Hi Phil

For a bubble of gas to exist, its internal gas pressure P must exceed P=T/R, where R is the spherical bubble radius and T is the "surface tension" (intermolecular attraction) of the liquid.

So it seems that bubbles cannot form - for at zero radius, infinite pressure is required. But of course the "large number of gas molecules" theory that gives that formula applies only once there are millions of gas molecules (and liquid too, of course).

In practice, a surface is required to "seed" the bubble - for example: from monolayers of gas physically ADsorbed (Van der Waals forces etc) onto a small particle - especially a corner or flat face of such a particle. In this case, the glass surface serves to seed the bubble.

Why preferentially at the bottom of the flute? For some sparkling wines, that is where the microparticles have sunk to. It is also where the hydrostatic pressure is highest.

Experiments show that the solubility of a gas (CO2 and air, in the case of champagne) increases with pressure and decreases with temperature. If the champagne is served chilled, the bottom of the flute (being the most narrow point) heats up quickest, making the dissolved gas less soluble in the liquid, increasing its "pressure" to that critical bubble-formation pressure.

As poured from the bottle, champagne is already supersaturated (unstable) with CO2 and air. The fierce bubbling initially is due to turbulence - which causes microbubbles to form. The smaller these microbubbles are, the SLOWER they rise to the surface. Stokes formula tells us the force to drag a sphere at speed V thru a liquid, viscosity, t, is 6Pi times tRV (see Stokes Formula).

So the bigger bubbles rise rapidly, soon being lost as seeding centers. The tiniest bubbles are still at the bottom, not yet having risen, where the hydrostatic pressure actually makes them a little smaller yet. CO2, especially cold CO2, is highly compressible - a very "imperfect gas" (good refrigerant).

The hydrostatic pressure of a bubble of given radius R can only exist (not shrink) if its internal pressure exceeds T/R PLUS the hydrostatic pressure. But the "gas" dissolved in the liquid is already at that "Pressure". So I doubt that the champagne depth has any effect to discourage CO2 bubble formation - only the encouragements mentioned above.

I think there is an urban myth attached to fizzy drink; namely that shaking the bottle increases the pressure. All it does is fills the liquid with microbubbles (of the free compressed gas in the bottle), making it ready to foam immediately on release of that pressure. It would be easy to test this by attaching a pressure gauge to a soda bottle and shaking it (but not enough to warm it!).

John