Too Comfortable for Ice

The room was too hot, the water was too warm, and Alfred Stieglitz had colour portraits to develop. The Lumière instructions called for solutions and washing water at 60°F, about 16°C. His chemicals were at 75°F, about 24°C. The room had reached 84°F, nearly 29°C.

There was no ice to be had, he wrote, before qualifying the excuse: perhaps he had simply been “too comfortable to get it”.

This was May 1908, less than a year after the Lumière Autochrome had gone on sale. Stieglitz, an American photographer and editor who had enthusiastically introduced it to New York audiences, was still learning how to keep its delicate image layer from coming off the glass. In the first of six experiments he described, the plate began to frill at the edges during the bleaching bath. Within three minutes the damage had spread across it. The portrait was useless.

The work inside the plate

Auguste and Louis Lumière, inventors of the Cinématographe, manufactured Autochromes in Lyon, France. Their colour plate reached the market in 1907, after years of development. It could be used in an ordinary plate camera, without making three separate exposures through different coloured filters.

Colour photography itself was older. Other experimenters had produced colour images through separate records or optical interference. What the Lumières offered was a ready-made material that brought a great deal of that work into one sheet of glass. Every part of the scene could be recorded during the same exposure.

The factory put down a sticky coating and spread a mixture of potato-starch grains across it. The grains had been dyed orange-red, green and violet-blue. They formed a random mosaic, packed close together in a single layer. Black carbon filled the remaining gaps, preventing stray white light from passing between the filters.

The grains had to let enough light through.

A rolling press flattened the screen. A protective varnish separated the starch and its dyes from the water used in photographic processing. Above it went the light-sensitive silver-gelatin emulsionPhotographic emulsion×The light-sensitive coating on photographic glass, film or paper. It commonly contains silver salts suspended in gelatin.Merriam-Webster · emulsion ↗. Making a plate meant joining materials that had different jobs and different tolerances of moisture.

The grains supplied the colours before there was a subject. The emulsion would determine how much light could pass through each tiny coloured patch in the finished photograph.

A red object did not need to be painted red afterwards.

Its reflected light passed more readily through the orange-red grains than through the neighbouring green and violet-blue ones. The light-sensitive material directly behind those grains received the stronger exposure. The camera was making a microscopic pattern in silver salts, selected by the colours already present in the screen.

This arrangement dictated how the plate went into the camera. Its bare glass faced the lens, so that the incoming light met the colour mosaic before it reached the emulsion. A special yellow filter over the lens corrected the balance of light reaching the plate.

All that filtering cost light. The process favoured steady subjects, good illumination and people prepared to hold a pose. It could make portraits, but it gave the photographer less freedom to catch a passing movement. Nor was an exposed plate yet a colour picture ready to hold up to a window.

Make the negative, then remove it

First development turned the exposed silver salts dark. Behind the grains that had passed most of the subject's light, the photographer now had most of the black silver. Viewed at this stage, the image had the wrong relationship of light and dark for a positive transparency.

An ordinary negative would next be fixed, removing the unused light-sensitive salts. With an Autochrome, those salts had to stay. They were the material from which the final image would be made.

The first black silver image was bleached away. The salts that had escaped the original camera exposure remained in the emulsion. The worker deliberately exposed them to light and developed the plate again, turning that remaining material dark. This second image reversed the first.

For the red object, the paths behind the orange-red grains that had received strong exposure were now relatively clear. More competing paths were blocked by silver formed during the second development. White light sent through the finished plate emerged as a multitude of coloured points, which blended at ordinary viewing distance.

The colour screen and photographic emulsion remained on the same glass throughout processing. There were no separate colour pictures for the photographer to line up afterwards. There were, however, several baths in which a fragile gelatin layer could be damaged, and an exposure decision that could not be revisited with another print from the same negative. The plate itself was the finished original.

A box passed from hand to hand

Stieglitz's enthusiasm had begun in Paris in June 1907. He and Edward Steichen were there when the new process was demonstrated. Illness kept Stieglitz away from the presentation; Steichen attended and bought some plates.

According to the account Stieglitz published that October, Steichen brought his first two pictures to the hotel. They were comparative failures. Even so, they convinced Stieglitz that practical colour photography had arrived. Within a week, he reported, Steichen was making much better ones.

Stieglitz had organised the Photo-Secession to press photography's claims as an art. He exhibited work at his New York gallery and edited Camera Work, which combined carefully made reproductions with criticism and technical discussion. Steichen and Eugene were fellow photographers, pursuing their own experiments.

Steichen took his experiments to Britain, photographing George Bernard Shaw, Lady Hamilton and a group on George Davison's houseboat. He gave a box of plates to R. Child Bayley, editor of the British journal Photography, so that Bayley could try them himself. Stieglitz and Eugene continued testing at TutzingTutzing, Germany×A town on Lake Starnberg, south of Munich in Bavaria. Stieglitz and Eugene tested Autochrome plates here.Open in Google Maps ↗, near Munich in Bavaria.

The opportunity to experiment depended on obtaining more glass. Stieglitz reported shortages in France during July and part of August, caused by trouble at the Lyon factory. Photographers who had read enthusiastic reports could not necessarily buy the material needed to test them.

Stieglitz was already asking how long the pictures would last.

At Tutzing, Stieglitz investigated permanence and keeping qualities as well as picture-making. He admitted to many failures and remained undecided about varnishing the completed image. Steichen, meanwhile, was changing his chemical treatment to obtain the tone and quality he wanted.

Stieglitz sailed from Europe on 18 September with pictures by Steichen, Eugene and himself. After landing in New York, he invited the press to see them at the Photo-Secession galleries, 291 Fifth Avenue, on 27 and 28 September. As he wrote up the story, commercial plates were still expected on the American market.

The pictures could travel ahead of the supply. But a journalist looking at one still had to be shown it in the right light. It was a transparency, intended for transmitted illumination or projection, rather than an opaque print. One exposure had made one original; distributing equivalent colour copies was another problem.

Keeping the image on the glass

By the following spring, Stieglitz had ample reason to distinguish the plate's built-in colour from the work of using it. In his hot room that May, the manufacturers' recommended bath temperature was well below what came from the tap. He described himself as a photographic vagabond, working wherever he could, without a permanent workplace in which to control conditions.

He tried formalin as a hardener, although Lumière and photographic acquaintances had advised against it. One plate was partly rescued after its edges began to lift. Another treatment produced a good result, then an unsatisfactory repeat. Later experiments, including protecting the edges or hardening before development, gave him plates he considered successful.

His conclusions from the six tests were provisional. He still needed to determine whether other emulsions would tolerate formalin and whether hardening should come before or after the first development. The printing forms of Camera Work had already been locked when he had them reopened to insert his account, in time for other photographers facing the summer.

A closer look · 3:04 · English

Looking through the glass

Play film

A successfully developed plate could still lose its colour in the light required to see it. At the Metropolitan Museum of Art, conservator Luisa Casella investigated whether removing oxygen would slow that damage. Tests showed substantially better stability, though fading did not stop altogether.

In January 2011 the museum announced a brief display of five original Autochromes by Stieglitz and Steichen, in individual low-oxygen enclosures under controlled light. They would be shown from 25 to 30 January. For the rest of the exhibition, facsimiles would take their places.