Daguerre's 1839 process outline

The book its first English translator called “the first Manual of a new science” cost one franc and gives its whole chemistry in five numbered operations. Two of those operations are a vapour meeting a metal, one is an exposure, and only two are anything a formulary can hold: a phial of diluted nitric acid at the beginning and a dish of saturated salt water at the end. That imbalance is the most important thing about the daguerreotype and the reason it sits in this book at all. Every other formula here is a bottle. This one is a box of iodine, a box of mercury, and two solutions that between them do no more than prepare the plate and take the sensitive layer off it afterwards.

Acide nitrique étendu d'eau — nitric acid extended with water — the plate-cleaning acid of the first operation

Ingredient	Quantity	Form the source specifies
Nitric acid	100 mL	One part by volume. Daguerre states no strength for the acid itself, no vessel and no batch size; the hundred millilitres are the course's scaling of his ratio.
Water	1600 mL, added	Sixteen parts of distilled water to one of acid, by volume. This is a proportion of water added and not a make-up volume, so no final volume, no per cent and no molarity is stated for it anywhere on this page.
Applied on a cotton pad, spread evenly, and followed by dry pumice. The manual has it applied three times over, twice if the plate is being prepared in advance, and at least once more at the moment of making a picture.

Eau saturée de sel marin — water saturated with common salt — the fixing bath of the fifth operation

Ingredient	Quantity	Form the source specifies
Sodium chloride	250 mL	Dry salt, measured by the space it fills: Daguerre fills a wide-mouthed bottle to a quarter of its height with common salt and fills the rest with clear water. The quarter is his and the litre is the course's, because he names a vessel and not a volume. The undissolved excess is deliberate.
Water	to make 1000 mL	"que l'on remplit avec de l'eau claire" — the bottle is filled up, which is a make-up volume stated as a vessel rather than as a number. Shaken from time to time until the water can dissolve no more salt, then filtered through grey paper until perfectly limpid, and prepared in quantity in advance so that it need not be made for each plate. Used warmed but never boiling; a saturated solution of this salt is barely stronger hot than cold.
A saturated bath, so the working strength is a physical constant rather than a formulation choice: the encyclopaedia gives 36.0 g of salt per 100 g of water at 25 °C from PubChem's HSDB entry. Daguerre's own alternative, a weak solution of pure sodium hyposulfite, is published with no strength at all and so cannot appear here.

Used in this order — the fifth operation, which is the only part of the process that is a bath at all

1. Water — Plain ordinary water, in the second of the two tinned copper dishes. The plate is plunged and withdrawn at once, so that the surface is merely wetted. Daguerre gives the reason as a fault rather than a benefit: without it the salt water or the hyposulfite leaves indelible stains.
2. Eau saturée de sel marin — water saturated with common salt — until the yellow colour has entirely disappeared — An end point and not a duration; the manual publishes no time for this bath. About three centimetres deep in the dish, warmed without boiling, with the plate lifted and lowered on a tinned copper hook to keep fresh solution against the surface.
3. Water — Back into the first dish of pure water at once, and then onto an inclined stand, where not less than a litre of hot but unboiled distilled water is poured over the plate in a sheet. Any drops left standing are blown off before they can dry, because a drop that dries leaves what was dissolved in it behind.

On trempe d'abord la plaque dans l'eau pure contenue dans la bassine … puis, sans la laisser sécher, on la plonge de suite dans l'eau salée … Quand la couleur jaune a tout-à-fait disparu, on enlève la plaque … et on la plonge immédiatement dans la première bassine d'eau pure.

**Recorded because it is what the source publishes, and not as something to follow.** It is set down here for two reasons and neither of them is instruction. It is one of the two earliest published fixing-and-washing sequences in photography — Talbot's own, published in London on 21 February 1839, is the earlier — and every fixer page in this formulary descends from the problem these two were attacking. And it is the only part of the daguerreotype that contains no hazard at all: warm salt water, a copper dish and a litre of distilled water. Nothing here makes a daguerreotype, because what makes a daguerreotype is the iodine box and the mercury box, and this page describes neither as a procedure.

☠Level D: this page describes a process and gives no procedure

The daguerreotype depends on iodine vapour, on mercury vapour deliberately warmed, and in later practice on bromine, chlorine and potassium cyanide. Mercury and iodine are both Level D in this course’s own chemical encyclopaedia, and the process is classified Level D under the safety rubric and the Level D policy.

No procedure for making a daguerreotype appears on this page. The two solutions above are published because a formula that cannot be read cannot be identified, and identification is the use a reader has for this entry. The fixing sequence in the object is the fifth operation, which is warm salt water in a copper dish; the iodine box and the mercury box are described as chemistry and history and are given no quantities, no apparatus, no charge and no sequence. There is no protective equipment that makes a warmed mercury bath acceptable in a domestic darkroom, and none is offered.

The route to this chemistry that the course can publish is the Becquerel daguerreotype, which removes the mercury and does not remove the classification; its own entry says why. The route to a silver image a reader can actually make is salted paper, whose sensitiser and thiosulfate fixer are two pages away.

The process atlas entry owns the image, the optics of its ambivalence and its conservation; the Part I lesson owns the history. This page owns the manual: what Daguerre actually published as quantities, what he published as proportions and vessels instead of quantities, and what the chemists around him could and could not explain in the same volume.

Purpose

To convert the surface of a silver mirror into a light-sensitive salt, find the invisible change that light makes in it, and then take the unchanged salt away without disturbing the picture.

Each clause of that is one of the manual’s operations, and each of the three is strange. The conversion turns the support into the sensitive layer instead of coating one onto it, which no other entry in this formulary does. The finding is physical development, and behind it lies the discovery the whole of modern photography rests on: that light can leave a change too small to see, and that a later chemical operation can find it. The taking-away is a fixing operation set out in a book of working directions — Talbot had published his own, in London, on 21 February of the same year — and Daguerre states its purpose in one clause that could serve as the definition of fixing for the whole of the century that followed: to remove the iodine which, if the picture were exposed too long to light, would go on decomposing and destroy it.

The two solutions in the object serve the first and last of those. The acid’s purpose is to leave nothing on the silver but silver, so that the iodine meets metal and not grease, pumice dust or copper. The salt bath’s purpose is to dissolve away the silver iodide the light did not reach, so that the polished metal beneath it can do the work of the shadows.

There is a second purpose that is this course’s rather than Daguerre’s. The daguerreotype is where the course’s whole account of fixing begins, and it begins with the hardest case in the subject. Silver iodide is by a long way the least soluble of the three photographic halides, and the first fixing problem anyone had to solve was the worst one there is. Understanding why Daguerre’s salt bath was imperfect and why hyposulfite was not is most of what a reader needs in order to understand every fixer page in this formulary.

Recommended uses

None. This is a historical-study entry, and the honest answer to “what is this for” is: reading, identifying, and understanding where the rest of the formulary came from.

What the manual recommends the process for is worth having, because it is a list of what a three-to-thirty-minute exposure can photograph. Daguerre tells the operator to choose objects lit by the sun, because the operation is then quicker, and notes that the action is faster the more strongly lit the subject and the whiter it is by nature. Everything follows from that: architecture, monuments, statuary, landscape, still life. Arago’s report printed in the same volume makes the case in the terms of 1839 — that with such an instrument one man could copy in a few years the millions of hieroglyphs at Thebes, Memphis and Karnak that would otherwise need decades and legions of draughtsmen.

The one use the manual’s own directions never propose is the one the process became famous for, and the three official reports printed with it disagree about whether it would ever be possible. Arago states the dilemma exactly, and it is a photographic argument rather than a chemical one: general opinion is little disposed to admit that the same instrument will ever serve to make portraits, because the problem contains two apparently irreconcilable conditions — for the image to be formed quickly, that is, within the four or five minutes of stillness one may demand and expect of a living person, the face has to be in full sun. The minister’s statement to the Chamber of Deputies says the smallest advance will be enough for Daguerre to portray living people; Gay-Lussac’s report to the Chamber of Peers says the problem of the process’s application to the portrait is nearly solved. Within two years the accelerating vapours and a faster lens had settled it, and the daguerreotype became a portrait medium almost to the exclusion of everything on this list.

The uses this entry serves in a modern course are three.

• Identification. A cased plate in a family collection or a dealer’s tray is identified by the properties this chemistry produces: a mirror surface, an image that flips between positive and negative as it is turned, and a lateral reversal. Those follow from the chemistry described below and from nothing else in the atlas.
• Reading a nineteenth-century text. A reader who meets “the fifth operation”, “the iodine box” or “eau saturée de sel marin” in a reprint has somewhere to look them up.
• Understanding fixing. The mechanism section below is the course’s hardest fixing case, and it is worth working through before reading any of the fixer entries.

When another formula is preferable

Always, for anything a reader intends to do.

For a direct positive on a plate, without mercury: the Becquerel daguerreotype develops the same iodised silver plate by red light instead of by mercury vapour. It is slower by roughly an order of magnitude, and this course still classifies it Level D, because sensitising a plate over elemental iodine remains sensitising a plate over elemental iodine.

For a silver image made from published chemistry a reader can handle: salted paper over Reilly’s plain salting solution, fixed in a plain thiosulfate bath. That is the same silver chemistry, at Level B, with every quantity published.

For the fixing chemistry specifically: any fixer in this formulary is a better teacher than a saturated salt bath, and the plain hypo fixing bath is the direct descendant of the alternative Daguerre himself preferred.

For the history without the hazard: Talbot’s photogenic drawing paper is the other 1839 process, published within weeks of this one, and it is safe to make.

Mixing

Daguerre publishes a mixing instruction for one of the two solutions and a vessel for the other, and the difference between them is instructive.

The acid is a proportion. One part of acid to sixteen parts of distilled water, by volume, with “(en volume)” stated twice in the French so that nobody reads it as weights; Memes adds the sentence outright in English. There is no strength given for the acid, no vessel, and no batch size. Nothing is said about the order of addition, which by modern practice would matter a great deal — the nitric acid page records the instruction printed in capitals on the ILO-WHO safety card, never pour water into this substance — and the manual’s silence on it is one of several places where an 1839 text assumes a competence a modern reader should not assume they have.

!A ratio is not a strength, and an 1840 reader proved it

The volume bound after the manual in the copy scanned here is a separate anonymous pamphlet of 1840, Le daguerréotype considéré sous un point de vue artistique, mécanique et pittoresque, written by somebody working the process. Its author records exactly the trouble a ratio without a strength causes. Having been told to use nitric acid extended with water, he bought commercial acid from one of the first chemical manufacturers in Paris; what he was supplied “portait soixante degrés”, and his mixture made up with sixteen parts by volume of distilled water came out stronger than the one he had been using before.

The course does not interpret the sixty degrees, because the pamphlet does not name the scale it belongs to. What the passage establishes without any interpretation at all is the general point: a proportion of concentrate to water specifies a solution only when the concentrate is specified, and Daguerre’s is not. It is the same defect the schema note above records, seen by a practitioner within a year of publication.

The salt solution is a vessel and an end state. Common salt into a wide-mouthed bottle to a quarter of its height; the bottle filled with clear water; shaken from time to time to help the salt dissolve; and when the water is perfectly saturated — Daguerre defines the term in the next clause, when it can dissolve no more salt — filtered through grey paper until it is perfectly limpid. He adds that it is made in quantity in advance and kept in corked bottles, so that it need not be made for each picture.

That is a formulation instruction of a kind the rest of this formulary never uses, and it is worth seeing clearly. The quarter-bottle of salt does not set the strength of the bath. Saturation does, and saturation is a property of the salt and the temperature rather than of the recipe. The quarter is there to guarantee that undissolved salt remains, which is the only way to be certain the solution is saturated when you have no way to weigh anything. The filtration is there because undissolved salt in the dish would scratch a mirror.

The hyposulfite solution has no mixing instruction at all, and the book compensates for it twice. In the fifth operation it is “une solution faible d’hyposulfite de soude pure”, weak, pure, and otherwise unspecified. In the publisher’s notice at the back it is a piece of hardware: when the bottle is exhausted, refill it with distilled water and as much hyposulfite in crystals as the little silver-plated measure supplied with the apparatus will hold, filtering it as the sea salt is filtered. A reader who bought the Giroux outfit had the formula in their hand as an object. A reader who did not had a word.

Behaviour

The acid behaves as a cleaning agent whose effect is judged by eye. Daguerre’s own test is optical and precise for what it is: the acid is evenly spread when the surface of the plate carries a regular film over its whole extent, and where it has broken into globules that have not taken, those places will make stains. The 1840 pamphlet gives the same test in the negative and adds the reason: you know the plate is properly stripped and degreased when the acidulated water spreads on it without forming droplets. That is a contact-angle test in all but name — the reading is this course’s rather than the pamphleteer’s — and it is the same test a modern worker uses to decide whether glass is clean enough to coat.

The manual is also clear that acid is not the only thing happening in the first operation, and that the sequence of pumice, oil, acid, heat and more pumice is not decorative: after heating, a slight whitish film forms on the silver, which has to be polished off; then the acid goes on again, three times in all, each time followed by dry pumice on fresh cotton. What that whitish film consists of is not stated by Daguerre, by Arago or by any source in this course’s corpus, and the page does not guess.

The salt bath behaves visibly and stops when the colour goes. The yellow of the iodised layer is the indicator, and its disappearance is the end point. There is no time, and no time would be much use: the bath is warm, its concentration falls as it works, and Daguerre has the operator lift and lower the plate on a hook to keep fresh solution against it. The manual’s honest sentence about the bath is the one that matters most: it does not always remove the iodine entirely, and it is worst at it when the plates have been made some time before.

And no time is published for it, which is itself informative. Every other bath in this formulary carries a duration somewhere. This one cannot, and the chemistry below says why: salt water is a poor solvent for silver iodide, and a bath that only just manages the job has an end point that moves with how much iodide is on the plate, how long it has stood, and how warm the dish is. A published time would have been a published lie.

⚗Why both liquids are warmed and why the salt bath gains almost nothing by it

Daguerre has both dishes heated without boiling. For the plain water that is a rinsing convenience. For the salt solution it looks like an attempt to make a weak solvent stronger, and the encyclopaedia entry for sodium chloride shows how little it can have bought: one gram of the salt dissolves in 2.8 mL of cold water and in 2.6 mL of boiling water. A saturated salt solution is barely stronger hot than cold.

What warmth does buy is rate rather than capacity. Diffusion is faster warm, the reaction at the surface is faster warm, and a bath that is only just capable of dissolving the salt at all will get through it sooner. That is the same reasoning behind every warm fixing bath since, and it is worth noticing that a fixing bath published in 1839 already had it, without any of the theory to justify it.

Image characteristics

The image is a pattern of amalgam on a mirror, and every property in the object above follows from that fact rather than from anything in the two solutions.

It has no density scale. Ware’s account is the one this course follows: 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. Tilt the plate so the mirror throws a dark surround into the eye and the polished areas read as deep shadow; tilt it the other way and the picture inverts. Nothing changes but the geometry, which is why this is the one entry in the formulary whose contrast field has to say that the concept does not apply.

Its “grain” was measured in 1839 and again in the twenty-first century, and the two figures are not the same measurement. Dumas and Adolphe Brongniart put the spherules of the lights and half-tones very regularly at one eight-hundredth of a millimetre — 1.25 µm — with the microscopes of the day. The AIC’s conservation page gives image particles ranging from 0.1 to 50 µm. The AIC figure is a range across the plates conservators actually meet, gilded ones among them; the 1839 figure is what two chemists saw down a microscope in the first year, on plates made one way. The course reports both and does not reconcile them.

It is a direct positive and a unique image. No negative exists at any stage, so there is no printing, no edition and no enlargement, and it is laterally reversed unless the camera carried a mirror or a prism. Ware puts the general rule for every direct process: the image is a mirror of reality, and restoring the true chirality means re-photographing it.

It resolves what it resolves because there is nothing in the way. No binder, no paper fibre and no coating sit between the eye and the image. The course has not verified any of the resolution figures commonly quoted for daguerreotypes and repeats none of them.

Ungilded, it can be wiped off with a finger. That is the AIC’s sentence, not a paraphrase, and it is the physical fact that the Fizeau gilding bath of 1840 was invented to fix.

And its rendering of colour was diagnosed correctly in the same volume that announced it. Gay-Lussac’s report to the Chamber of Peers, printed with the manual, states that coloured objects are not reproduced with their own colours, and that because the different luminous rays do not act in the same way on Daguerre’s reagent, the harmony of lights and shadows in a coloured subject is necessarily altered — “a stopping-point traced by nature herself for the new process”. That is spectral sensitivity described as a defect, and described correctly, long before dye sensitisation gave anybody a way to alter it. It is the same fact that makes a red curtain photograph black on an unsensitised silver halide.

The mechanism

Four events, in the order the plate meets them. The first and the last are ordinary chemistry that this course can state exactly. The second is the latent image, which nobody in 1839 could explain and which is taught elsewhere. The third was argued over for years by people with plates in front of them, and is not adjudicated here.

Silver into silver iodide, at the surface only

Iodine vapour meets metallic silver and converts the outermost layer of it into silver iodide:

2 Ag + I₂ → 2 AgI

The second operation: a metal surface becomes a light-sensitive salt without anything being coated onto it

Two things about that layer are worth holding. It is astonishingly thin. Arago’s note records Dumas’s weighings, from which the thickness of the gold-yellow layer does not appear to reach a millionth of a millimetre; a note in the same passage adds that the plate does not gain appreciable weight in taking the iodine on, but gains very noticeably under mercury vapour. And its colour is the assay. Daguerre’s whole specification for the operation is a colour — a fine golden yellow, not carried on to violet, because a violet layer is less sensitive to light, and not stopped short of it, because a layer that is not yellow enough reproduces the image only with great difficulty. Five to thirty minutes, rarely more, and never timed: looked at.

Light on silver iodide, producing nothing visible

The exposure produces no visible change whatever. That is the sentence the whole of modern photography rests on, and Daguerre states it plainly: the impression of the image exists on the plate but is not visible. What light does to a silver halide crystal — photon, photoelectron, silver ion, speck of metallic silver — is taught in Part IV and not here, because the daguerreotype’s own literature could not say it.

Mercury on the exposed places, forming an alloy

Mercury vapour condenses on the plate and combines with the silver where light had acted, forming a silver-mercury amalgam. The alloy is the picture. This is physical development in its purest form: nothing is reduced by a developing agent, a metal simply joins another metal in the places light prepared.

⚗Four period explanations, and a manual that admits it does not know

Arago’s note in Daguerre’s own book sets out the obvious theory and then destroys it. The obvious theory is that light drives the iodine off wherever it strikes, the metal is laid bare there, mercury acts freely on the bared parts, and the hyposulfite wash then removes the iodine the light did not release — artistically, uncovering the mirror-bright parts that are to make the blacks.

Two objections follow immediately, and both are Arago’s. First, what then are the numberless and marvellously graded half-tones the pictures show? Second, and decisively, Pelouze weighed the plate. It gains no appreciable weight from the iodine layer, gains noticeably under mercury vapour, and yet after the hyposulfite wash — with amalgam demonstrably on its surface — it weighs less than before the operation began. Chemical examination of the liquid confirmed why: the hyposulfite is removing silver.

Arago’s own closing sentence is the model for how this course treats a mechanism nobody has settled: thousands of fine drawings will perhaps be made with the daguerreotype before its mode of action has been completely analysed. The process atlas entry sets out the four competing nineteenth-century accounts — Arago’s chemical one, Donné’s physical one, Moser’s condensation argument and Choiselat and Ratel’s subiodide — and this course adjudicates between none of them. What is not in doubt is that mercury goes where the light went.

One detail in the manual has never been explained and is worth keeping in view precisely because it has not been. The plate is carried at forty-five degrees to the rising current of mercury vapour, and Arago records that if it were horizontal at the moment the mercury precipitates, the picture would have to be looked at at forty-five degrees to show its maximum effect. He speculates about crystalline needles orienting in a liquid or half-liquid and does not claim it. A page that quietly dropped that sentence would make the process look better understood than it was.

Taking the iodide off, which is the hardest fixing problem in the subject

Here the two solutions in the object rejoin the story, and here the modern chemistry can be stated exactly, because it is the same chemistry the rest of this formulary runs on.

Silver iodide is the least soluble silver halide by a wide margin. OpenStax’s solubility-product table gives AgCl 1.6 × 10⁻¹⁰, AgBr 5.0 × 10⁻¹³ and AgI 1.5 × 10⁻¹⁶ at 25 °C; the silver iodide page gives the same fact as a mass, 28 × 10⁻⁷ g/L at 25 °C. It is roughly three thousand times less soluble than the bromide and about a million times less soluble than the chloride.

Thiosulfate dissolves a silver halide by taking the silver ion out of the halide’s solubility equilibrium into a complex ion:

AgI + 2 S₂O₃²⁻ → [Ag(S₂O₃)₂]³⁻ + I⁻

Fixing a daguerreotype, in modern notation

Combining the published solubility product with the published formation constant of 2.9 × 10¹³ for the two-coordinate thiosulfate complex — the course’s own arithmetic, marked as arithmetic, and set out in full on the iodide hypo check page — gives an overall constant for that dissolution of about 0.004, against about 15 for silver bromide.

K = K_sp × K_f = 1.5 × 10⁻¹⁶ × 2.9 × 10¹³ ≈ 4 × 10⁻³

Thiosulfate against silver iodide, from two published constants

Four thousandths. Even hyposulfite, the strongest fixing agent of the century, dissolves silver iodide barely at all compared with what it does to a bromide emulsion. That is the number to hold on to when reading Daguerre’s fifth operation, because it says that the first fixing problem in photography was the worst one the subject contains, and that the man who solved it did so with the one reagent on earth that could.

Against that, common salt. PubChem’s HSDB entry lists silver iodide as soluble in aqueous solutions of sodium chloride, and gives no concentration for it — as it gives none for any of the solvents in that list. So the salt bath is not doing nothing; chloride does complex silver, and a saturated chloride solution is about as much chloride as water can be made to hold. The course does not publish an equation or a constant for that reaction, because it has not read a source that gives one, and writing a plausible one would be exactly the invention Rule 1 forbids.

What can be stated without any invention is the observation, and it comes from the two best-placed witnesses there were. Daguerre, in his own manual: the hyposulfite is preferable because it removes the iodine entirely, which the salt solution does not always do, above all when the plates have been made some time before. Eder, writing the history: in his early practice Daguerre knew only the imperfect fixation with a warm common salt solution, which gave the plates a mottled appearance. A weak solvent, working on the least soluble salt in photography, on a layer that had had time to age.

And Pelouze’s weighing adds the fact neither of them stated: the hyposulfite bath removes not only the iodide but silver. Every thiosulfate fixer since has done the same thing, which is why fixing times matter and why an exhausted bath and an over-long one damage a picture in different ways.

Function of every ingredient

Two substances are ingredients of the object above. Four more are the process and cannot be ingredients of anything, because they are never in solution; they are set out beneath, because a page that explained only the acid and the salt would have explained the packaging and not the photograph.

Nitric acid, one part in sixteen of distilled water

What it is. Nitric acid, HNO₃, a strong mineral acid and a powerful oxidiser; the aggregated ECHA classification carries signal word Danger with H272, H290, H314, H318 and H331, and three pictograms at once — oxidiser, corrosion and skull. In the dilution the manual gives, one volume of it stands against sixteen of distilled water.

Why it is there. It is the last step of the plate preparation and the one that decides whether the second operation works. Pumice and olive oil bring the silver to a mirror; the acid takes off what the polishing leaves behind. Daguerre applies it on a cotton pad, spreads it until the surface carries an even film rather than beads, powders the plate with pumice again, wipes it dry, and repeats — three times over, or twice if the plate is being prepared in advance, in which case at least one more application is indispensable at the moment of use. A plate acidulated the day before is not a plate acidulated now.

What it does chemically. The period explanation is Pelouze’s, reported in Arago’s note and carefully hedged there: the useful influence the acid exerts might well lie in its removing from the surface of the silver the last molecules of copper. That is an interpretation offered in 1839 by a competent chemist, and this page reports it as one. What can be added from modern data without extending it is that nitric acid attacks both of the metals present: CAMEO’s reactivity profile for the acid records that it attacks metals, and that with silver and copper it does so with the production of nitrogen dioxide. On a silver-faced copper plate, an acid that dissolves copper preferentially is doing exactly what Pelouze proposed, and an acid left too strong or too long is dissolving the silver face as well.

What follows on the picture. Everything, by the manual’s own account, because the second operation cannot correct the first. A plate whose acid was unevenly spread carries stains where the globules stood. A plate that was not properly degreased will not take an even iodine layer, and an uneven iodine layer is an uneven picture; Arago’s note is explicit that for the perfect gradation of shadows and lights the thickness of that layer must be exactly the same everywhere.

More or less of it. More acid, or a stronger acid, and the 1840 pamphleteer’s experience applies: his mixture came out stronger than he was used to, it destroyed the fine blacks his first polishing had produced, and he had to add a third and sometimes a fourth pumicing to get them back. He also records the opposite failure — an acidulation badly wiped off, whether with pumice or with tripoli, produces slate-coloured tones. Less acid, or too few applications, and the plate is not stripped: his own test for whether enough has been done is whether the acidulated water spreads without forming droplets.

What it interacts with. The olive oil and the pumice that precede it, which is why the plate is heated between the polishing and the final acidulations: the 1840 account has the plate warmed strongly enough to vaporise the oil and reduce its residue to ash. And, dangerously, everything organic — see Incompatibilities.

Sodium chloride, to saturation

What it is. Sodium chloride, NaCl, common salt, in the state the formula actually specifies: a saturated aqueous solution, filtered clear, with the excess left undissolved in the stock bottle. PubChem’s HSDB figure for the saturation point is 36.0 g per 100 g of water at 25 °C.

Why it is there. It is the fixing agent of Daguerre’s first practice, and the coincidence with London is one of the neatest things in the subject: in the same year, on a different support, in a different country, Talbot’s photogenic drawing paper was being fixed in a strong solution of common salt too. Chloride does complex silver, and PubChem’s HSDB entry lists silver iodide as soluble in aqueous sodium chloride, so the bath is not doing nothing. That two independent inventors reached for brine, and that both baths turned out to leave halide behind — Talbot’s entry sets out what conservation science found still sitting in the highlights of a salt-fixed sheet — is why Herschel’s hyposulphites mattered as much as they did.

What it does chemically. It takes unexposed silver iodide off the plate, and the manual gives you the assay for it: the yellow colour disappears. How completely it does so is the whole question, and the manual answers it honestly in the same paragraph — not always entirely, and worst on plates that have stood. The course states no equation and no equilibrium constant for the chloride reaction, because none of its sources publishes one; what it can state is the comparison in the mechanism section, where even thiosulfate manages only about 0.004 against silver iodide.

What follows on the picture. Iodide left in the layer is iodide that goes on decomposing in light, which is the destruction Daguerre says the operation exists to prevent. Eder describes the visible consequence of an incomplete salt fix in one word: mottled.

More or less of it. There is no more. The bath is saturated, so its strength cannot be increased by adding salt — only by a temperature change that the encyclopaedia shows buys almost nothing, since one gram dissolves in 2.8 mL of cold water and 2.6 mL of boiling. A saturated formula is a formula with no adjustable variable in it, which is a rare and instructive thing to meet: the only things a worker could change were the temperature, the agitation and the time, and Daguerre specifies the first two and refuses to specify the third. Less salt than saturation, on the other hand, changes everything, because a bath already at the edge of what it can dissolve fails completely a little below it.

What it interacts with. The plain-water dip before it, without which — and this is Daguerre’s own statement, given as a fault — the salt water or the hyposulfite makes indelible stains on the plate. And the distilled water after it, because salt dried on a mirror is a permanent mark: he has the plate flushed with not less than a litre of hot distilled water and any surviving drops blown off before they can dry.

The four substances that are the process and are not in the object

Iodine, I₂. The sensitiser, used as a vapour and never as a solution. A capsule of it in the bottom of a closed box, with the crystals divided so the source of the emanation is broad, a gauze stretched over the capsule to regularise the evaporation, and the plate face down over it. It supplies the halide that makes silver photosensitive, and it does so by converting the plate’s own surface rather than by being coated on. Two operating facts in the manual are pure vapour-pressure physics: the box works faster the more it has been used, because the wood inside is penetrated with iodine vapour that comes off it evenly from every surface; and the inside of the box must be at the same temperature as the outside, or a plate going from cold to warm covers itself with a film of moisture that ruins the effect. Iodine is Level D in this course.

Mercury, Hg. The developer, used as a vapour from a warmed pool. About a kilogram of it stands in a capsule under the plate, enough to cover the bulb of a thermometer. Nothing dissolves, nothing is reduced, and no agent is consumed: mercury alloys with silver wherever light has acted. It is Level D in this course, and the reason is in Safety below.

Silver, Ag. Not a reagent but the support, and the only support in this formulary that is also an ingredient. The plate is copper faced with silver — plaqué — and Arago records Daguerre’s claim that the image forms better on plaqué than on a sheet of silver alone, with the suggestion that electricity may play a part in these curious phenomena. The course records that as a period conjecture and adopts none of it.

Sodium thiosulfate, Na₂S₂O₃. The alternative fixer of the fifth operation, published without a strength, and the substance that turned the process from a curiosity into a permanent record. Herschel had described the solvent action of the hyposulphites on silver salts in 1819, twenty years before anyone needed it; his 1819 paper states that newly precipitated muriate of silver is soluble in all the liquid hyposulphites, and in that of soda with great ease and in large quantity. Eder gives the transmission route — Herschel to Talbot, who acknowledged the improvement as early as 1 May 1839, and Daguerre soon after — and Ware records that by the end of 1839 the process was enjoying widespread success having adopted Herschel’s method of fixation with thiosulphate. In the manual it needs no warming and less of it is needed, since it is enough that the plate be covered in the bottom of the dish.

Interactions

Acid and iodine, through the surface between them. The first operation exists for the second, and the manual treats an imperfect plate as unrecoverable: if the iodine layer overshoots to violet, the layer cannot be used and the whole of the first operation must be begun again. There is no salvage step anywhere in the book.

Iodine and the latent image, in both directions. Iodine makes the plate sensitive, and the manual sets a limit of one hour between the second operation and the third, after which the combination of the iodine and the silver no longer has the same property. The same limit applies between exposure and development. Later workers found the reverse effect as well — exposing an already-exposed plate to halogen vapour destroys the latent image, though the plate will take a new one — and the process atlas entry carries that with its sources.

Mercury and the plate, after the picture is finished. Mercury left on the plate ends by adhering to the silver, which is why the manual has it taken off with pumice and oil and frequent changes of cotton each time a plate is reused, and why the little metal bands that clamp the plate down are cleaned after every picture — they carry both an iodine layer and part of the image.

Salt water and hyposulfite against the same layer. These are alternatives and not a sequence, and the paragraph that sets them against each other is a comparison of two fixers made by the one person who had used both on the same material. The hyposulfite wins on completeness, on not needing to be warmed, and on quantity — enough merely to cover the plate in the bottom of the dish, against three centimetres of salt water.

Thiosulfate and the image silver. Pelouze’s weighing, again, because it is the interaction every fixer page in this book has to reckon with. A fixing bath that dissolves the halide also dissolves the image, slowly, and the fact was discovered in the first year of the first process.

Varnish and the amalgam, which is an interaction nobody wanted. Daguerre’s footnote records that he had tried to protect his pictures with varnishes of amber, copal, rubber, wax and several resins. Two things went wrong. Optically, any varnish considerably attenuated the lights and veiled the shadows, which is precisely what a coating does to an image that works by specular against diffuse reflection. Chemically, the mercury was decomposed by combining with the varnishes tried, an effect that only began to develop after two or three months and ended by destroying the picture. The answer he settled on was mechanical: put the plate under glass and seal it.

Variants

Hyposulfite of soda in place of salt (1839, Daguerre’s own). The manual’s own alternative, printed inside the fifth operation and recommended in preference to the bath it replaces. It is the most consequential variant in the history of the subject and it has no quantities, which is why it is not a second solution in the object above. Every thiosulfate fixer in this formulary is a descendant of the choice that paragraph records; the plain hypo fixing bath is the nearest thing to it that this course can publish with numbers.

Bromine and chlorine acceleration (1840 to 1841). The vapours that turned a three-to-thirty-minute architectural exposure into a portrait. They belong to the process rather than to this entry’s two solutions, they are set out in the process atlas, and both are Level D substances in their own right; HSE’s EH40 lists bromine at 0.1 ppm long-term and 0.2 ppm short-term.

Fizeau’s gilding (1840). A bath of hyposulphite of soda containing chloride of gold, which Eder gives as 300 parts of hyposulphite, 1,000 parts of water and one part of chloride of gold, and which greatly increased both the beauty and the permanence of the plate. Fordos and Gélis afterwards identified the double salt as hyposulphite of gold and sodium — later sodium aurothiosulfate, sold in the trade as sel d’or. It has its own planned entry at Fizeau gilding, and it is a Level D context rather than a Level D formula in its own right: the chemistry is gold and thiosulfate, and what makes it unusable here is what it is applied to. Eder dates it to 1840; the AIC’s conservation page gives 1841 for the introduction of gold toning, and this entry follows Eder without claiming the year is settled.

The Becquerel variant (1840). Development by red or yellow light rather than by mercury, on an iodised plate. It is the only route to a daguerreotype that removes the worst hazard in the process, and it does not remove the classification, because the iodine remains; see its process entry.

Potassium cyanide, in later practice. Cyanide entered daguerreotypy for cleaning plates and, in the wet-plate decades that followed, for fixing. This course’s account of it is at cyanide fixing and toning, and it is Level D for reasons the Level D policy states plainly: contact with any acid liberates hydrogen cyanide at once, and a darkroom is a room with an acid in it.

⁇Did the hyposulfite paragraph stand in the first printing?

Two accounts run alongside each other and this page does not merge them.

Eder’s is a sequence: in his early practice Daguerre knew only the imperfect fixation with warm salt solution; Herschel pointed Talbot to the hyposulphites; Talbot acknowledged the benefit as early as 1 May 1839; and Daguerre, learning of it, abandoned salt and adopted hyposulphite of soda in 1839. On that account the manual’s hyposulfite paragraph records a change of practice that had already happened by the time of publication.

The manual’s own text presents the two as alternatives in a single paragraph, with salt described first and at length and hyposulfite offered as a substitute that is “even preferable”.

The complication is bibliographic. Both French scans this course holds carry the Alphonse Giroux et Cie imprint and describe themselves on the title page as corrigée, et augmentée du portrait de l’auteur, so neither is demonstrably the first impression of a book that went through many printings and translations within months. This page therefore does not claim that the paragraph stood in the first printing. What it claims is what it can show: the paragraph stands in two independent scans of the French text and in an English translation whose preface is dated 13 September 1839, and it says hyposulfite is preferable. Anyone wanting to settle the priority question needs to collate printings, which is a bibliographical exercise this course has not carried out.

Safety

This is the section the classification exists for, and it is short, because the argument does not need length.

Mercury. Mercury carries the harmonised classification under Regulation (EC) No 1272/2008 with H330, fatal if inhaled, together with H360D, H372, H400 and H410. HSE’s EH40 gives mercury and its divalent inorganic compounds a long-term workplace exposure limit of 0.02 mg/m³, measured as mercury. Ware’s summary is the plainest sentence anyone has written about it in a photographic context: the vapour is notorious for its insidious neurotoxicity.

The property that defeats intuition is worth stating alone. Mercury is a liquid metal that gives off vapour at ordinary room temperature, and the vapour has neither colour nor smell, so nothing about a room tells you what is in it. NIOSH’s entry sets its respirator ladder against the vapour rather than against the liquid, for that reason.

And then the fact that decides this page. The historical procedure does not merely keep mercury in the room; it deliberately heats it. The manual has about a kilogram of mercury in an open capsule in a wooden box, with an alcohol lamp under it and a thermometer standing in the metal, brought to 60 °C and forbidden to exceed 75 °C, the plate then left over it until the thermometer has fallen to 45 °C. Those figures are printed here once, as the ground of the hazard assessment and not as a direction. A liquid whose room-temperature vapour already requires a 0.02 mg/m³ limit is being warmed by forty degrees in an enclosure that is opened by hand in a small dark room, and there is no combination of gloves, goggles and an open window that answers it. An exposure limit implies monitoring; monitoring implies equipment and competence a domestic darkroom does not have.

Iodine. Iodine’s harmonised classification is milder — signal word Warning, H312, H332 and H400 — but the encyclopaedia entry records the widest classification spread in the whole course behind it: the aggregated ECHA notifications give signal word Danger and ten statements, and a Japanese NITE-CMC block gives H330, fatal if inhaled. EH40 gives a short-term limit of 0.1 ppm. CAMEO’s sentence is the operational one: it emits toxic vapour at room conditions, and the vapour becomes visibly purple as it builds up in a confined space. The second operation is a closed box of exactly that vapour, opened repeatedly to look at the plate.

Nitric acid. The one substance in the object that a supervised laboratory does handle, and the course still classifies it Level C on its own page. Three pictograms at once — oxidiser, corrosion and skull — is an unusual thing to find on a darkroom shelf, and it is the reason. The dilute mixture of this formula is much less aggressive than the concentrate, but the concentrate is what a reader would have to buy, dilute and store in order to make it.

What is not a hazard here, and why it is worth saying. The fifth operation is warm salt water in a copper dish, a hook, and a litre of distilled water. There is nothing in it that would trouble a kitchen: sodium chloride is not on EH40’s list at all, and its classification is a mild one. Naming that plainly is part of an honest assessment and part of the point of this entry. The daguerreotype is not uniformly dangerous; it is dangerous in two of its five operations, and those two are load-bearing. Take out the mercury and you have the Becquerel process, which is still Level D because the iodine is still there. Take out the iodine as well and you no longer have a daguerreotype.

One caution that applies to reading rather than doing. EH40 carries a warning worth repeating in any course like this one: the absence of a substance from the list does not indicate that it is safe.

Storage

The salt solution is made in quantity in advance and kept in corked bottles, filtered before it goes in. No keeping time is published, and none is needed for a saturated solution of a stable salt: what will change in the bottle is the temperature, and with it the amount of undissolved salt lying in the bottom.

The hyposulfite solution is kept in the bottle it is made in and refilled when exhausted. The publisher’s notice is the only capacity statement in the manual: a refill will serve to wash three plates.

The iodised plate keeps for not more than an hour before exposure, and the exposed plate for not more than an hour before development. Those are not storage instructions so much as statements about how quickly the latent image and the sensitised layer decay.

The developed plate, before the fifth operation, keeps for several months without alteration provided it is not often looked at in full daylight. That is worth pausing on, because it is chemically odd and entirely correct: the amalgam image already exists, and the unexposed silver iodide is still sitting on the plate around it, so what is being stored is a finished picture on a sensitive support.

The finished plate goes under glass and is sealed, with paper strips over the joints of the box lid where a traveller cannot frame it properly. Daguerre calls the result inalterable even in sunlight, which is his claim and is chemically defensible and mechanically wrong: Eder records that the early daguerreotypes were generally kept in paper wrappings, which injured the delicate images, and that Daguerre was protecting his pictures under glass as early as 1839. The AIC states the reason in one line — an ungilded plate’s image can be wiped off with a finger. The historical convention that followed is a cover glass held off the surface by a spacer, sealed at the edges, in a case, and the conservation of surviving plates is a specialist field with its own literature.

Incompatibilities

Nitric acid with everything organic. The acid’s page carries the list from CAMEO and the ILO-WHO safety card: alcohols, amines, ammonia, hydrazines, hydrocarbons, powdered metals and thiols, all of which it can ignite on contact; combustible materials and reducing agents generally, because it is an oxidant first and an acid second; and acetic acid, with which CAMEO records that the mixture can explode if not kept cold — which puts it on a different shelf from the stop-bath concentrate. It attacks metals, with silver and copper producing nitrogen dioxide. And it is never diluted the wrong way round; the safety card prints the instruction in capitals.

Read that list against the first operation and something uncomfortable emerges. The plate arrives at the acid carrying olive oil, and both texts answer it by heating the plate in between — Daguerre until a slight whitish film forms on the silver, and the 1840 pamphleteer explicitly to vaporise the oil and reduce its residue to ash. Nobody in 1839 was reasoning about an oxidising acid meeting an oil, and the practice happens to remove the oil before the acid meets it in quantity, but the sequence is not one this course would design.

Iodine with the plate you have already exposed, which destroys the latent image; and with metal fittings and anything above an open dish of it, which it stains and corrodes.

Mercury with the picture, in the specific sense that mercury left on the plate adheres to the silver, and with any varnish, which decomposes it over two or three months.

Thiosulfate with acid, which is the incompatibility that governs every fixer page in this book and is not stated in the 1839 manual because nobody had met it yet: thiosulfate in acid decomposes to sulfur and sulfur dioxide, and the reaction is written out on the 1928 bisulphite fixing bath page. It matters here only because a reader assembling the manual’s two solutions in one room would have a bottle of dilute nitric acid and a bottle of hyposulfite on the same bench.

And the general incompatibility that decides the page. Mercury and iodine together in one small room, one of them warmed, is not a combination that any home control regime reaches.

Waste

Nothing on this page is a waste stream a reader will generate, because nothing on this page is a procedure. What follows is what the substances are, for anyone who encounters them.

The salt bath and its washings are silver-bearing, because a fixing bath’s whole function is to put silver into solution — and Pelouze established in 1839, by weighing the plate before and after, that the hyposulfite bath takes image silver as well as unexposed halide. A spent daguerreotype fixing bath, salt or hyposulfite, is a silver-bearing liquid waste and belongs to a silver-recovery route, exactly as the modern fixers in this formulary do.

Mercury waste is not a domestic waste stream in any jurisdiction the course has looked at. Elemental mercury and mercury-contaminated materials are controlled wastes handled by licensed contractors, and mercury’s own classification carries H400 and H410, very toxic to aquatic life with long-lasting effects.

Iodine’s harmonised classification carries H400 on its own account.

Nitric acid waste is acidic, oxidising and, in this use, metal-bearing — dissolved copper and silver from the plate. The acid’s page sets out why that fails both the pH and the metals limits of an ordinary sewer code, and why it belongs to a licensed route rather than a drain.

Local regulation governs all of it, everywhere, and this course states general practice and never a jurisdiction-specific instruction.

Troubleshooting

The manual is unusually good at faults, because Daguerre was writing for people who had never seen the process and would fail at it. Most of what follows is his; the three rows that are not are attributed in place. Every one of them is a diagnosis rather than a remedy, which is what a Level D page can carry.

What is seen	What the manual says caused it
Indelible stains after the fixing bath	The plate went into the salt water or the hyposulfite without the plain-water dip first
Numerous indelible spots after the final wash	Drops of the wash water were allowed to dry on the plate; the test for suitable water is a drop evaporated on a burnished plate leaving no residue
Stains where the acid stood in globules	The acid was not spread evenly; the places it did not take make stains
Slate-coloured tones	An acidulation badly wiped off, whether with pumice or with tripoli — the 1840 pamphleteer’s diagnosis, not Daguerre’s
The iodine layer violet instead of gold-yellow	Left too long over the iodine; the layer is then less sensitive to light and cannot be used, and the first operation must be begun again
The iodine layer not yellow enough	Too short an exposure to the vapour; the image of nature reproduces only with great difficulty
The layer darker on one side than the other	Uneven vapour; the board is turned end for end, not upside down, when it goes back on the box
A film of moisture on the plate in the iodine box	The inside of the box was at a different temperature from the room
Highlights not white but blackened	The camera exposure was too long
The picture vague and without detail	The camera exposure was not sufficient
Fine blacks lost after acidulation	The acid was stronger than the one previously used; the 1840 account recovers them with a third and sometimes a fourth pumicing
Highlights attenuated and shadows veiled, worsening over two or three months	A varnish was applied; the mercury is decomposed by combining with it
The image wiped away by a touch	Ungilded amalgam on a bare metal surface — the AIC’s observation, and the problem gilding was invented to solve

Two of those deserve a sentence more. The mottled appearance Eder attributes to the warm salt fix is not in Daguerre’s own fault list, for the obvious reason that it is a fault of the method rather than of the operator. And the yellow colour disappearing is the only positive end point in the whole manual; everything else is judged by a colour that should not have been reached, a time that should not have been exceeded, or a stain that should not be there.

Experiments

None of these is a daguerreotype, and none of them requires mercury, iodine or nitric acid. They are the parts of this page a reader can actually put a hand on.

Fix a silver iodide against a silver bromide, and watch the difference the mechanism section predicts. The iodide hypo check exists because thiosulfate handles the two halides about three thousand times differently, and the same arithmetic is the reason Daguerre’s fifth operation was hard. Run the check on a fresh fixer and on an exhausted one and the equilibrium in that page’s equations becomes something you have seen rather than something you have read.

Measure a saturated solution instead of trusting the number. Make a saturated salt solution Daguerre’s way — excess salt, shaken, filtered — then evaporate a weighed volume of the filtrate to dryness and weigh the residue. Compare with the encyclopaedia’s 36.0 g per 100 g of water at 25 °C, note your temperature, and note how much the answer moves between a cold room and a warm one. The point of the exercise is the one the formula makes: saturation is a specification with no adjustable variable, and the only way to hit it is to leave an excess behind.

Test a wash water the way the manual does. Put one drop of your tap water and one of distilled water on a piece of polished metal or clean glass, evaporate them with gentle warmth, and look at what is left. Daguerre’s instruction is exactly this, and the residue ring is the reason every alternative-process page in this course cares about water.

Reproduce the ambivalence without making a plate. Find a daguerreotype in a museum collection — the Library of Congress and the Met both publish theirs online, and many collections have them on open display — and, in front of the object, work out which way you have to stand for the picture to read positive. Then write down what the highlights and the shadows are made of. If the account in Image characteristics above is right, you should be able to predict the angle before you find it.

Follow the weight. Pelouze’s experiment is the most elegant thing in the 1839 literature and it needs only a balance: he showed the plate gains no weight from the iodine, gains from the mercury, and finishes lighter than it started after the hyposulfite bath. Design the modern equivalent for a sheet of film — weigh it dry, expose and develop it, fix it, wash and dry it, weigh it again — and predict the sign of the change before you run it. The prediction is the exercise; the balance you need is the one on the planner.