> This suggests that the six fold symmetry is the consequence of EM
> forces on the surface of the generating molecules that link together
> as a planar tessellation, with secondary perpendicular linkages that
> create another plane either above or below the inital lattice.
>
> But such sheet like constructions have no intrinsic center. I would
> suggest (I would like to believe) that some how the gravitational vector
> acts as an axial center, which despite all the variations of shape,
> all these shapes have in common.
In principle a sheet has no center, but a hexagon or a symmetrical tiling of hexagons does, and I imagine a crystal growing from a nucleus/seed in the absence of gravity would take a relatively "pure" form of that crystal. I recall that making high-quality crystals in microgravity was one of the excuses for the space station.
I don't see how the gravitational vector could act as a center, since it exists througout the crystal.
Another thing to consider is that near-microscopic snowflake in a turbulent atmosphere would have a lot of surface friction for it's mass, rendering it something like weightless. Plankton sometimes develop similarly high-surface area shapes to hinder sinking out of the zone where they can pick up sunlight.
So I don't see how ice crystals would be significantly different in the absence of gravity, esp. when snowflakes seem to reflect ice's geometry at a molecular level.
I don't have much success googling for examples of ice crystals grown in orbit.
> ``The growth of snow crystals depends on a balance between faceting (see
> Crystal Faceting) and branching. Faceting tends to make simple flat
> surfaces, while branching tends to make more complex structures. The
> interplay between faceting and branching is a delicate one, depending
> strongly on things like temperature and humidity. This means snow
> crystals can grow in many different ways, resulting in the great
> diversity we see in snow crystal forms.''
>
> All well and good. But why don't the global results of these
> interplays resemble feathers for example, instead of radially symetric
> shapes?
Perhaps because they're effectively weightless?
Also, think of how ice crystals form in the presence of certain constraints, like on a windowpane. Kinda like feathers sometimes, no?
-- Andy