Libbrecht, the chair of the physics department at Caltech, says he was drawn into this frosty work when he read a reference to the physics of ice crystal formation. "I realized I didn't understand how these things work very well, and then I looked it up, and realized that no one understood how they worked," he says.

He started trying to figure out why crystals grow into such dramatically different shapes in different atmospheric conditions, and, well, "It just sort of snowballed, you could say." Now his Caltech lab spends considerable time growing "synthetic snowflakes" under different conditions (the results can be ogled on Libbrecht's Web page). These needles are slender hexagonal columns that form at mid-humidity and relatively warm temperatures, around 23 degrees Fahrenheit.