> > As for the water level in the beaker: as the temperature drops, the water level also should drop -- technically speaking -- because the average density of the water-in-water system is decreasing. And as Don says, the water level will rise once more after the ice is completely melted. However, for most practical purposes, the water level does not change. That's because the density of water at 25ºC is 0.9970479 g/cm3; at 17ºC it's 0.9987779; and at 3.98ºC it's 1.0000 g/cm3. At 0ºC it has slightly decreased again to 0.9998425 g/cm3. These differences in density are so miniscule that to the naked eye there will be no change in water level. The only big difference occurs when all the water actually freezes into pure ice at 0ºC, where it has a expanded and lighter density of 0.917 g/cm3.
>
> Yes, that was something I wished to emphasize a little more in my original post-- the drop in water level due to the density change will be VERY slight. I like the word "slight", but at times like this, it does not convery the meaning very well. Words like "miniscule" (as Wolfspirit said) or "infinitesmal" or "invisible to the naked eye" (I don't care if that one is four words) are better ways of describing the drop in water level. [...]

Well, I forgot emphasize that water is a very anomalous substance -- it has discontinuous volume changes during its cooling, causing its solid form to be lighter than its liquid form. The reason it does this is due to its high affinity for aligning hydrogen and oxygen atoms into very regular hydrogen-bonding pairs. These regular 'H-bonded' molecules take on an ice-crystal lattice structure that's similar to diamond, and the entire structure spreads out over a greater space (volume) for a given amount of mass. In other words, ice is bigger but less dense. If ice does not have room for expansion when it is created, it may rupture its container.