mental_floss magazine
SUBSCRIBE >
GIFT SUBSCRIPTIONS >
DIGITAL SUBSCRIPTIONS >
subscriber services >
Up in the winter sky, water vapor in a cloud condenses into a droplet and freezes into a tiny bit of ice, with the water molecules bonding together as a hexagonal crystalline lattice with a six-fold symmetry. As water vapor condenses on its surfaces, the ice crystal grows into a hexagonal prism. As the crystal gets larger and larger, branches begin to form at the corners of the hexagon. When the crystal is heavy enough, it falls through the atmosphere toward the ground, where we call it a snowflake.
Many of those snowflakes have fallen onto the small town of Jericho, Vermont, the home of Wilson Alwyn Bentley. As a teenager, Bentley became interested in snowflakes, and he attempted to draw them while looking at them through a microscope his mother had given him. He found that he couldn’t get the complex structures of the flakes down on paper before they melted, so he attached a camera to a microscope using an adjustable bellows mechanism and photographed his first snowflake on January 15, 1885.
Over the next few decades, Bentley continued to study snowflakes, taking 5,381 photographs of them and developing a system to categorize over 80 different flake types and shapes. In 1920, he became a Fellow of the American Meteorological Society and was awarded the Society’s first research grant ($25). Bentley sometimes told people that he had never seen two snowflakes that looked alike and published several magazine articles arguing that no two flakes are identical. That idea stuck in the public imagination, which brings us to today’s question: was he right?
Scientists have discovered that as an ice crystal gets blown around in the air while it grows, the environmental conditions it is exposed to and the timing of the exposure determine the shape of the snowflake. With different factors determining the snowflake’s shape, and that shape changing as the growing snowflake moves through different conditions, you get a lot of variety in snowflake shape. Here’s a handy little graph from a Caltech physics professor that shows which shapes occur in which conditions:
If two growing snowflakes are exposed to the same temperatures and humidity and water saturation levels at the exact same time (live out the exact same lives, if you will), they may look exactly alike at the macroscopic level. In fact, in 1988, the Nancy Knight was studying snowflakes as part of her work with the National Center for Atmospheric Research and found two identical snowflakes of the hollow column type in a Wisconsin snowstorm.
But Caltech physics professor and snowflake expert Kenneth Libbrecht (the man who made the above graph) points out that if you look at any two flakes – even seemingly identical ones – on the atomic level, you’ll find numbers of water molecules and different layouts of those molecules (most water molecules contain an oxygen atom of 16O, but one molecule in every 500 has an 18O). One thing you won’t find? Two snowflakes that are exactly alike.
If you’ve got a burning question that you’d like to see answered here, shoot me an email at flossymatt (at) gmail.com. Twitter users can also make nice with me and ask me questions there. Be sure to give me your name and location (and a link, if you want) so I can give you a little shout out.
Comments (19)
Comment
My seven year old daughter made a comment to me about no snowflakes being alike yesterday. It’s nice to know that the little facts she finds interesting and is taught are actually true, whether or not she ever understands the complex meaning behind it.
Nice article.
posted by Nicole on 12-6-2008 at 10:21 am
This is exactly how mental_floss sucked me in the first time I came across the page while scanning CNN.com for a decent read. I don’t remember the first article, and it doesn’t really matter. Once I found this wonderful site, I kept coming back. Thank you mental_floss. Keep up the fantastic breaks from the day to day dulls of the News and Information world. Knowledge is Power!
posted by Sponge-Brain on 12-6-2008 at 11:18 am
Great article, Matt!
I didn’t know that the ones that hit your face and sting might actually be shaped like needles! I thought they just felt like it. I guess because when it is those conditions, who really feels like taking time to examine the flake?
We’re supposed to get snow tonight and tomorrow. Wonder what kind of flakes we’ll get? My guess is sectored plates or dendrites (lake effect=high saturation?). If I can work that into a conversation today, I’m going to sound incredibly intelligent. :)
posted by kate on 12-6-2008 at 12:11 pm
I have tertiary syphilis.
posted by Brad on 12-6-2008 at 1:43 pm
Below is an excellent site on snow/ice crystals.
posted by mike on 12-6-2008 at 3:26 pm
Thanks! Great article. :)
And what’s up with Brad’s comment? Can’t you guys delete crap like that?
posted by Dawn on 12-6-2008 at 5:57 pm
Good point, Kate . . . I know the sensation you’re talking about, and it’s going to feel all the more intense next time because I’ll be visualizing all these icy needled spikes jabbing at my face! Ha ha, yikes!
posted by kikadee on 12-6-2008 at 6:21 pm
I live only a few miles from Jericho, and I’m pretty sure there’s a little Snowflake Bentley museum up there. Always meant to check it out. (the old never-been-to-the-local-tourist-attractions routine, and the recaptcha is “traveled one”)
posted by steph on 12-6-2008 at 9:17 pm
Ha ha – this is awesome. I used to joke around with people and say that I recently read about scientists who actually discovered two snowflakes that were alike. (As if anyone is actually looking)
Anyway, if you follow the logic of the Caltech professor that on “the molecular level” they are different, then you could say that no two of anything are the same (with the exception of very basic molecules)
posted by Benji on 12-7-2008 at 3:39 am
what are snowflakes?
posted by southern california guy on 12-7-2008 at 3:55 am
Fascinating short article, thank you!
Something I’ve always wondered – when snowflakes with apparent symmetry are studied, is the symmetry perfect all or much of the time? The picture above shows perfectly symmetrical dendrites and sectored plates. I can understand that the physical conditions (temperature, pressure etc) are identical across the snowflake when it is being formed, which might result in uniform growth, but if perfect symmetry of shape is often seen, how does the snowflake ‘know’ what is happening to it as a whole so that it can ‘make’ the growth symmetric? If, as I guess, it doesn’t, are such ‘perfect’ snowflakes the exception rather than the rule? Are snowflakes with fractured symmetry seen?
posted by JM on 12-7-2008 at 5:13 am
I don’t think this is really accurate. True, no snowflake will be identical when comparing them on an atomic level.
But then, you could say no 2 leaves would ever be identical on the atomic level. Or rocks. Or ice cubes.
posted by John on 12-7-2008 at 5:45 am
I’m sorry, but that just seems a bit silly to point out that any structure, no matter how similar at first glance, is actually constructed differently at an atomic level. I’d be interested in knowing what structures *would* be *exactly* alike, once you take it down to an atomic level.
posted by Kell on 12-7-2008 at 5:50 am
Is it true that no two snowflakes are alike | CommentURL.com…
\r\nUp in the winter sky, water vapor in a cloud condenses into a droplet and freezes…
posted by CommentURL.com | A world of interesting web pages on 12-7-2008 at 6:18 am
i think it’s all about chaos theory…no two can be alike!
posted by J. Acai on 12-7-2008 at 6:56 am
I’m pretty sure you know very well that
the www is the greatest technological creation in modern history. kosher-restaurants.tarew.co.cc
posted by tanys on 12-7-2008 at 7:10 am
John and Kell have it exactly right. It’s absurd to compare anything at the atomic level. I’d dare say that nothing larger than 1mm cubic volume could ever be exactly alike, man or nature made.
posted by Bob on 12-7-2008 at 2:30 pm
Hey y’all
Let me try this counter argument that there have been two snowflakes that are exactly the same on a atomic level.
Suppose there is a mole of atom(10^23) in a millimeter of water and say that this is the size of an average snowflake. We make the simplifying assumption that the this represents on the order or the number of different atomic configurations.
The size of antartica is is 14million kilometers. So if we suppose that it snows once a year covering the area of antartica then you get something on the order 1.4 × 10^19. Also suppose that the average age of the earth is measured on the order of billions of years. Add another 9 to the 10^19 and you get 10^28. Of course the number of combinations of 10^23 is much much bigger. But we do get close to it. If you consider the possibility of snow bearing planets you get even closer.
The fact that you get similar configurations on a macroscopic level tends to imply a much smaller search space than is usually assumed.
-opengre
posted by gary on 12-7-2008 at 9:40 pm
I assume it is theoretically possbile for two snowflakes to be the same, aslthough I would not like to go look for them. I also assume it is theoretically possible for two people to have the same left index-finger fingerprint, but I would not like to go look for them either.
posted by Amy W on 12-9-2008 at 7:04 am