171

THE MICROSCOPE • Vol. 62:4, pp 171–181 (2014)

The Hidden Secrets of Snowflakes

C R I T I C A L C R I T I C A L C R I T I C A L C R I T I C A L C R I T I C A L FOCUSFOCUSFOCUSFOCUSFOCUSBrian J. Ford

Snowflakes have an intrinsic beauty andhave been studied for over 450 years, yetmany of the pioneer researchers of snowcrystals are unknown to modern science.

Snow swirled aroundmy sleeves as I stepped

out from my rented roomin Prague and climbed intoa tram. I alighted near themuseums and spent a fewhours in research before Iscrunched through thegathering drifts past the church of St. Francis of Assisiand down to the Charles Bridge. It is an awe-inspiringstructure, lined by gothic statues (now all replicas; theoriginals have been removed for safe keeping) andmarked by vast towers at either end. Snow was fall-ing heavily as I stopped to look at the view.

My mind drifted back more than four centuriesearlier, when a distinguished philosopher had walkedacross that same snowy bridge in winter. He not onlylooked at the snowflakes on his lapel, but set out tospeculate on how they formed. He was JohannesKepler, renowned to us now as a pioneering astrono-mer. Kepler had been born near Stuttgart, Germanyin 1571, and at the age of six his mother took him up anearby hill to watch the great comet of 1577. Threeyears later he wrote about the striking red color of themoon during a total lunar eclipse, although he wasnot yet 10 years old. The young Kepler’s hands weredamaged by an attack of smallpox, which limited hisability to use telescopes, but he became fascinatedby the quest for a mathematical explanation of theuniverse.

He was transfixed by the power of the lens, and

was diverted by the studyof small objects, as well asthose that lay at incalcu-lable distances from us. Itwas in 1610 that Keplerhad walked across theCharles Bridge. He was 39,and as the feathery snow-

flakes settled on his coat he peered at them inquisi-tively. Kepler’s attention was captured by their ex-quisite structure and, being of a mathematical turn ofmind, he mused on how the packing of atoms mightrelate to the six-rayed structure of each snowflake. Hehad published his great astronomical work MysteriumCosmographicum in 1596, and now he resolved to writean account of snowflakes. In 1611, it became a booklettitled De Nive Sexangula and was presented to his pa-tron Baron Johannes Matthäus Wackher vonWackenfels as a holiday present. People often get thedetails wrong. The work is often described as“Strenasue de nive sexangula,” but that should be twowords, Strena sue; “Strena” means a New Year’s gift.

Kepler speculated: “There must be a cause whysnow has the shape of a six-cornered starlet. It cannotbe chance. Why always six? The cause is not to befound in the material, for fluids are formless and flow,but in an agent.” His “agent” was the physical natureof matter — how small atomic elements of waterpacked together to make ice. He noted that the pack-ing would always be “the tightest possible, so that inno other arrangement could more pellets be forced

172 THE MICROSCOPE 62 (2014)

into the same container.” Kepler had been consideringthe mathematics of the packing of spheres, for the En-glish mathematician Thomas Harriot had written tohim around 1608 about the way cannon balls werestacked aboard ship. Harriot was the navigator forSir Walter Raleigh’s pioneering voyages to the NewWorld in 1584–5 and had been asked by Raleigh toinvestigate the most efficient way to pack cannon ballsfor use against rivals. This was the model that Keplerused. Curiously, Kepler did not describe the structureof snow, nor did he picture them in his booklet.

The first portrayal of a six-rayed snowflake I havelocated was published in 1555 by Olaus Magnus, theexiled archbishop of Uppsala, Sweden, in his Historia

de gentibus septentrionalibus (a history of the northerntribes). The book contained a section titled “de variisfiguris nivium” — the various figures of snow — inwhich, for the first time, the appearance of snow-flakes was recorded. I say “recorded” with a wry smilebecause most of the illustrations are pure invention:His plate showed snowflakes looking like a tiny eyeor a miniature hand; one looks like a bell, another likethe crescent moon. But one is historic. Halfway acrossthe lower portion of the plate is an image of a perfecttiny star. It is an unmistakable six-rayed snowflakeshown for the very first time in history. I revealed allthis in my book Images of Science (1993), and it hassince been widely reported. I still believe it is the ear-liest record of a “microscopic” subject in the historyof science.

FIRST GLIMPSE AT REALITY

It fell to the French philosopher René Descartes toportray snowflakes with some semblance of realismin 1637. Descartes read Kepler’s work and published aplate illustrating a variety of snowflakes. It was in-cluded in his Discours de la Méthode — his discourse onthe method of reasoning to find scientific truth —where he wrote: “These were little plates of ice, veryflat, very polished, very transparent, about the thick-ness of a sheet of rather thick paper [and] so perfectlyformed in hexagons, and of which the six sides wereso straight, and the six angles so equal, that it is im-possible for men to make anything so exact.” The draw-ing shows some imaginary snowflakes with up to 18branches, but most show six sides, as they do in thereal world. Two of them (labeled F in his figure) showtwo crystals connected by an axle, and — unlikely asthey seem — crystals of this sort do occur in nature.Wrote Descartes: “I only had difficulty to imagine whatcould have formed and made so exactly symmetricalthese six teeth around each grain in the midst of freeair and during the agitation of a very strong wind,until I finally considered that this wind had easilybeen able to carry some of these grains to the bottomor to the top of some cloud, and hold them there, be-cause they were rather small; and that there they wereobliged to arrange themselves in such a way that eachwas surrounded by six others in the same plane, fol-lowing the ordinary order of nature.”

But Descartes skated round the essential question.He wrote that the snowflakes “were obliged to ar-range themselves in such a way that each was sur-rounded by six others in the same plane” and con-cluded that this was “following the ordinary order of

Olaus Magnus of Sweden published this snowflake study in 1555.The shapes are unreal, except for the one that shows a six-rayedflake. He was a talented man and a great traveler. In 1539, heproduced the first map of Scandinavia to show the coast andIceland in a form we would recognize today.

René Descartes published these snowflakes in 1637 in Discoursde la Méthode, which offered the first accurate view of snowcrystals. Many of the multi-branched types are exaggerated, butthe six-rayed flakes are relatively true to life, and the paired flakesthat are joined by an axle (upper right) are accurately portrayed.

173

CRITICAL FOCUS | BRIAN J. FORD

nature.” This is a cop-out: There should be a reasonother than the “order of nature.” Kepler had specu-lated on some kind of atomic packing, but the ideawas not taken up by that arch-rationalist Descartes.And so this crucial concept was ignored until 1660,when Thomas Bartholin turned to the investigationof snowflakes.

Bartholin’s account was published in a paper, Denivus usu medico observationes varium (on various medicaluses of snow), addressed to his brother Erasmus. Herewe had images of snowflakes made from low-powermicroscopy, which were recognizable, even thoughthey were drawn as cartoons.

He produced neat diagrams of atomic packingwhich you would not guess were published as earlyas 1660, for they look strangely modern. Bartholin usedthese as evidence for the six-rayed structure of a snow-flake, and his work had resonances through the re-search of a French priest René Just Haüy, known to-day as the father of crystallography. His first greatwork was the Traité de minéralogie published in 1784. Hereferred to the ideas of Kepler, whose ideas on closepacking of spheres became known as Kepler’s Conjec-ture. Although the idea of stacked spheres underpin-ning crystal structure was becoming popular, the proofof the concept using computational mathematics wasnot published until 2005 (Annals of Mathematics, 162,pp 1065–1185).

HOOKE ORIGINALS?

Look into the standard accounts and you will findthe 29-year-old Robert Hooke praised as the first per-son to study snowflakes successfully with a micro-scope. Yet his famous drawings were not original —Hooke flagrantly copied them from Bartholin’s earlierwork. I presented this conclusion to a startled audi-ence at the 2009 Inter/Micro conference in Chicago andagain as my presidential lecture to the CambridgeSociety for the Application of Research in October 2009.And the topic becomes more intriguing when we lookfurther in to Robert Hooke’s great work Micrographia(1665). His Observation XIII, under the heading “ofsmall diamants,” deals with crystal structure. Howdoes Hooke rationalize the regular predictability ofcrystals? Why, by considering their structure aspacked spheres, much as Kepler and Bartholin haddone before him. His diagram in Figure VII showstightly packed spheres (here he calls them bullets, orglobules) and relates that to crystal morphology. It isa neat piece of discussion, marred only by Hooke’s de-cision not to give due credit to those whose work he

Thomas Bartholin, a Danish physician and mathematician,produced fanciful engravings of snowflakes. He modeled snowcrystals in terms of assemblages of microscopic spheres as Keplerhad tried to do.

The pioneering scientist Robert Hooke complained about others(notably Isaac Newton) plagiarizing his work. But in his snowflakestudies published in 1665, he clearly plagiarizes Bartholin’sengravings. Hooke also discusses crystal structure, using Kepler’sconjecture and Bartholin’s ideas but gives neither of them credit.

174 THE MICROSCOPE 62 (2014)

was repeating. A decade later in 1675, Dr. FriedrichMartens, a German physician, classified snow crys-tals into 24 discrete types. Then a Tuscan mathemati-cian and philosopher, Donato Rossetti, published hisLa figura della neve (the figures of snow) in 1681. My latecolleague Joseph Needham of Cambridge Universitypointed out that his images of hexagonal snow crys-tal plates were ground-breaking.

It was from the Netherlands that research intosnowflakes next emerged and, for the first time, wewere approaching the real appearance of a snowflakein nature. A Dutch natural historian, Jan Floris Marti-net, published his Katechismus der Natuur (1779) in fourvolumes in Amsterdam. His work is largely forgotten(you will not find Martinet in the standard reference

sources) but his eye for detail is worth commemorat-ing. So diligent was he that his book was reprinted forover a century after his death — and among his lega-cies were some good studies of snowflakes. Yes, manyof his images of “sneeuw” are fanciful, but there ishere a hint of accuracy in many of them. His book wastranslated into several languages including French,German and Spanish — and then it appeared in a Japa-nese edition.

We tend to think that there was little contact be-tween the Far East and the western world at that time,but the Netherlands remained a focus of interest forthe Japanese, and Martinet’s Katechismus der Natuur wasa popular read amongst natural philosophers (the titlebecame Kakuchi mondo in Japanese). One leading Japa-nese artist was Shiba Kokan, who specialized in com-missions that would emulate the Dutch masters, andhe became the first Japanese to include snowflakes inhis drawings. Another Japanese enthusiast was DoiToshitsura who later became Lord of the Koga Domainin Shimousa Province and held high office with theShogun authorities. Around 1800, still in his mid-teens,he began to study snow crystals and drew them neatlywith pen and ink. Then in 1833, there was a cata-strophic collapse in the Japanese food supply from theTenpo Famine, caused by a severe winter. One ofToshitsura’s entries in his research papers records thefact: “On Jan. 29, 1833, there was a heavy snow andthe snow crystals could be very distinctly seen andsome exhibited shapes not seen in other years.” Hewas based around Osaka at the time, which rarelyhad very low temperatures. Snowflakes need a tem-perature of 25° F or colder to be observed under a mi-croscope, which the winters in Osaka rarely reach. Sothe catastrophe for the community seems to havegiven Toshitsura the stimulus he needed to continuehis observations.

Years later, his research was privately publishedby Doi himself under the title Sekka Zusetsu (1832).It includes 86 observations made by Doi Toshitsuraalong with 12 illustrations from Martinet’s Katechismusder Natuur. Naturally, this limited-edition book had asmall circulation, but the illustrations were copied bySuzuki Bokushi for his book Hokuetsu Seppu (whichmeans “snow stories from North Etsu Province”) pub-lished in 1837. This brought the drawings to a largeraudience, and the two books launched the study ofsnowflakes in Japan. For all the acknowledged influ-ence of Jan Martinet, Doi Toshitsura never gave theimpression that he had really studied snowflakesfor decades. His figures were distorted and more likecartoons.

Snowflakes were first divided into discrete morphological catego-ries by Donato Rossetti of Turin, Italy, who published La figuradella neve in 1681. His diagrams are hardly true to life, but he didmuch to popularize the notion that snowflakes had an intrinsicbeauty. Joseph Needham pointed out that this was the firstportrayal of the hexagonal plate crystals often seen in nature.

175

CRITICAL FOCUS | BRIAN J. FORD

HAND-CUT SNOW

While the Japanese philosophers were publish-ing their own visions of unreality, an English scientistwas about to give the world a glimpse of how snow-flakes really appeared. He was William Scoresby, bornin Yorkshire, England, in 1789. From the age of 11,when he first accompanied his whale-hunting fatherinto the arctic, he longed to be at sea. The youngScoresby studied natural philosophy and chemistryin Edinburgh, and used a portable microscope to ob-serve fresh snowflakes. His results were published in1820 as Snow crystals as observed and drawn by WilliamScoresby, in an account of the Arctic regions with a history anddescription of the northern whale-fishery. His figures are rea-sonably realistic, and he set out to give some order tothe variations of structure found in nature. Scoresbywent on to publish more than 60 papers with the RoyalSociety of London and traveled round the world study-ing the earth’s magnetic field, energy and the socialrelationships of science.

There is one Victorian book that almost capturesthe realism of snow crystals, yet the illustrations wereproduced in an extraordinary fashion. In 1864, Frances

Chickering of Portland, OR, released her remarkablebook, Cloud Crystals: A Snow-Flake Album, published inNew York by D. Appleton & Company. The imageswere made by hand-cutting paper silhouettes, forChickering would observe flakes with a strong handlens and then quickly cut replicas. Her results are notliteral images, but they do convey some of the reali-ties. She was identified in her book only as “A Lady”since publishing was then regarded as a somewhatindelicate indulgence for a woman. In the introduc-tion she explained: “The present collection originatedin the accidental observation of the beauty of a snowcrystal upon a dark window sill … crystals are caughtupon dark fur or cloth, a strong magnifier placed overthem to assist the eye, and the figure immediately cutfrom memory.” Others attempted to emulate Mrs.

Martinet’s book was translated into Japanese, and in Osaka thisprivately printed volume Sekka Zusetsu was produced by DoiToshitsura. It included these cartoons of snowflakes (1832). Most ofthese are fanciful, but his studies caught the attention of others andwere later reprinted. The book is on permanent public display inthe National Museum of Nature and Science in Tokyo.

Jan Floris Martinet was a Dutch theologian and physician, whobecame devoted to the idea of educating the public. In 1777–1779,he published the four volumes of his great Katechismus derNatuur, which included these intricate portrayals of snowflakes.

176 THE MICROSCOPE 62 (2014)

Chickering, though nobody equaled her skill. So at-tractive were her images that a neighborhood printer,J.F. Richardson, produced prints of her images. Theywere printed in sets of seven using deep maroon ink, acurious choice that worked surprisingly well.

It was the dawn of photography that eventuallygave us our first true image of snowflakes. A greatRussian investigator named A.A. Sigson had set out

to systematically photograph images of snowflakesin 1872, and over the following 30 years slowlyamassed a collection of 300 micrographs. He pub-lished them as Snowflakes; micro-photographic images fromnature around 1895.

Sigson’s specimens were obliquely illuminated onthe microscope stage since his aim was to create es-thetically attractive images. The result was the award

William Scoresby produced another attempt to categorize snow-flakes in his book published in 1820. He shows many of the maintypes of structure that we find and sets them out in order. Evenhere, there is so much distortion of reality that the diagrams havelimited scientific merit.

During the Victorian era, hand-cut paper silhouettes were apopular form of portraiture, and Frances Chickering compiled anartistic book of snowflakes made by cutting out the shapes sheobserved with scissors. She was identified only as “A Lady” as theauthor of her work, and the results were so spectacular that theywere reprinted as maroon tinted plates for the cognoscenti.

We will never know who was the first scientist to photograph asnowflake under the microscope, but Russian investigator A.A.Sigson must have been the first to do the job systematically. Startingin 1875, he gradually acquired a collection of artistic micrographs.Many of them won awards and prizes, and they were published inspectacular art nouveau collections as works of art.

177

CRITICAL FOCUS | BRIAN J. FORD

of many prizes, including a gold medal at the Exposi-tion Universelle that took place in Paris from April 15to Nov. 12, 1900. In Germany, micrographs of snow-flakes were taken on Feb. 1, 1879 by Dr. JohannHeinrich Flögel of Ahrensburg, Germany. They turnedup in a dirty cardboard box during the renovation ofFlögel’s house in 1975. The micrographs had been takenat low magnifications (between 16X and 45X) using aburning magnesium ribbon as the light source. Thena collection of micrographs of snow under a micro-scope was published in Germany by RichardNeuhauss as Schneekrystalle: Beobachtungen und Studien(snow crystals: observations and studies) in 1894.

‘SNOWFLAKE’ BENTLEY

And then there came Wilson “Snowflake” Bentley.You were wondering when he would arrive on thescene, and here he is. So many people think first of thisphenomenal farmer that it becomes easy to overlookthose who came before. Everyone seems to say thatBentley was the first person to photograph the crys-tals of snow through a microscope. The statement isall over the web, and even in the resources who shouldknow better. If you check out the Smithsonian’s de-scription, for instance, you will find Bentley authori-tatively described as “the first person to photographa single, unique snowflake.” What nonsense; single,unique snowflakes had likely been photographed by

hundreds of microscopists by this time and severalbooks devoted to snowflakes had been published inseveral countries by the end of the 19th century.

Bentley’s claim to fame is not that he was the firstto image snowflakes through the microscope, nor thathe was the first to publish them or even the first to

Gustav Hellmann, a German meteorologist, asked his technicianRichard Neuhauss to take micrographs of snowflakes in 1892.Their images were inferior to those taken by Wilson “Snowflake”Bentley in the U.S. Hellman later accused Bentley of fraud.

Wilson Bentley (above) ran a farm in Jericho, VT, and begancollecting his matchless snowflake photographs in 1885. Hisresearch was widely published and was ultimately collated in SnowCrystals, a book he compiled with William J. Humphreys. It waspublished in 1931, shortly before Bentley’s death.

Johann Flögel of Ahrensburg, Germany, took his first images ofsnowflakes in 1879. His collections were forgotten at the time of hisdeath and turned up only when his former home was beingrenovated in 1975. The negatives were taken as glass plates andprinted off onto paper. Damp storage conditions caused them tofade; this image was digitally optimized for clearer viewing.

178 THE MICROSCOPE 62 (2014)

write a book about snowflakes. It is that he was sim-ply the best that the world had ever seen. Bentley wasborn in 1865 and raised in Jericho VT, where he hadlittle formal schooling but went into agriculture. Whenhe was 15 his mother gave him a microscope, and itwas the intrinsic beauty of snowflakes that most at-tracted his attention. By the time he was 20, on Jan.15, 1885, he attached a bellows camera to his micro-scope outside in the yard and began to collect photo-micrographs. The quest for new images of what he

called “these tiny miracles of beauty” became a life-long preoccupation, and every winter he would takemore pictures until he had a library of some 5,300glass plates. Some of his earliest images were firstpublished in a scientific paper, “Twenty Years Studyof Snow Crystals,” which appeared in 1901 in theMonthly Weather Review published in Washington D.C.But Bentley did not confine himself to scientific jour-nals. In Harper’s Monthly Magazine (No. 104, December1901) he wrote a popular article called the “Story ofSnow Crystals,” which read: “Quick, the first flakesare coming, the couriers of the coming snowstorm.Open the skylight, and directly under it place the care-fully prepared blackboard, on whose ebony surfacethe most minute form of frozen beauty may be wel-come from cloud-land. The mysteries of the upper airare about to reveal themselves, if our hands are deftand our eyes quick enough.”

It was not just Bentley’s photographic skills thatmattered, so much as the way he treated the nega-tives. He painstakingly removed all the exposed emul-sion around each image, so that the snowflake stoodout dramatically from its background. This is how hecaptured the eye-catching beauty of the crystallinestructure, shining out from the page against a back-ground of velvet black. George Henry Perkins, profes-sor of natural history at the University of Vermont,agreed to co-author a paper with Bentley arguing thatno two snowflakes were ever alike. The idea caughtthe popular imagination and his ideas were publishedin magazines including the National Geographic, Natureand Scientific American, and it was Bentley who con-tributed the entry for “Snow” in the 14th edition ofEncyclopaedia Britannica. Eventually, Bentley decided topublish a major book on all his snowflake researchand joined forces with William J. Humphreys of theU.S. Weather Bureau, who wrote the text. Their greatwork Snow Crystals was illustrated with more than2,400 of Bentley’s photomicrographs. It came out latein 1931, just before Bentley died. He had walked homefor six miles, marveling at the snowflakes in a bliz-zard, and on Dec. 23 he died of pneumonia.

Bentley became an inspiration to other investiga-tors. A German microscopist named J.M. Pertner in-cluded some of Bentley’s images in his book on snowcrystals published in 1906, and Alfred Lothar Wegenerreproduced them in his influential book Thermodynamikder Atmosphäre (1911) published when he was only 31.Wegener was a German meteorologist, who was fasci-nated by ice and snow. He became an avid polar ex-plorer and coined the concept of continental drift in1912. This became a matter of great controversy, and

Wilson “Snowflake” Bentley painstakingly retouched his glassnegatives so that the snowflakes stood out dramatically from thebackground. Bentley is now claimed to have been the first tophotograph snowflakes and the first to publish a book on thesubject, none of which is true. He was, however, the greatestsnowflake micrographer of the 20th century.

Alfred Wegener, the pioneer of continental drift theory, includedmicrographs of snowflakes in his publications on meteorology,Thermodynamik der Atmosphäre (1911). He used this plate fromJ.M. Pertner that owed much to “Snowflake” Bentley. It originallyappeared in Pertner’s book Schneekristalle, published in 1906.

179

CRITICAL FOCUS | BRIAN J. FORD

the scientific establishment of the U.S. rebelled againstit heartily. The president of the American Philosophi-cal Society dismissed the theory as “utter, damned rot,”which was remarkably frank language at the time. In-deed, American teaching did not begin to accept thetheory until the 1960s. It was the beauty of the arcticsnows that excited Wegener’s passion, which is whyhe included snowflakes in his writings. In 1930, hewent on his fourth exploration of Greenland, and notrace of him has ever been found. He died knowingthat hardly anyone believed he was right; how curi-ous that the science of plate tectonics is now knowneverywhere and has no detractors in sight.

JAPAN’S FAKE FLAKES

There was one further step to take: creating artifi-cial snowflakes in the laboratory. For this crucial re-search we must return to Japan, a country where, aswe have seen, interest in the microscopical structure

of snow dates back to the 1700s. This was the life’swork of Dr. Nakaya Ukichiro. He was born on July 4,1900, close to the area where snowflakes were de-scribed in Hokuetsu Seppu, the book that had been pub-lished in 1837. He was inspired by this historic workas a young scientist at the Imperial University of To-kyo. In 1928 he moved to London to study underDr. O.W. Richardson, and then settled in as professorof physics at Hokkaido University in Sapporo, remain-ing attached to that university for the rest of his life.Nakaya specialized in low-temperature research andset out to try and create artificial snowflakes. Thegreatest problem he faced was the lack of a suitablesubstrate, and he soon found what he wanted: rabbithair. Placed in a sub-zero chamber with high-humid-ity air, Nakaya soon found that crystallization wouldbegin. In 1936, he created the world’s first artificialsnowflakes. During the Second World War he sufferedperiods of ill health, but in 1952, he was invited by theInternational Glaciological Society to tour the U.S. andto attend the conference that established SIPRE (Snow,Ice and Permafrost Research Establishment) of whichhe remained a fellow from 1952 to 1954. He lived inWinnetka, IL, a town just north of Chicago, and wrotehis masterwork Snow Crystals: Natural and Artificial forHarvard University Press in 1954. There was grow-ing popular interest in snowflakes at the time. In 1951,Rosemary Clooney released Suzy Snowflake, a nov-

In 1936, Japanese scientist Nakaya Ukichiro became the first tosuccessfully create snowflakes in the laboratory. Snowflakes initiatearound a tiny nucleus (often a speck of mineral dust), and Nakayafound that a rabbit hair would support crystal growth. He grew hiscrystals in freezing air that was supersaturated with water vapor.

Nakaya published this first useable chart of snowflake types in hisbook, Snow Crystals: Natural and Artificial, published in 1954.Unlike the earlier versions we have seen (published in 1681,1777, 1820 and 1832), he set out to produce diagrams that agreewith what we observed under the microscope. Descartes’ “axle”from 1637 can even been identified (lower left).

180 THE MICROSCOPE 62 (2014)

elty seasonal song produced by the Columbia RecordCompany. The song became the soundtrack for a car-toon enjoyed by Chicago children every holiday sea-son. On Dec. 28, 1953, Suzy Snowflake was the subjectof a poorly produced animated short film by CentaurProductions, featuring in Garfield Goose and Friendson WBBM-TV (CBS).

My interest in the beauty of snowflakes dates backto my student days. One of my early snowflake stud-ies was printed off for use as a Christmas card andwas reported in the Western Mail newspaper on Dec.14, 1963. Color photography was costly back in theday, so I used a ferric ammonium citrate toner to adda blue tint to the prints and then rinsed them in a paleyellow bath (I cannot remember what that was) togive the resulting photograph the appearance of a colorprint. A series was published in an article titled “TheBeauty in a Snowflake” in the New Knowledge encyclo-pedia, Vol. 5, 1965.

CHILLED MICROSCOPY

Today we have our own version of “Snowflake”Bentley in Dr. Kenneth Libbrecht. If I told you hisprofession, it may ring a bell. Libbrecht is the profes-sor of physics at California Institute of Technology(Caltech), and would be the head of the departmentwhere the character Sheldon Cooper in the televisioncomedy “The Big Bang Theory” does his research. Fans

of the show know that the head of Caltech’s physicsdepartment in the TV program is the characterDr. Eric Gablehauser (though Libbrecht tells me thathe has rarely watched the show and knows nothingof its characters). Libbrecht takes his micrographswith a chilled microscope fitted with a camera, which(since cameras tend to resent the cold) is enclosed ina warmer chamber. Many of his published imageshave been taken by Patricia Rasmussen; both say theywere originally inspired by perusing Wilson Bentley’sstunning photomicrographs. Libbrecht has certainlymade most of his subject-matter: his published booksinclude The Art of the Snowflake, Winter’s Secret Beauty,Field Guide to Snowflakes, The Magic of Snowflakes andthe Little Book of Snowflakes. Of particular scientificinterest is his 40-page paper on the physics of snow-flake formation in Report on Progress in Physics (Vol. 68,pp 855–895, 2005). His research on the “AerodynamicStability and the Growth of Triangular SnowCrystals” appeared in The Microscope in 2009 (57:4, pp157–163).

Both the U.S. Postal Service and Sweden’s post of-fice have produced stamps featuring Libbrecht’s mi-crographs. In July 2009, he received the Émile ChamotAward presented by the State Microscopical Societyof Illinois. His is a breathtaking career.

Many people observe snowflakes in the comfortof a warm laboratory by making plastic replicas ofthem. John Delly wrote one such technique in this jour-nal (The Microscope, 38:2, p 223, 1990). James Benko andThom Hopen have also taken great pictures, and I havesome of Benko’s fine replicas under the microscopeby me as I write. The traditional method was to dis-solve polyvinyl acetal resin in ethylene dichlorideto make about a 1% solution of the plastic. You canalso drop chilled liquid superglue onto snowflakes,and even clear acrylic spray paint. But these are thetechniques for wimps. The true devotee will use a mi-croscope outside in the cold. Among the investiga-tors currently working on the micrography of snow-flakes are Alexey Kljatov in Moscow, Don Komarechkain Canada, Hermes Sarapuu in Estonia and MarkCassino at Kalamazoo, MI. From time to time, scan-ning electron microscopes (SEMs) have been used tostudy snowflakes, but the results are synthetic andlook like hunks of plastic. A selection has been takenat the Electron Microscopy Unit of the Beltsville Ag-ricultural Research Center in Beltsville, MD. They’rehorrible.

Finally, can it be true that no two snowflakes areidentical? Yes it is true, and here is my proof. It is thesubtleties of constantly changing humidity, tempera-

There was relatively little interest in snowflakes in the 1960s. Theauthor took these micrographs in the winter of 1964. One becamea holiday greeting card and was reported enthusiastically in thepress. This plate is taken from New Knowledge, which publishedfour of the images of varying complexity.

181

ture and air currents that dictate the crystallizationof water vapor into ice crystals in snowflakes. Thesefluctuations dictate how the six rays form. Clearly,the variations between one part of a minute snow-flake and its other branches are minuscule — and yetthere are subtle differences between their fine struc-ture. If no two branches on the same flake can be iden-tical, no two flakes could ever be the perfect match. Itis the breathtaking beauty, the crystalline intricacyand sublime subtlety of each feathery constructionthat makes snowflakes so captivating.

And when were snowflakes first written about?This dates back to 135 B.C., when the Chinese philoso-pher Han Ying pointed out that although flowers hadfive petals (he must have been looking at dicotyledon-ous species), “blossoms of snow are always sixpointed.” More than 17 centuries later (and over fourcenturies ago) Kepler was attracted by them on theCharles Bridge. On that wintery day when I was inPrague, I thought back to the centuries of research thathad taken place before our own efforts started to sur-face. I tightened my scarf and pulled up my collar as alarge snowflake settled onto my glasses. I didn’t havethe heart to crush it with a tissue, so I blew it awaycarefully and watched it slowly settle, just as a myriadpeople have done before.

CRITICAL FOCUS | BRIAN J. FORD

Our modern-day version of “Snowflake” Bentley is KennethLibbrecht, professor of physics at CalTech University. He andPatricia Rasmussen have produced an array of stunning snow-flake micrographs. Libbrecht has authored a number of books onsnow, and his images have been used for postage stamps.

Don Komarechka, a macro photographer based in Ontario,Canada, uses a hand-held camera outside in freezing conditionsand takes up to 60 high-speed frames of the same flake at varyingfocal settings. He then uses software to interpolate the sharpestimages from which he assembles his astonishing results.

Alexey Kljatov from Moscow has published some remarkablesnowflakes, and he reminds us that ingenuity matters more thancostly equipment by relying on a Helios 44M-5 lens from an oldRussian Zenit camera. Kljatov pays special attention to lighting andsucceeds in capturing a snowflake’s startling beauty.