Why doesn't the liquid in a liquid-in-glass thermometer slide down when you turn it upside-down?

Question

This question applies to both mercury and alcohol thermometers. I assume the answers would be similar, but maybe not. My understanding is that the capillary in a liquid-in-glass thermometer consists of a liquid and a vacuum on the other end. However, if there is a vacuum, then wouldn't the liquid easily slide down when you turn the thermometer upside-down? That doesn't happen from what I remember (back in the old days when those were popular).

The main problem is that I feel there can be many false explanations here. For example, maybe the empty end is a partial vacuum and that pushes the liquid to one end. I don't know. Maybe the liquid has some surface tension and that prevents it from sliding down. Maybe its viscosity is very low. This site (archived link here) says,

Mercury is attracted only weakly to glass, so it doesn’t really adhere to the walls of its channel.

but I just don't know who or what to trust here. Can anyone give me an answer based on actual physics to back up?

Specific citations of values of properties and/or verifiable calculations are highly preferrable. No speculations please!

Answer 1

The movement of liquids in narrow tubes is governed by the balance between accelerations and gravity (which wants to pull the liquid in the "down" direction) and surface tension effects (which make the liquid want to stick, or not, to the tube walls). The viscosity of the liquid couples to the diameter of the tube to set the overall time scale for establishment of that balance. Small diameters and large viscosity slow things down, and vice versa.

These effects are in turn governed by something called the Weber number which is a dimensionless ratio containing inertial effects in the numerator and surface tension forces in the denominator. Small Weber numbers mean surface tension effects are in control. Big Weber numbers mean inertial effects are in control.

When gravity dominates, the liquid will flow out the bottom of the tube as air flows in to take its place when you tip the tube upright or shake it. When surface tension dominates, you can turn the tube any which way you want and shake it, and the liquid will not slide out past the air and come out its end.

So then, bulb thermometers have a bore size chosen to prevent air from coexisting with liquid within the tube, and to minimize gravity and acceleration effects to yield a very small Weber number.

Resetting a bulb thermometer to zero can be done by large accelerations (whipping the thermometer around by flicking your wrist) which overwhelm the adhesion forces, but dropping the thermometer against a hard surface creates accelerations strong enough to allow mixing of the air and the liquid in the bore, producing separate "slugs" of liquid distributed along the bore length.