OK, but does Grandpa's knee ache, too?

A reader writes:

This page (via this page via this page via this page...) says that if it's going to rain, the surface of a parallel-walled cup of strong coffee will be slightly convex, with the bubbles in the middle. If it's not going to rain, the bubbles will be around the edge. Apparently this has something to do with atmospheric pressure. I am skeptical.


I'm skeptical, too. I can't imagine how this is supposed to work.

Any ordinary liquid (which is to say, not liquid helium, supercritical carbon dioxide or other such substances not easy to find at the supermarket) has surface tension, which causes it to form a meniscus, a curved surface, when put in a container.

If the molecules of the liquid stick to each other better than they stick to the material the container is made from, the meniscus will be convex, higher in the middle. If the liquid molecules stick to the container material better than to each other, the meniscus will be concave, lower in the middle and higher around the edge.

Water in most kinds of household cup or glass forms a concave meniscus; water in a silicone cup forms a convex one. Coffee behaves much the same, as far as I can tell; foam or crema or whatever could be piled up in different ways, and really strong coffee might be oily enough to give a concave meniscus in almost any container, but that's the extent of the differences as far as I can tell.

Weather is definitely related to atmospheric pressure, and to relative humidity, for that matter. Falling pressure and rising humidity generally indicate a higher probability of rain. But pressure and humidity won't have any effect on the behaviour of a liquid in an open container, unless the pressure is so low that the liquid starts to boil at the ambient temperature. If the liquid is water then it'll evaporate faster when the humidity is low and not at all if the humidity is 100% (or higher).

One thing definitely does affect the distribution of bubbles on top of a cup of coffee, though; it's called a teaspoon. If you stir your coffee round and round, the bubbles will pile up in the middle. If you don't, they'll probably stick to the edges.

I think the bubbles ending up in the middle when the liquid is spinning is analogous to the behaviour of similarly spun flames. If you make an apparatus that can spin candles on a platter or arm while shielding them from the wind of their movement...


...their flames bend inwards. Centrifugal force makes them bend in, not out, for the same reason the undisturbed flames go up, not down; the hot flames are lighter than the air surrounding them. Helium balloons behave the same way, but the rig to demonstrate it is more cumbersome.

(The above is an unusual version of this classic physics demonstration, which is usually done with a two-candle apparatus that looks more like this.)

If the weather-predicting coffee is meant to operate by mystic unknown forces, like the much weirder "storm glass", then of course observing that normal atmospheric pressure variations have no effect on coffee is irrelevant. The burden of proof is on the claimant, though, and this is a pretty extraordinary claim; I'd like to see someone actually test this peculiar alleged phenomenon.

Psycho Science is a regular feature here. Ask me your science questions, and I'll answer them. Probably.

And then commenters will, I hope, correct at least the most obvious flaws in my answer.

4 Responses to “OK, but does Grandpa's knee ache, too?”

  1. ix Says:

    Concerning the title, it always seemed to me that there might be something to that. It's also pretty universal that people complain about old injuries when the weather is somewhat humid/cold. Any idea what the mechanism might be there?

    • Stark Says:

      There is some evidence that barometric pressure can have an effect on folks with arthritis. When the pressure drops, as before a big storm, the already inflamed tissues of arthritic joints can indeed inflame a bit more - and since the folks are already experiencing inflammation induced pain the change could be registered as a bit more pain.

      However, there is relatively little actual research on the subject, largely due to their being no reason to do the research. It would serve no purpose - you aren't likely to develop a treatment regimen based on the variability of the weather after all.

      I do have a first hand anecdote of barometric pressure relieving some arthritis pain though. I scuba dive and have had the unfortunate experience of one of my dive buddies getting bent (nitrogen bubbles in the blood - no fun, potentially fatal) for which the treatment is an extended ride in a decompression chamber. This particular person has moderate arthritis in his left wrist. He always said it didn't bother him near so much when at depth - but he never was sure if the reduction in pain was the pressure or the temperature or just his imagination. Whilst in the deco chamber for 14 hours he had a good chance to test things and determined that under pressure the arthritis was indeed lessened... which makes sense since arthritis pain is caused by inflammation and high pressure should help a bit with that issue. Still, the effect was only temporary... and 14 hours in a deco chamber is the definition of "No Fun At ALL" so he doesn't recommend it as a treatment.

  2. matt t Says:

    A simple way to demonstrate the unintuitive movements of the lighter-than-air object is a helium balloon in a car - suspend it above a seat, and go for a drive.

    As you accelerate, the balloon leans forward, as you brake it moves backwards. As you corner left, it leans left. The effect is quite remarkable, and stronger than you might expect.

    But be warned - this experiment will tempt you to accelerate, brake, and corner like a hoon. :)

  3. pjcamp Says:

    Oh dear.


    1. The surface of a liquid is always in equilibrium with the atmosphere. Barometric pressure cannot change fast enough for it to not be a quasiequilibrium process. This means the shape of the meniscus depends entirely on the interaction between the fluid and the container.

    2. There's no such thing as centrifugal force. All forces come from interactions between objects. Can you see the second object? Neither can I. There is a centrifugal effect but it is an artifact of viewing a process from a rotating reference frame. It give rise mathematically to something that looks like a force and is often called "centrifugal force" but is actually a component of the acceleration. Similarly for "Coriolis force" which also doesn't exist.

    If you want an inertial explanation of this (and the helium balloon effect as well) you have to ask yourself why balloons float and flames rise in the first place. Both are due to a buoyant force driven by a density difference and gravity.

    Buoyant forces exist because of a pressure gradient in the surrounding fluid (in this case, the atmosphere). Consider balloons. Because of the difference in height between the top of the balloon and the bottom, the pressure on the top is slightly less than the pressure on the bottom. Gravity pulls air molecules to the ground, but their thermal motion means they don't all stay there all the time. They are, however, concentrated toward the ground. This means there is a net upward force because more air molecules collide with the bottom of the balloon than with the top (a buoyant force -- note: interaction between balloon and air molecules so a real force). If that force is greater than the gravitational force on the balloon, it floats.

    Similarly with flames. The heat of the flame reduces the density of the gases within it. Think of it as a balloon without a shell. Buoyant force drives the hot air upward. Voila. Convection. This is why flames in space are spherical. They're still plenty hot but being in free fall, the atmospheric pressure gradient has gone away.

    Now suppose you could create a horizontal pressure gradient in the air. Then you would have a horizontal component to the buoyant force. The most straightforward way to do this is to tie your helium balloon to the gear shift in your car and tromp on the acceleration. The air in the car, obeying Newton's first law, stays where it is until something (the back of the car) forces it forward. There is therefore a back to front pressure gradient and the balloon moves forward, just the opposite of the fuzzy dice hanging from the mirror.

    In this case, the acceleration is centripetal, caused by the rotation of the platform. Air isn't very viscous but it isn't nonviscous (else airplanes would not fly -- story for another day). So it does rotate along with the container with a reasonable delay (most of the air initially obeys Newton's First Law but not forever). In order to turn in a circle, something must provide the air with a centripetal ("toward the center") acceleration. The thing pushing toward the center is the wall of the container, the only thing the air is interacting with. Voila. Horizontal pressure gradient, largest toward the outside edge so horizontal component of the buoyant force, directed toward the center.

Leave a Reply