Understanding a simple statement regarding the measurement of temperature

[JC2000]

Summary: 1.In the context of calibrating a scale to correlate volume change with temperature, my book states: "Since all substances change dimensions with temperature, an absolute reference for expansion is not available." What do they mean by an absolute reference in this instance?

My understanding of it is as follows. Different substances expand to a different extent at the same temperatures and thus an 'absolute' scale would not be possible to construct. Since choosing one substance would mean temperature is now measured relative to expansion of that substance.

Overall context : (from my book) : "Thermometers are calibrated so that a numerical value may be assigned to a given temperature in an appropriate scale. For the definition of any standard scale, two fixed reference points are needed. Since all substances change dimensions with temperature, an absolute reference for expansion is not available. However, the necessary fixed points may be correlated to the physical phenomena that always occur at the same temperature. The ice point and the steam point of water are two convenient fixed points and are known as the freezing and boiling points, respectively. These two points are the temperatures at which pure water freezes and boils under standard pressure. "

2. My other question : (reference to the underlined sentence).
Here the author mentions that two points are required to define a scale. Am I right to assume that two points are required since the scale is linear. Are there non-linear scales? Or are they adjusted for by using logarithms?

3. "Liquid-in-glass thermometers show different readings for temperatures other than the fixed points because of differing expansion properties.".

Does this mean that depending on liquid the scale would vary. But since the Celsius and Fahrenheit scales are with reference to boiling and freezing of water, thermometer makers would convert the observed expansion for a specific liquid to a celsius/fahrenheit value so that the celsius/fahrenheit scale could be applied to a thermometer with any liquid?

 

[sophiecentaur]

Am I right to assume that two points are required since the scale is linear.

The linearity is an axiomatic idea and any practical thermometer will depart from it. Four common forms of thermometer technology are Liquid in glass (the most convenient for centuries), Gas, Platinum Resistance and Pyrometric. Their scales do not track perfectly, for practical and fundamental reasons.
It's worth while reading over this Wiki article.

Personally, I question the whole idea of an overall 'linear' temperature scale because it ignores of the 'effort' involved in getting down near 0K. Temperatures that we come across in our lives fit a linear scale fine but, as with 'relativistic' speeds, things change at the bottom end of the Temperature scale. But I would not like to be the one to decide at which temperature the scale should start to be defined in a non-linear way. (Total nonsense or worth thinking about?)

 

[Zeke137]

It's worth while reading over this Wiki article.

The Wiki link doesn't work :(...

Temperatures that we come across in our lives fit a linear scale fine but, as with 'relativistic' speeds, things change at the bottom end of the Temperature scale.

I'd often mused on that myself, what with recent attempts to achieve temperatures as close as a few tenths, hundredths or even thousandths of a Kelvin degree to 0K, absolute zero.

But I would not like to be the one to decide at which temperature the scale should start to be defined in a non-linear way

I had thought the triple point of helium might have been a good place to start, but on closer reading, find that helium has four triple points! Maybe the helium lambda point?

 

[sophiecentaur]

The Wiki link doesn't work :(...

Dunno what went wrong there - sorry. Try again

As for choosing what sort of scale would be best for near 0K temperatures, I have no idea. Expect there are many opinions from qualified people. Otoh, they may just say that the K scale is fine and to stop whining!