations in c, the speed of sound in an inert gas. The advantage would be that the way c varies is predictable by fundamental principles so can be used to calibrate other thermometers.

And that brings up an interesting point about measurement devices.

Some measurement devices are not based on rigorous, physical laws. In the world of temperature measurement, this includes platinum resistance thermometers (PRTs), thermocouples, thermistors, and solid-state sensors. These make wonderful interpolation instruments, but physicists at NIST and other primary standards labs have a bit of “contempt” for them. (Not really true. I am just trying to make a point here.) Why? Because their behavior cannot be mathematically modeled based on physical laws; their “behaviors” are strictly based on empirically-derived data when compared to devices based on laws. OTOH, physicists love gas thermometers because they’re based on physical laws. The fact that they’re exotic, expensive, and a pain in the ass to use only adds to their “contemptness” for interpolation sensors. :wink:

Same goes for humidity measurements. Capacitive-type RH sensors are used for interpolation, and they do a good job of it. But when a physicist at NIST wants to generate a known RH, they will use a device like this.

Not that it matters, as both F and C are now derived units, but the reason for 32F and 212F is they are 180 degrees apart.

IMHO except for convention we should be using K, and it is annoying that Fahrenheit missed his intended round number mark (human body temperature was intended to be 90).

But the reason that older systems tended to use numbers with factorials is you could easily produce more accurate graduations using geometry and simple tools like dividers.

Base 10 measures, which is only convenient because we ended up with a number base related to the number of digits, requires more modern tools to try and produce accurate divisions.

It’s been way too long since I took Chemistry in school, but doesn’t boiling it hard for a full minute assure that all the sugar is dissolved? Forget the fruit and pectin parts, and assume that you have a pot of water boiling on the stove. You add sugar to the water, bring it back to a boil and boil it hard for a minute. Won’t all of the sugar be completely dissolved regardless of whether you stirred it in bit by bit or dumped it all in at once?

Factorials? 2, 6, 24, 120 … I think you might have meant something different, as I assume you want to include 12 and 60 in the numbers used.

You can totally easily geometrically divide a line segment into any number of equal parts; it does not have to be a factorial or any sort of magic number.

True, but elements of mass, time, distance, represent physical quantities, down to the level where quantum mechanics becomes important. A carbon 12 atom has a particular rest mass, a cesium atom transitions at a specfic frequency, and light traveling at that frequency has a particular wave length. But temperature refers to the random motion of molecules, therefore any effect that is going to be used to measure this temperature is going to be random. If we define the kelvin scale in terms of the triple point of water, then we are defining it in terms of the temperature at which a molecule turning into ice, vapor or water. But whether any given molecule changes state is random. So it is possible that a particular arrangement of molecules at one temperature will have more ice forming than a different arrangement at a lower temperature, and there is no way to avoid this.

The fact that the behavior of individual molecules is unpredictable does not mean that a sample of pure water does not have a very well-defined temperature. There are a lot of molecules in there.

ETA your nice, hot cup of tea will not randomly, spontaneously freeze.

Maybe I should say, it won’t partly freeze and the rest boil away, any more than a gas in a box will randomly occupy half the volume.

It has been 42 years since I took chemistry in school and I don’t remember much about solubility, other than sugar dissolves faster in hot water and salt doesn’t. But I think the solubility rate depends on the ratio of sugar to water, not just the temperature. If the solvent becomes saturated with the solute, the solute won’t dissolve at all. If you put the sugar in all at once, it will all dissolve at the same rate, but I believe that if you add sugar later it will dissolve slower than the sugar you put in first.

If the solution is not saturated, and you boil long enough to make sure that the last batch of sugar is dissolved, then it will *all *be dissolved. But this is a recipe and they may be trying to make it foolproof.

I hasten to add that your question was more complex than just about dissolving the sugar, and I have no idea of the impact to the rest of the recipe of adding the sugar in batches.