"Greg Cook" wrote in message Exactly right.
Crystal formation is actually a very difficult process.
Crystals begin to form and grow on defects in glass or on an already
crystalline particle of the substance. If you have a completely smooth
surface, crystals will not precipitate out. Thus, by adding some potassium
bitartrate, you are Seeding your wine to initiate crystallization of the
bitartrates. The second point perhaps has to do with having a high
concentration of fine particles for the crystals to grow on.
Sorry, should have explained it all, I forget that people on this forum
usually also want to know why!
Greg is correct - work sone by Roger Boulton and others has shown that gross
crystal growth of KHT is dependent on:
- Degree of agitation
- Number of initial crystals
- Difference between current and equilibrium concentration (equilibrium
concentration is temperature dependent, so chilling more increases the
driving force)
And a few other things that escape me right now!

These happen to be the things we can affect in the winery. It is
theoretically possible to cold-stabilise a wine in as little as four hours,
in a process known as the contact process. You can do it in less time if you
want to cut corners, but most techniques for checking cold stability take
longer than that anyway! The process chills down the wine to -4°C, chucks in
4g/L of crystals, and stirs it vigorously.