On: Snowflake macro photography (2013)

[allenz]

The edges grow much faster because "molecularly flat regions... have fewer dangling chemical bonds and are thus less favorable attachment sites [for condensing molecules]." Also, snowflakes can take on 3D column shapes depending on humidity and temperature. See http://www.its.caltech.edu/~atomic/publist/AmSci2007.pdf

[adrianratnapala → allenz]

Right, but from that we might also expect very long thin hexagonal crystals. Maybe that is what rime is?

[jamie_ca → adrianratnapala]

There are some shots of those near the tail end of the article. Both pure cylinders (https://c6.staticflickr.com/9/8240/8661684397_3e2953b2ca_b.j...) and capped columns (https://c8.staticflickr.com/6/5613/29417525663_448be7dac8_o....).