Daguerreotype

☠Level D: studied, never performed

The daguerreotype depends on mercury vapour, on iodine and later bromine and chlorine vapours, and in some later practice on potassium cyanide. It is classified Level D under the safety rubric and the Level D policy: described for its chemistry and its history, and never carried out.

This entry contains no procedure of any kind. No quantity, no temperature, no time, no apparatus and no sequence detailed enough to follow. Where a step has to be named for the chemistry to make sense, it is named and nothing more.

The Part I lesson owns the history; Part XXVI owns the mercury chemistry and the toxicological argument, in a lesson written for that purpose.

The chemistry

The plate is not coated; its own surface is converted. A sheet of copper faced with pure silver is polished to a mirror and sensitised in a closed box over iodine, taking on a yellow-rose colour. Iodine vapour meets metallic silver and turns the outermost layer of it into silver iodide — by a long way the least soluble of the three common silver halides, at 2.8 × 10⁻⁶ g per litre at 25 °C against 0.135 mg for the bromide and 1.93 mg for the chloride.

Two consequences follow and both matter. There is no binder, no gelatin and no paper fibre between the viewer and the image, which is why a daguerreotype resolves detail that nothing else of its century approaches. And a salt that insoluble is correspondingly hard to wash out again, which is the fixing problem below.

The development is an alloy. Mercury vapour condenses on the plate and combines with the silver where light had acted, forming a silver–mercury amalgam. That is the whole operation, and it is strange beside everything else in this atlas: a metal vapour alloys with a metal surface, selectively, and the alloy is the picture.

⚗Four period explanations, and a course that does not adjudicate

Eder records the arguments about why the amalgam forms where it does, and they do not agree. Arago in 1839 gave a chemical account — silver iodide reduced by light to metallic silver, which then absorbs the mercury. Donné, the same year, a physical one: the exposed layer goes powdery so mercury reaches the metal beneath. Moser in 1842 argued from his breath pictures that condensation alone explained it, and showed that a fully exposed plate could be developed with steam. Choiselat and Ratel in 1843 proposed a silver subiodide.

Four accounts, from people with the plates in front of them, and no agreement. This course does not adjudicate between them and does not substitute a modern mechanism it has not read. What is not in doubt is that mercury goes where the light went.

The halogens are part of the same operation. Iodine made the plate sensitive; bromine from 1840 and chlorine from 1841 made it fast enough for a portrait. Eder records the reverse effect too, from Gaudin in 1841 and from Shaw and Percy two years later: exposing a plate that already carries a latent image to any of those vapours destroys it, though it will take a new exposure and develop from that.

Historical workflow

Polish, sensitise over iodine, expose, develop over mercury, fix, gild, case. Each of those is a named operation in the historical record and none of them is given here as a step.

The discovery that matters is the second one. Everything in Part I before this asked light to finish the job on its own, and Daguerre spent months exposing iodised plates in the camera and getting no picture at all. The plates were not blank: they carried a change too small to see, and the discovery was that such a change exists and can be found by a later chemical operation. Everything modern photography does rests on that sentence.

Fixing. Daguerre’s manual states the purpose of the last operation in one clause: to remove the iodine which, if the picture were exposed too long to light, would go on decomposing and destroy it. His first answer was a hot saturated solution of common salt, and Eder is blunt about the result — an imperfect fixation which gave the plates a mottled appearance. The second answer is in his own published manual, which offers a weak solution of pure sodium hyposulfite as an alternative and says plainly that it is preferable, because it removes the iodine entirely, which the salt does not always manage.

Gilding, 1840. The AIC’s conservation page states the problem in the plainest terms available: an ungilded plate is so delicate that the image can be wiped off with a finger. Fizeau found that a bath of hyposulphite of soda containing gold chloride greatly improved both beauty and permanence, and the treatment was adopted everywhere. Eder dates it to 1840; the AIC gives 1841 for the introduction of gold toning, and this entry follows Eder without claiming the year is settled. The chemistry was worked out shortly after by Fordos and Gélis, who identified the double salt and called it hyposulphite of gold and sodium — later sodium aurothiosulfate, sold in the trade as sel d’or.

The image

Six properties, all of them consequences of the picture sitting on the surface of a mirror.

It is a direct positive, and it is unique. No negative exists at any stage. The plate that was in the camera is the finished object, and copies were made by photographing it again or by having an engraver copy it. It is a unique image.

It is laterally reversed unless the camera carried a mirror or a prism. Ware puts the general rule: a direct photographic image is a mirror of reality, and to restore the true chirality it has to be re-photographed.

It is neither positive nor negative but ambivalent. Ware’s account is exact: virtually none of the light falling on a daguerreotype is absorbed by it. The highlights are microscopic globules of silver amalgam, which scatter light diffusely; the shadows are polished silver, which reflects it specularly. Turn the plate so the mirror throws a dark surrounding into your eye and the polished areas read as deep shadow and the picture is a positive; turn it so the mirror throws a bright source at you and the picture inverts. Nothing changed but the geometry.

It resolves extraordinary detail. The AIC gives the image particles as 0.1 to 50 micrometres. The course has not verified any of the resolution figures quoted for daguerreotypes and does not repeat one.

It tarnishes and it abrades. The exposed surface of the plate is the image, which is why the AIC’s guidance is that it must not be touched and why the historical convention is a cover glass held off the surface by a spacer, sealed at the edges, in a case. Eder records Daguerre protecting his pictures under glass as early as 1839, after early plates kept in paper wrappings were damaged. Precisely which compounds the tarnish consists of is a conservation-science question this course does not settle.

It is monochrome, and often not. Many surviving plates were hand-coloured with dry pigment, and the AIC lists applied colour among the identification features.

Permanence

The permanence story is a two-stage one. Unfixed, the plate destroys itself, and Daguerre said so. Fixed but ungilded, it survives chemically and not mechanically: the image can be wiped away. Gilded and cased, it is the most physically fragile durable object in this atlas — stable in substance, vulnerable in surface, and dependent for its survival on a sealed package rather than on its own chemistry.

That is a different permanence argument from every paper process here, and it is worth holding beside them. A salted paper print fails because of what was left in it. A daguerreotype fails because of what touches it.

Hazards

Three of the four reagents that make a daguerreotype have workplace exposure limits, which is a convenient way of saying that people who work with them professionally are monitored. HSE’s EH40 lists mercury and its divalent inorganic compounds, as mercury, at a long-term limit of 0.02 mg/m³; bromine at 0.1 ppm long-term and 0.2 ppm short-term; and iodine at a short-term limit of 0.1 ppm. Potassium cyanide, which entered later practice for fixing and cleaning, carries an aggregated GHS classification as fatally toxic by every route.

EH40 carries a warning worth quoting in any course like this one: the absence of a substance from the list does not indicate that it is safe.

Those figures are not a hazard assessment. They are the reason one is needed, and the reason it is not attempted here: an exposure limit implies monitoring, and monitoring implies equipment and competence a domestic darkroom does not have. One property defeats intuition and is worth stating alone — mercury is a liquid that gives off vapour at ordinary room temperature, and that vapour has neither colour nor smell, so nothing about the room tells you what is in it.

Why we study it and do not reproduce it

Four reasons, and each of them fails if the process is left out.

The chemistry is the same chemistry. Mercury amalgamating with the silver of an exposed plate is physical development; the cyanide fixer is complex formation with a different ligand from thiosulfate. A reader who understands the safe processes but has never met the dangerous ones has a partial mechanism.

The history is not optional. This is one of the two processes photography began with, and the one that won the argument of 1839 and lost the century — which makes it the course’s clearest case of a technical decision settled on grounds other than picture quality. Nothing on paper came near it for detail, and it could not be copied.

Identification and conservation depend on it. A plate in a collection has to be identified before it can be cared for, and its hazards belong to whoever handles it now.

Assessing a hazard is a skill. Part XXVI’s assignment asks a reader to assess a historical process for hazard themselves, which is only possible if they have seen the assessment made on cases where the answer is unambiguous. This is one of them.

The Becquerel variant removes the mercury and does not remove the classification; its own entry says why.