Finished with a Penknife

Wilson Bentley could see the snow crystal on the screen at the back of his camera. The screw that brought it into focus was at the other end. With the bellows extended, he could no longer reach it.

His solution was a cord. It passed around the microscope's focusing screw and over a wheel on either side of the camera, bringing the adjustment within reach of the man watching the image. In 1922, when he explained the arrangement to readers of Popular Mechanics, Bentley included a photograph of the wheels. The crystal was changing while he worked. Walking round the table to make another adjustment would consume some of the time he had left.

A cold room facing the window

Bentley lived in JerichoJericho, Vermont, United States×A town in north-western Vermont, east of Burlington. Bentley lived and photographed snow crystals on his family farm here.Open in Google Maps ↗, Vermont, on the family farm. Born in 1865, he grew interested in the structures he could examine through a microscope his mother gave him. Drawing offered a way to keep them, but snow crystals changed too quickly for a careful record of their internal detail. During the 1880s he began making photographs through a microscope instead. He would continue until the year of his death, 1931.

The working arrangement depended on a room kept approximately as cold as the air outside. In his instructions, the room, apparatus and accessories all had to be clear of artificial heat. The camera stood near a window and faced it. Ordinary daylight passed through the crystal and the microscope into the camera.

The microscope supplied the camera's optics.

It was turned horizontally and joined to the camera by a connection that excluded stray light. The microscope eyepiece was removed; no separate lens was needed in the camera. Its objective projected an enlarged image towards the photographic plate. Different objectives and different extensions of the bellows gave Bentley a range of magnifications, with the cord allowing him to focus from the rear.

There were several ways to lose a specimen before the plate was exposed. A collecting board could take warmth from the hands holding it. A branch could break during transfer. A crystal might be placed at an awkward angle, or slide while the exposure was under way. Bentley's equipment included thick mittens, a wooden splint and a feather, each answering a different part of the problem.

He caught falling crystals on a black surface where their outlines were visible. His 1922 collecting board was about a foot square, fitted with handles so his hands need not touch and warm it. He repeatedly brushed it clean with a turkey wing or similar duster, keeping an eye on what arrived. When promising specimens fell, he took the board immediately into the cold room.

A pointed wooden splint touched the face of the chosen crystal gently enough for it to adhere. Bentley could then lift it onto a glass slide. Too much pressure crushed it. After a brief inspection under the observation microscope, he used the edge of a feather to press a suitable crystal flat. Throughout, he had to keep his breath off it.

For especially frail specimens, he recommended leaving five or six other crystals around the one being photographed. These helped slow its loss to the air. The board, tools and slide were all part of keeping a specimen available long enough to record it.

The slide went onto the photographic microscope, where Bentley centred the crystal and focused its image. A black card covered the objective while he inserted the plate holder and withdrew its protective slide. Uncovering the objective began the exposure; replacing the card ended it. In his 1922 instructions, exposures could run from eight seconds to a hundred or more, depending on the conditions and magnification. A brighter day did not give the same exposure as a dull one, nor did a small image demand the same time as a large one.

His 1898 article had given a longer range, forty to three hundred seconds. These were working instructions from different points in his career, rather than a single fixed recipe. Either way, the crystal had to remain still throughout. Bentley later described specimens slipping on the slide and leaving blurred negatives, or breaking as he picked them up. A fine snowfall could arrive when the temperature was just too high; usable daylight could end while good crystals were still falling.

Clearing the glass

The developed negative preserved what the camera had recorded. It did not automatically produce the familiar Bentley print: a pale, intricate crystal standing alone against deep black.

The difficulty came from the material. A snow crystal is transparent ice. Under transmitted light, much of its body could be too similar to the surroundings for the clear separation Bentley wanted. Fine internal lines might register well while the whole form remained difficult to read. He called his solution “blocking out”.

The black background took handwork.

Bentley supported the negative over light and traced closely round the crystal's image with a sharp blade. He then scraped away the surrounding photographic emulsionPhotographic emulsion×The light-sensitive coating on photographic glass, film or paper. It commonly contains silver salts suspended in gelatin.Merriam-Webster · emulsion ↗, leaving clear glass outside the outline. When printed, that cleared area became dark. Following the indentations between delicate branches could take hours.

He recommended doing the cutting on a copy negative, protecting the original from an irreparable slip of the knife. In a separate account for American Photography in 1922, he said that the original was usually kept intact as evidence of the image's fidelity. He described working on the copy under a microscope, moving from a penknife to larger blades as he cleared more of the surrounding area. He estimated that, added together, this labour had occupied a year or more of his life.

The finished print therefore combined a photographic record with an outline followed by hand. Bentley explained the intervention openly: he wanted the crystal to stand out as people expected white snow to stand out against a dark surface. The appearance of an effortless, isolated specimen came after both the cold work at the camera and the slow work on its reproduction.

Keeping a record of the storm

Other people helped turn the photographs into a larger undertaking. George H. Perkins, a naturalist at the University of Vermont, joined Bentley in publishing A Study of Snow Crystals in 1898. Perkins made his contribution explicit: he had compiled the paper from Bentley's notes and photographs; the observations, theories and illustrations belonged to Bentley.

For his study of the winter of 1901–02, Bentley thanked E. H. Nash for changing and numbering exposed plates, freeing him to spend more time finding and photographing crystals. Weather maps supplied through the US Weather Bureau and John W. Smith, its New England forecasting official, helped him compare what fell at Jericho with the movement of storms.

He wanted to connect the forms with the weather that produced them. His tables recorded dates, crystal types, temperature, pressure and wind. He also recognised a serious weakness in his photographic record. Years of looking for beautiful and interesting specimens had made it difficult to collect a representative sample of a snowfall. Broken and unsymmetrical crystals were liable to be passed over.

The distinction mattered. A collection of outstanding examples could reveal an immense variety of structure while saying little about how common each form was. Nor did a thermometer at ground level measure the temperature in the cloud where a crystal had grown, a limitation Bentley acknowledged. His proposed accounts of conditions high in the storm went beyond what he could observe directly from the farm.

The familiar claim that no two snowflakes are alike needs a similar distinction. Bentley and Perkins's 1898 paper allowed that two crystals could be nearly, perhaps exactly, alike in outline. It was the detailed internal figures that seemed almost impossible to match. They were comparing selected specimens, not every crystal that had ever fallen.

Modern experiments make the role of growing conditions easier to examine. The Caltech physicist Kenneth Libbrecht has grown pairs of complex crystals side by side, giving them closely matched changes in temperature and humidity. He calls them identical twins while specifying that they are not exactly identical. Bentley's photographs preserve differences visible in the crystals he obtained; a collection of them cannot establish a universal rule about all snow.

Pictures that could travel

The photographs had uses beyond Bentley's weather studies. In 1928 he listed schools, museums, lecturers, silk manufacturers, metalworkers and other designers among their users. A crystal that had lasted minutes on a slide could supply a pattern long after the storm.

He also prepared boxes of lantern slides: transparent images that could be projected at a size a roomful of people could see. Colleges and universities bought sets. The University of Wisconsin's surviving collection includes snow crystals alongside frost, dew and clouds, preserving the broader range of water forms that interested him.

William J. Humphreys, a physicist with the US Weather Bureau, helped give the photographs another durable form. For Snow Crystals, published by McGraw-Hill in 1931, he sorted through some 4,500 photomicrographs. The Wisconsin collection's account records 2,453 pictures selected for the book, most of ice crystals, with about a hundred of frost and dew.

Bentley died on 23 December that year. The book was reissued in full by Dover in 1962. The lantern slides followed a different route into the present: Wisconsin's librarians digitised their collection and cross-referenced its images against the book and the Jericho archive. Individual crystals could be found again across sets that had travelled to different places.