3 ms·
This SciAm article ( http://www.scientificamerican.com/article/why-do-snowflakes-crystal/ http://www.scientificamerican.com/article/why-do-snowflakes-... ) help
by andrewflnr 11y ago
This SciAm article ( http://www.scientificamerican.com/article/why-do-snowflakes-crystal/ http://www.scientificamerican.com/article/why-do-snowflakes-... ) helped me. It's hard to summarize, but it looks like branching starts (in certain conditions) on corners and amplifies existing projections. Corners start hexagonally symmetrical, and to the extent that the formation of "dendrites" proceeds in lockstep around the crystal (due to uniform conditions?), they stay symmetrical. The amplification effect is weaker for the smaller sub-branches, so those are somewhat but not always symmetrical.
It doesn't help that there's a couple missing diagrams, but I think the first one is just a hexagonal prism and I didn't really need the second.