Talbot's photogenic drawing paper

A chemist asked in 1834 how to put silver chloride on paper would have said: dissolve your salt, dissolve your silver nitrate, and use the proportions that convert one exactly into the other. Ware computes what those proportions are — silver nitrate needs just over a third of its own weight of salt, 34.4 per cent — and records what happens to a paper made that way. It turns lilac, stops, and leaves a feeble impression that is not a photograph.

Talbot supplied about one twentieth. Everything else on this page follows from that.

The salting solution — "a weak solution of common salt" — the halide, deliberately far short of the silver

Ingredient	Quantity	Form the source specifies
Sodium chloride	6–23 g	0.6 to 2.3 per cent w/v, the band Ware reads from Notebook P; Talbot's customary trough held 1.1 per cent, which is 11 g/L, of domestic culinary salt
Water	to make 1000 mL	Talbot states no volume and no batch size; he states a strength, and not even that numerically. The litre is the course's scaling of Ware's per cent w/v, which is grams in 100 mL of finished solution.

The sensitising solution — "not saturated, but six or eight times diluted with water" — the silver, in gross excess, and the source of the halogen acceptor

Ingredient	Quantity	Form the source specifies
Silver nitrate	180–230 g	18 to 23 per cent w/v, Ware's reading; Notebook P records 20 per cent and 11 per cent. A saturated solution is around 150 per cent w/v, so Talbot's "six or eight times diluted" lands between about 19 and 25 per cent
Water	to make 1000 mL	As above, a strength rather than a batch. Talbot spread this on one surface only, presumably with a brush, and dried the sheet at the fire.

The strong salt solution — "a saturated solution of salt" — Talbot's usual fixer, and the salt half of the alternate washings that made his camera paper

Ingredient	Quantity	Form the source specifies
Sodium chloride	317 g	31.7 per cent w/v, Ware's figure and explicitly his assumption that "strong" means saturated at room temperature; Talbot's own word is "saturated"
Water	to make 1000 mL	Talbot immerses the exposed picture, wipes off the superfluous moisture and dries it. He specifies no rinse in water afterwards, and Ware notes that there was none.

The iodide fixer — "iodide of potassium, much diluted with water" — the first fixer Talbot found, in 1834, and the one he warned about

Ingredient	Quantity	Form the source specifies
Potassium iodide	10–20 g	1 to 2 per cent w/v, deduced by Ware from Notebook P; Talbot gives no figure at all
Water	to make 1000 mL	Talbot "washed over" the picture with it, which Ware reads as a watercolourist's wash rather than a bath. No final rinse is mentioned.

Used in this order — Ordinary photogenic paper, fixed by chloride — Talbot's usual method

1. The salting solution — "a weak solution of common salt" — about two minutes — The sheet is dipped and wiped dry, "by which the salt is uniformly distributed throughout its substance". Talbot gives no time; the two minutes are Malone's, describing practice at the Reading Establishment, as reported by Ware.
2. The sensitising solution — "not saturated, but six or eight times diluted with water" — Spread on one surface only and dried at the fire. "When dry, the paper is fit for use."
3. The strong salt solution — "a saturated solution of salt" — After exposure. The picture is immersed, the superfluous moisture wiped off, and the sheet dried. No water rinse is specified.

"I dip it into a weak solution of common salt, and wipe it dry, by which the salt is uniformly distributed throughout its substance. I then spread a solution of nitrate of silver on one surface only, and dry it at the fire. The solution should not be saturated, but six or eight times diluted with water. When dry, the paper is fit for use." And, for the fixing: "It consists in immersing the picture in a strong solution of common salt, and then wiping off the superfluous moisture, and drying it."

The exposure falls between the second and third steps and is not a bath, so the schema cannot hold it. The published times are in the development table below, whose field name is wrong on this page for the same reason.

Used in this order — Ordinary photogenic paper, fixed by iodide instead

1. The salting solution — "a weak solution of common salt" — about two minutes — As above.
2. The sensitising solution — "not saturated, but six or eight times diluted with water" — As above.
3. The iodide fixer — "iodide of potassium, much diluted with water" — After exposure, washed over with the dilute solution. Talbot warns that if it is too strong it attacks the dark parts of the picture, and that the proper proportions have to be found by trial.

"The first thing which gave me a successful result was the iodide of potassium, much diluted with water. If a photogenic picture is washed over with this liquid, an iodide of silver is formed which is absolutely unalterable by sunshine. This process requires precaution; for if the solution is too strong, it attacks the dark parts of the picture."

Talbot calls this the first successful preserving process he found, in 1834, and says the lace specimen he showed the Royal Society had been preserved this way five years earlier. By 1844 he had published its failure himself.

Used in this order — The far more sensitive paper Talbot used in the camera

1. The strong salt solution — "a saturated solution of salt" — The finished ordinary paper is washed with the saturated salt solution and dried, at which point its sensibility is greatly diminished and sometimes "seems quite extinct".
2. The sensitising solution — "not saturated, but six or eight times diluted with water" — It is then washed again with a liberal quantity of the silver solution and dried, and becomes sensitive again — "and even more so than it was at first". The pair is repeated.

"In this way, by alternately washing the paper with salt and silver, and drying it between times, I have succeeded in increasing its sensibility to the degree that is requisite for receiving the images of the camera obscura."

The number of repetitions is not stated by Talbot or recoverable from Ware. In the Pencil of Nature Talbot adds that the paper was also used moist in the camera, and that the two changes together brought a bright-day exposure down to ten minutes.

☠What this page will not tell you how to do, and why the reason is not the usual one

This is a historical-study entry. It carries no working procedure, and the reason has nothing to do with hazard: at Level B, silver nitrate and common salt are exactly the chemistry the course teaches for real under salted paper.

The reason is evidential. Talbot published no quantity of any kind. His account of 21 February 1839 is complete, primary and entirely qualitative: “a weak solution of common salt”, a silver solution “not saturated, but six or eight times diluted with water”, “a saturated solution of salt”, “iodide of potassium, much diluted with water”. Every number in the table above is Mike Ware’s reading of Talbot’s unpublished manuscript Notebook P, or of Thomas Malone’s account of the Reading Establishment, or — for the strong salt bath — Ware’s stated assumption that “strong” meant saturated. Printed as a recipe, those numbers would be a modern reconstruction wearing a nineteenth-century name, and Rule 6 forbids that as squarely as it forbids inventing one.

There is a second reason and Talbot supplies it himself. He says the whole art lies in finding, by trial, the one proportion that “answers best and gives the maximum effect”, and that he prepared sheets with proportions slightly different from one another, numbered them, exposed strips side by side for a quarter of an hour, and chose the winner. A formula whose author tells you it has to be found each time is not a recipe, and pretending otherwise would misrepresent the process as much as inventing a weight would.

If you want to make a print by this chemistry, the course’s actionable route is the plain salted paper salting solution and the salted paper sensitiser, with a thiosulphate fixer rather than either of Talbot’s. Read this page for what those two bottles are for.

Purpose

To assemble silver chloride inside the surface fibres of a sheet of writing paper, and to leave a large excess of unreacted silver nitrate around every crystal of it.

The first half of that is forced by physics. Silver chloride is insoluble in water and in ordinary solvents, so it cannot be dissolved and brushed on; Reilly states the consequence flatly, that it has to be formed in place by treating the paper first with a soluble chloride and then with silver nitrate. Every salted paper, every albumen paper and every printing-out paper of the nineteenth century inherits that indirection from this formula.

The second half is Talbot’s, and it is the part nobody would have predicted. The excess silver is not a margin of error or an economy; it is a reagent, and the section on the mechanism is about what it does. Ware puts the discovery in one sentence: a strong response to light depends on the amount of halide being substantially less than that required for chemical equivalence to the silver nitrate. Talbot put it in one sentence too, ten years after finding it, and his is better — “a lesser quantity of salt produced a greater effect”.

There is a third purpose, which is this course’s rather than Talbot’s. This is the formula that makes printing out intelligible. Everything a modern silver material does — a latent image, a developer, a fixed speed, an enlarger — is a departure from what this paper does, and the departures are only legible if you understand the thing they departed from.

Recommended uses

Historical, all of them. Talbot lists his own in the paper of 31 January 1839, and the list is worth reading because it is the first published account of what photography was thought to be for.

Contact photograms of flat objects — “flowers and leaves, either fresh or selected from my herbarium”, lace, textiles, painted glass. The sheet is dried, the object laid on it, a sheet of glass pressed down tight, and the whole put in the sun until the open ground goes dark. Ware gives about ten minutes in direct sunlight. This is the use the paper is actually good at, and the surviving work proves it: Talbot showed a piece of lace to visitors who replied that they were not to be so easily deceived, for it was evidently no picture but the piece of lace itself.

Copying engravings and manuscripts by pressing the original face-down on the sheet. Half an hour through thick paper. Talbot notes that the copy is tonally reversed, and then says the thing the whole of photography turns on: preserve that copy so it will bear sunshine, use it as the object for a second copy, and the lights and shadows come back to their original disposition. That is the negative-positive principle, stated in 1839 as a way of multiplying rare prints cheaply.

The solar microscope — a quarter of an hour, at seventeen linear diameters, and Talbot’s argument for it is that a microscopist could not draw what he saw in less than days.

Silhouette portraits and casts of sculpture, both proposed rather than developed.

The camera obscura, badly. An hour or two at a hundred yards from a building in a large box camera; half an hour in the small cameras of 1835; ten minutes on the best paper he had, used moist. The negatives are underexposed outlines and Talbot said so in print.

And, in this course, as the thing to understand before the tray. The one recommended use of this page is to read it beside the plain salting solution: the shape of the modern formula is Talbot’s exactly, and its numbers are not.

When another formula is preferable

Every time, for every practical purpose. The list is short because the honest answer is “always”, and what changes is only the reason.

• To make a salt print today, the plain salted paper salting solution and the salted paper sensitiser. They are the same two-bath chemistry with three differences that matter: published quantities, a trace of gelatin so the result does not depend on what the mill happened to size the paper with, and a fixer that removes the residual halide instead of leaving it in the sheet.
• To fix anything at all, a thiosulphate fixer or Reilly’s alkaline fixing bath. Talbot’s two fixers both fail, in opposite directions, and he published both failures himself. Ware’s measured comparison is not close: a threshold exposure of order 200 lux hours for a salt-fixed sheet against 30 kilolux hours or more for a thiosulphate-fixed one.
• For a different surface and a different scale, the arrowroot salting solution or the albumenised salting solution. Each page carries Reilly’s own comparison of what its binder buys and what it costs; both are the same silver chemistry with the binder changed.
• For a camera negative, anything else whatever. The calotype is the historically correct answer and Ware puts the gain at about a hundredfold, because it stops spending the entire exposure on visible silver and spends a little of it on a latent image instead. A modern film is a further factor of thousands.
• For a hand-coated print that costs almost nothing and needs no fixer at all, the classic cyanotype sensitiser. It answers a different question, but it answers it reliably.
• For a paper that behaves like a modern one, the silver chloride contact emulsion — a developed image, in a binder, on a support, with a fixed speed.

Mixing

There are two chemistries here and they must never meet in a vessel. That is the structural fact about this formula, and it is why the entry is several solutions rather than one — two salt baths at opposite ends of the strength range, a silver bath, and an iodide bath, each made and kept on its own. Silver nitrate and sodium chloride in the same beaker give a useless curd of silver chloride; in the same sheet of paper they give a photograph. Everything else follows from keeping them apart until the paper brings them together.

What Talbot describes, in his own order. Choose paper of good firm quality and smooth surface — he does not know that anything answers better than superfine writing paper. Dip it in the weak salt solution and wipe it dry, “by which the salt is uniformly distributed throughout its substance”. Spread the silver solution on one surface only and dry it at the fire. When dry, the paper is fit for use. That is the whole published preparation, and it is four sentences long.

What is missing from those four sentences is every number. There is no weight, no volume, no immersion time, no drying temperature and no keeping time. Ware supplies concentrations from Notebook P and Malone supplies “about two minutes” for the dip, and neither of those is Talbot publishing a procedure. The course will not close the gap by inventing the rest.

🧮Where the numbers came from, and what they actually say about the ratio

Ware states the strengths as per cent w/v, which means grams of solute in 100 mL of finished solution. Scaling to a litre is multiplication by ten and nothing else:

1.1 % w/v = 11 g of sodium chloride per litre = 0.19 mol/L

Talbot's customary salting trough

20 % w/v = 200 g of silver nitrate per litre = 1.18 mol/L

The sensitiser, at Ware's midpoint

0.19 mol/L × 169.87 g/mol = 32 g/L, a 3.2 % bath

Silver that would exactly consume that chloride

So the bath Talbot used carries about six times the silver needed to convert the chloride it meets, and roughly five sixths of it survives the reaction as free silver nitrate lying in the paper. Ware arrives from the other side, by weight: the silver nitrate encounters about one twentieth of its own weight of salt, against the 34.4 per cent that chemical equivalence asks for. Eleven grams against two hundred is one eighteenth, which is as close as either reading deserves to be pushed.

One step in that is the course’s, and it is an assumption: going from bath strengths to what is inside the sheet assumes the paper carries away comparable volumes of the two solutions, and nobody has published that measurement. The direction is not an assumption, because three sources state it outright.

Now set the modern formula beside it, and take care not to mistake two sixes for each other. Reilly’s salting solution is 2 per cent against a 12 per cent silver bath, and Reilly describes Talbot’s proportions as “about six times more nitrate” — but that six is a ratio of bath strengths, 12 against 2. Talbot’s bath-strength ratio is 20 against 1.1, which is eighteen. The molar comparison says the same thing:

2 % w/v NaCl = 0.34 mol/L; silver to consume it = 58 g/L, a 5.8 % bath; the bath he specifies is 12 %

Reilly, worked the same way

Reilly’s silver bath is about twice what his own chloride needs. Talbot’s is about six times. He ran his paper roughly three times richer in free silver than the modern formula does, which is exactly consistent with the boundary he describes — a sheet that darkens by itself if you push a little further. He was working at the edge deliberately, because he needed every scrap of speed. Reilly, writing for people who want a repeatable print rather than a possible one, backed away from it. What has not changed in a hundred and ninety years is the sign of the deficiency; the size of it has changed a great deal.

The paper is an ingredient and Talbot names it as one. Ware finds from watermarks that he commonly preferred J. Whatman’s Turkey Mill, a gelatin-sized rag paper, and from his letters that he thought the product of some years better than others. This matters more than it looks: the Getty Conservation Institute’s account of salt prints records that English mills preferred gelatin as an internal size while French and other continental mills used starch, so the same two solutions on two sheets bought in two countries are not the same material. Reilly makes the same point from the other end and applies it to Talbot directly: Talbot’s prints were reddish brown because the gelatin was already there as the manufacturer’s sizing, not because he put any in, and French photographers found their papers would not behave like English ones for exactly that reason. Talbot added no binder of his own; the mill supplied it. That is why the modern formula adds two grams of gelatin per litre on purpose — to stop the result depending on a decision made in a paper mill.

Drying at the fire is not a detail. Ware records it as radiant heat from a fire. It matters because the water content of the sheet is half the halogen acceptor, and because a sheet dried in the dark keeps and a sheet left out does not — Talbot warns that his sensitive paper, if prepared by daylight, “must by no means be left uncovered, but as soon as finished be shut up in a drawer or cupboard and there left to dry, or else dried at night by the warmth of a fire”.

Behaviour

It is unstable by design, and Talbot says so in the plainest possible terms. A paper made with enough excess silver is a paper close to darkening on its own: “It happens sometimes that the chloride of silver is disposed to darken of itself, without any exposure to light: this shows that the attempt to give it sensibility has been carried too far. The object is, to approach to this condition as near as possible without reaching it.” That is a description of working at the edge of spontaneous reduction, and it is the reason no single number can be the formula.

Sheet-to-sheet variation is the dominant behaviour, not a fault. Talbot’s own quality-control routine is in the published letter: prepare a number of sheets with proportions slightly different from one another, cut a piece from each, number them, lay them side by side in very weak diffused light for a quarter of an hour, and put the winner in the camera. A modern reader should notice what that implies — he could not predict which sheet was best, so he tested. It is a ranking rather than a measurement, and it is the direct ancestor of every step-tablet exposure in sensitometry.

It prints out, so you watch it and stop. There is no development, no fixed time and no temperature. The image arrives and keeps arriving as long as light falls on it, which is why a printing-out image is self-masking and why its scale is so long: the silver already formed in the shadows shades the silver chloride beneath it.

It stops arriving, though, and where it stops is the whole story. Ware’s figure for the limit in a pure silver halide crystal with no halogen acceptor is a density of about 0.02 — a barely perceptible darkening — because the photolytic silver particles cannot grow past about 10 nm inside the lattice. Talbot’s excess silver is what lifts the ceiling.

In the camera it barely works and Talbot knew it. Ware lists six conditions that all had to hold at once: excess silver, a small format, a wide aperture, patience, a bright subject rich in blue and ultraviolet, and some way of stopping the image changing. Neglect any one and the attempt fails. The famous “first negative” is 36 by 28 mm.

Image characteristics

A negative, and Talbot had to be willing to look at one. The images are white on a coloured ground. Ware makes the point that an aesthetic tolerance of the tonally inverted image was itself one of Talbot’s contributions, along with seeing its potential for reversal printing. It is easy to underrate: nothing in the chemistry obliged anybody to accept a picture with its lights and darks the wrong way round, and Talbot’s own §11 is where he works out what to do about it.

The ground colour is a variable, not a constant. This is the most surprising sentence in the 1839 paper and it is nearly always left out of summaries. By “merely varying the proportions and some trifling details of manipulation” Talbot could get sky-blue, yellow, rose-colour, brown of various shades, or black; green alone was absent except as a very dark shade. He liked the blue variety, which “has a very pleasing effect, somewhat like that produced by the Wedgwood-ware”, and records that it is subject to no spontaneous change and needs no preserving process at all. The course cannot tell you what those compounds were and neither could he: “These different shades of colour are of course so many different chemical compounds, or mixtures of such, which chemists have not hitherto distinctly noticed.”

After fixing, the colour belongs to the fixer. Ware’s three cases, from his accounts of the three fixing methods and from modern replicas — he cautions that too few originals have been confidently identified to serve as archetypal specimens, and that no guarantee exists that surviving works look as they did:

Fixed with	Highlights	Image
Strong sodium chloride	Pale lilac, deepening to mauve or violet with sunning	Deep maroon
Potassium iodide	Pale primrose yellow	Fades over days, until the whole picture is obliterated
Sodium thiosulphate	Clear white	Reddish brown

Talbot preferred the first, and Ware records that he went on halide-fixing for three or four years after Herschel’s better answer existed — partly for the colour, partly because hypo was at first costly and scarce.

A curiosity worth knowing because it identifies the process. Talbot noticed that the primrose yellow of an iodide-fixed sheet “has the extraordinary and very remarkable property of turning to a full gaudy yellow whenever it is exposed to the heat of a fire, and recovering its former colour again when it is cold”. That is a reversible thermal colour change in silver iodide, observed and published in 1839. Talbot offered no explanation for it and this page has no sourced one to offer either; it is recorded because it is a diagnostic that an iodide-fixed sheet is what you are holding.

Matte, sunken and low in maximum density, for the reason every unbound silver image on paper is: the silver is inside the fibres, light reflected off the print is scattered by them, and the darkest areas look paler than the same quantity of silver would in a transparent binder. It is also why the sheet stays translucent enough to print through, which is the whole reason the negative-positive idea was available at all.

The mechanism

Step one, in the paper: a double replacement. The salted sheet holds sodium chloride through its substance. The silver solution arrives on one face, and because silver chloride is almost insoluble it precipitates on the spot.

Ag⁺ + Cl⁻ → AgCl

Silver chloride formed in situ, which is the only way it can be formed

The sodium nitrate takes no part and washes out later. What is left in the sheet is fine silver chloride, distributed as deep as the salting solution soaked, surrounded by roughly five times as much unreacted silver nitrate as was consumed in making it.

Step two: light, and a reaction that wants to run backwards. Blue and ultraviolet light reduce silver(I) to metal and oxidise chloride to chlorine.

2 AgCl + light ⇌ 2 Ag + Cl₂

Photolysis, and the back-reaction that stops it

Ware is emphatic that the second arrow is real. Chlorine is a strong oxidant, and with nothing to take it away it re-forms silver chloride from the silver just made. In a pure crystal the process therefore self-limits at about 10 nm of photolytic silver and an optical density around 0.02 — a darkening you would struggle to see. A paper made to the chemically correct proportions has nothing to remove the chlorine, and that is why it fails.

Step three: the free silver nitrate and the paper’s own water take the chlorine away. Cellulose paper carries about 8 per cent water at ordinary humidity. Chlorine disproportionates in water, and in the presence of free silver ions the equilibrium is driven hard to the right because silver chloride is so insoluble.

Cl₂ + H₂O + Ag⁺ → AgCl + HOCl + H⁺

Ware's halogen acceptor: water plus silver(I), acting together

The chloride made in that reaction meets more free silver and becomes fresh silver chloride, which light breaks down again. The halogen is partially recycled and the photolysis extends into the free silver ions in the surroundings. Ware’s conclusion is that the combined action of water and silver(I) ions is the major halogen acceptor and contributes the greater part of the image silver. Vogel reached the same conclusion in the nineteenth century by pure reasoning, and Reilly quotes him. More free silver nitrate therefore means more image silver and a greater maximum density, which is the practical form of everything above.

Step four: why the ratio, and not just the quantity, decides. A precipitated silver halide is not stoichiometric at its surface. Whichever ion is in excess in the surrounding liquid is adsorbed onto the crystal, and Ware sets out the two cases.

• Silver in excess — Talbot’s case. The crystal surface carries adsorbed Ag⁺ and a net positive charge, which attracts photoelectrons to the surface. Silver specks form and grow there, outside the constraints of the lattice, and the positive holes reach the surface to release halogen into a medium that will take it. Ware calls this “sensitized” silver halide.
• Halide in excess — the fixed case. The surface carries adsorbed chloride and a net negative charge, which repels photoelectrons inward; released halogen no longer meets free silver ions, so it is not scavenged and re-oxidises the surface silver instead. Only silver formed deep inside the crystal survives, and the sheet turns a dull violet and stops.

That is the entire formula, restated. Talbot’s 200 g/L silver bath would be exactly consumed by 69 g/L of chloride — a 6.9 per cent salting solution — and he used 11 g/L. The sheet salted at 69 g/L would not hold less silver chloride than his; it would hold silver chloride in the wrong ionic environment, and it would barely print.

⚗Talbot's two fixers are the same idea, run backwards

Talbot reasoned that if a deficiency of salt gives sensitivity, an excess should take it away. He was right, and the mechanism is the bullet-pointed pair above: the strong salt bath floods the sheet with chloride, the residual silver halide flips from adsorbed-silver to adsorbed-halide, and it stops responding usefully to light.

Ag⁺ + Cl⁻ → AgCl

The strong salt bath: the free silver nitrate is converted where it lies

Nothing is removed. Ware notes that the strong chloride also dissolves a small proportion of the silver halide by complexation — AgCl + Cl⁻ → AgCl₂⁻ → AgCl₃²⁻ — but not nearly enough to matter. The highlights of a salt-fixed photogenic drawing still contain large amounts of silver chloride, and under strong light with any ultraviolet in it they still print out slowly, which is the lilac veil.

Iodide works the same way and harder, because silver iodide is more insoluble than the chloride and iodine barely disproportionates at all — Ware gives equilibrium constants of about 500 for chlorine, 0.01 for bromine and 10⁻⁹ for iodine.

AgCl + I⁻ → AgI + Cl⁻

Iodide fixation: the residual chloride is converted to the far less sensitive iodide

And that is exactly why it destroys the picture. In the presence of excess iodide the nanoparticle silver of the image is easily oxidised by atmospheric oxygen back to pale silver iodide, faster in humid conditions and faster still in light. So an iodide-fixed sheet fades towards its ground while a chloride-fixed one fogs towards its image. Talbot published his first fixer’s failure himself, in 1844, in his own words: “the fixing process by iodine must be considered as not sufficiently certain to be retained in use as a photographic process.”

Herschel’s thiosulphate is a different kind of answer altogether — it removes the halide as a soluble complex rather than converting it — and Ware records that Talbot and Herschel, both careful about language, called Talbot’s treatments fixing and Herschel’s washing out. Modern usage has swapped the names round and calls Talbot’s methods stabilisation.

⁇Talbot's explanation was wrong, and it is worth knowing exactly how

Talbot’s account, 1844. What a deficiency of salt makes is “an imperfect chloride, or (perhaps it should be called) a subchloride of silver” — a distinct compound with less chlorine in it than silver chloride has, sensitive where the full chloride is not. He extends the same idea to the iodides, arguing that Davy must have observed “a sort of subiodide in which the iodine was deficient as compared with the silver”.

The modern account, from Ware. There is no subchloride. The crystal is ordinary silver chloride; what the excess silver changes is the surface, by adsorbing Ag⁺ onto it, and the environment, by supplying a halogen acceptor. The compound is the same; its ionic surroundings are not.

Why the distinction earns a callout. Talbot’s wrong explanation gave the right instruction, and for a century the instruction was all anyone had. Ware’s remark, in his account for the Talbot Catalogue Raisonné, is that the detailed mechanism was not understood until chemical science had reached the middle of the twentieth century, “which accounts for the historiographic observation that silver photography, for the first century of its life, was led by empirical practice rather than chemical theory”. This is the founding example of that. It is also a warning for anyone reading nineteenth-century manuals: the observations in them are frequently excellent and the mechanisms frequently are not, and the two have to be separated line by line.

Talbot’s own honesty on the point deserves quoting. Of the white insensible patches he could not explain, he wrote: “With regard to the theory of this, I am only prepared to state as my opinion at present, that it is a case of what is called ‘unstable equilibrium’.” That is a man saying he does not know, in print, in the paper announcing his invention.

Function of every ingredient

Sodium chloride, 6 to 23 g per litre in the salting bath — 0.6 to 2.3 per cent w/v. The halide, and the only thing here that becomes part of the light-sensitive substance. Dried through the sheet, it waits for the silver and is converted to silver chloride exactly where it lies. Talbot used ordinary domestic culinary salt; Ware’s reading of Notebook P puts his customary trough at 1.1 per cent, and Malone’s account of the Reading Establishment at 0.7 to 1.4 per cent depending on the paper.

Its quantity is the formula. More chloride makes more silver chloride, and up to a point that means more image; Reilly’s modern recommendation of 2 to 2.5 per cent for salted papers sits at the top of Talbot’s band and a little above it, and his reason is that lowering the chloride costs brilliance and density. But more chloride also consumes more of the silver bath, and past the point where the free silver runs short the print-out mechanism starves — which is the reading of the mechanism section, and it is also what Talbot observed without being able to explain: “If the strength of the salt is augmented beyond this point, the effect diminishes, and, in certain cases, becomes exceedingly small.” Less chloride gives a thinner image with less silver in it to reduce. The optimum is a maximum, with failure on both sides, and finding it is what his numbered strips were for.

Silver nitrate, 180 to 230 g per litre in the sensitiser — 18 to 23 per cent w/v. It does two jobs and the second one is the discovery.

Job one: it converts the chloride in the sheet to silver chloride, at the place the chloride is lying. Job two: everything it does not convert — about five sixths of it — stays in the paper as free silver nitrate and acts as the halogen acceptor, taking up the chlorine that photolysis releases so that the darkening can continue instead of reversing. Ware notes separately that silver nitrate on its own is not photosensitive at all: nitrate is fully oxidised and water is too hard to oxidise at that potential, so a silver nitrate paper only works when something oxidisable, such as the cellulose or the size, is present. It is not the sensitive substance. It is what lets the sensitive substance keep going.

More silver means more free silver, a higher acceptor concentration, more recycled halide, more image silver and a denser print — up to the point where the paper darkens spontaneously, which is Talbot’s stated boundary. Less silver means a paper that stalls, and at chemical equivalence it stalls almost completely. The gearing is steep enough that Talbot could see the difference at the edges of a badly brushed sheet, which is how he found it.

Sodium chloride again, 317 g per litre in the strong bath — 31.7 per cent w/v, saturated. The same substance, the same ion, the opposite job, and Talbot found this “sufficiently singular” enough to remark on it: “the same substance which is so useful in giving sensibility to the paper should also be capable, under other circumstances, of destroying it.” What changes is which ion is in excess when the light arrives. At 1 per cent it is silver; at 32 per cent it is chloride, and a chloride-jacketed crystal repels photoelectrons and cannot get rid of its halogen. Stronger is better here as far as saturation and no further; weaker leaves free silver unconverted and the sheet keeps printing. Ware’s 31.7 per cent is his assumption about the word “strong”, flagged as such by him and again here.

Potassium iodide, 10 to 20 g per litre — 1 to 2 per cent w/v. The alternative fixer, and the first one Talbot found. Iodide converts free silver nitrate to silver iodide and converts much of the residual silver chloride to silver iodide as well, because the iodide is the more insoluble; silver iodide is effectively insensible to light with no acceptor present, so the picture stops changing. Talbot’s own warning is the important part of the entry: “if the solution is too strong, it attacks the dark parts of the picture.” More iodide does two damaging things, and Ware separates them: it dissolves some of the silver iodide again by complexation (AgI + 2I⁻ → AgI₃²⁻), and — the fatal one — in the presence of excess iodide the nanoparticle silver of the image is oxidised by atmospheric oxygen back to pale silver iodide, faster when it is humid and faster still in light. That second reaction is the slow fading that killed the method. Less iodide leaves free silver unconverted, and the sheet is not fixed. There is no setting at which it is reliable, which is why Talbot abandoned it and said so.

Water, in all four baths, and it is not a diluent. Three separate jobs. In the salting solution it carries the chloride into the substance of the sheet, which is why the salt is dipped rather than brushed. In the sensitiser it is what the per cent w/v is measured against, so a solution made up short is a stronger solution and a different formula. And in the finished dry sheet the residual water — about 8 per cent by weight in cellulose at ordinary humidity — is half of the halogen acceptor in Ware’s mechanism. That last one has a practical edge, though the connection is the course’s rather than either author’s: Ware gives water as half the halogen acceptor, and separately records that Talbot’s best camera results came from paper used moist. Neither of them writes the sentence that puts the two together, and this page marks it as a reading.

The paper, which Talbot names and the schema cannot hold. Superfine writing paper of good firm quality and smooth surface, in practice J. Whatman’s Turkey Mill, a gelatin-sized rag stock. The size is the binder this formula never mentions: it holds the sensitiser and therefore the image in the surface fibres instead of letting it sink, and Ware suggests it may also protect the colloidal silver by surface adsorption and so influence the colour and stability of the print. What it does not do, on Ware’s evidence, is act as the halogen acceptor — gelatin is known not to be an effective scavenger of halogen at print-out exposure levels, so the sizing is not essential to the photochemistry, which can be demonstrated. Reilly’s 1980 account disagrees on that point, and the plain salting solution page sets the disagreement out in full rather than picking a side.

Interactions

Salt with silver, which is the only interaction that matters and never happens in a beaker. The two solutions meet inside the paper and their ratio is the design of the process. Change one without the other and you have made a different material: a weaker salting bath against the same silver gives a thinner but more vigorously printing sheet, a stronger one starves the acceptor and the image stalls short. Reilly states the modern form of the same rule — the sensitiser about four times the strength of the salting solution — and attributes the six-times version to Talbot.

Silver with the mill’s sizing. Two grams per litre of gelatin added deliberately, as in the modern formula, is a small adjustment to a gelatin-sized English sheet and the only protein present on a starch-sized French one. Talbot added none at all, so his results were entirely at the mercy of the stock, which is exactly why he cared which year’s Whatman he had.

Everything with the water in the sheet. Ware’s halogen acceptor is water and silver ions together. A sheet baked dry at the fire and used at once behaves differently from one equilibrated in a damp room, and a sheet used deliberately moist — Talbot’s 1835 trick — behaves differently again. This is a real variable that no nineteenth-century account controls.

The fixers with the free silver, and the order that follows. Both of Talbot’s fixers convert the free silver rather than removing it, so they can be applied straight to the exposed sheet. Thiosulphate cannot: Ware’s first pitfall is that the excess silver nitrate must be washed out with water before the hypo goes on, or thiosulphate reacts with silver nitrate and stains the print brown with silver sulphide. Herschel understood this and Talbot noted it in Notebook P, and Ware suspects Talbot did not always take enough care over it — which may be part of why he disliked hypo.

The exposure with the acidity. Ware points out a consequence of the disproportionation equations that is easy to miss: they generate hydrogen ions, so the sheet becomes more acidic during exposure, and that acidity could assist re-oxidation of the image silver by the nitrate present. It is one of the candidate slow back-reactions that degrade print-out images over decades, and it is also the proposed reason the ammonia-based variant below gives a stronger image — the ammonia buffers the acid away.

Variants

Ammonio-nitrate of silver paper, 1839. The one improvement Talbot adopted wholesale. Ware credits the formulation to Alfred Swaine Taylor early in 1839: ammonium hydroxide is added to a silver nitrate solution of about 20 per cent w/v until the brown precipitate of silver oxide that forms at first just redissolves, giving a colourless diammine silver solution.

2 Ag⁺ + 2 OH⁻ → Ag₂O + H₂O

The precipitate that forms first

Ag₂O + 4 NH₄OH → 2 Ag(NH₃)₂⁺ + 2 OH⁻ + 3 H₂O

And redissolves as the diammine complex

Used on paper salted at 2 per cent, it gave a more sensitive sheet, a colder image colour and, Talbot recorded, pictures more resistant to fading. It became standard practice at the Reading Establishment for making positive prints.

☠Why the ammonio-nitrate variant is described here and not published as a formula

Ware’s cautionary note is unambiguous: solutions of diammine silver can, with time, deposit a black precipitate of silver nitride, also called fulminating silver — “a highly sensitive contact explosive, detonating at a touch, and sometimes even when wet”. Storing such solutions is not recommended. The silver nitrate page carries the same hazard from CAMEO’s datasheet, including a recorded case in which concentrated ammonium hydroxide and silver nitrate treated with sodium hydroxide gave a black precipitate that exploded on stirring. Reilly’s own warning is narrower and worth having beside it: the danger arises if an ammoniacal silver bath is “boiled down”, as some old manuals recommend to rejuvenate an overloaded bath, and he says flatly that it is not recommended for anyone to undertake that dangerous operation.

The course therefore describes the chemistry, records the historical importance, and gives no procedure. If an ammoniacal silver solution is ever made by accident — it is easy to make by adding ammonia to a silver bath — the ammoniacal silver quench SOP says what to do with it.

“Waterloo” paper, Talbot’s bromide version. Ware reads Notebook P as follows: the finished ordinary photogenic paper was washed with a 10 per cent potassium bromide solution, which converts the silver chloride substantially to the more insoluble silver bromide, and then washed with 20 per cent silver nitrate or with the ammonio-nitrate to restore the excess of silver ions. Multiple alternate coatings raised it further. Talbot seems to have regarded it chiefly as a route to camera negatives, and it was made obsolete by the calotype before it went anywhere. He fixed it with 10 per cent potassium bromide.

The calotype, 1840, which is not a variant of this paper but its replacement, and belongs to its own process page. Silver iodide in the fibres, sensitised with gallic acid and acidified silver nitrate, exposed only long enough to lay down an invisible latent image, then developed. Ware puts the speed gain at about a hundredfold — an hour becomes a minute — and that is the moment photography stops being a print-out craft.

The modern descendant, and the honest comparison. Reilly’s plain salting solution is 20 g of salt and 2 g of gelatin in a litre against a 120 g/L silver bath. Set that beside Talbot’s 11 g/L against 200 g/L and four things stand out: the modern chloride is roughly double, the modern silver is roughly half, the excess of silver over what the chloride needs has therefore fallen from about six times to about two, and the gelatin is added deliberately instead of being whatever the mill supplied. The last of those is the convenience; the third is the substantive change, and it is the difference between a paper that prints out hard and one that keeps for a day and behaves the same way twice. The Getty’s description of salt printing as practised today has moved further still, to a citrate-and-ammonium-chloride salting at around 4 per cent with a 12 per cent silver bath.

Hypo, from 1839 onwards. Not a variant of the paper at all, but the change that turned photogenic drawing into the salted paper print. Ware’s naming convention is worth adopting: prints made on Talbot’s photogenic drawing paper and fixed with thiosulphate are generally called salt prints. Same sheet, different fixer, different object.

Safety

Level B, and the level is about the silver bath and nothing else.

Silver nitrate at 18 to 23 per cent w/v is the governing hazard. Its harmonised classification under Regulation (EC) No 1272/2008 carries H272 (may intensify fire; oxidiser), H314 (causes severe skin burns and eye damage), H400 and H410 (very toxic to aquatic life, with long-lasting effects), signal word Danger. It stains skin, cloth and every surface it touches, and the stain appears hours after the splash and has to wear off with the skin. The chemical’s own page carries the full classification, the workplace exposure limits, the incompatibilities and the argyria warning that goes with repeated exposure. Sealed splash goggles rather than safety glasses, nitrile gloves, an apron, and eyewash within reach before the bottle is opened — the full regime is in the silver nitrate handling SOP and on the chemical’s own page.

A hazard peculiar to this formula: paper loaded with an oxidiser, drying at a fire. Silver nitrate is noncombustible itself but accelerates the burning of combustible material, and a sheet of paper carrying 200 g/L of it and held to a fire is the textbook case. Talbot dried at the fire because he had no other heat source. Nobody should copy that, and nothing in the course asks anyone to.

The mild parts. Sodium chloride is food. Potassium iodide is a pharmaceutical at these strengths. Neither bath is light-sensitive, so both can be handled in white light.

The exposure. Direct sunlight was the source, and the ordinary sunburn and eye hazards apply. A modern UV unit substitutes its own; see the UV unit operation SOP.

And the reason there is still no procedure here. It is not the silver nitrate — the course teaches that chemistry properly under salted paper at the same Level B. It is that the quantities are a modern reading of an unpublished notebook, and Rule 6 does not let a reconstruction be published as a historical formula. The reasoning is set out in full in the callout at the head of this page and in provenance.note.

Storage

The solutions, as far as anything is known. Talbot publishes nothing about keeping any of the four baths, and the course will not invent it. What is established elsewhere and applies: a silver nitrate solution is decomposed by light faster than the solid is, so amber glass, capped, cool and dark, in a container that has never held food and never will, labelled with strength and date per the labelling SOP. A salt solution keeps indefinitely. A potassium iodide solution does not behave as well as it looks. Its own encyclopaedia page carries Kodak’s 1928 finding that such solutions often turn deep yellow from liberated free iodine, and the consequence: a yellowed stock is no longer only potassium iodide but iodine dissolved in iodide, and that mixture attacks a silver image — which is precisely the wrong thing to pour over one. Dark brown bottles for both the solid and the solution.

The salted, unsensitised sheet. Nothing in it is light-sensitive yet, and by analogy with the modern process it should keep indefinitely if kept dry — but that is the course reasoning from Reilly’s material to Talbot’s, not a statement either man made about this paper, and it is marked as such.

The sensitised sheet. Talbot’s only relevant instruction is a handling one and it is emphatic: a sensitive sheet prepared by daylight “must by no means be left uncovered, but as soon as finished be shut up in a drawer or cupboard and there left to dry, or else dried at night by the warmth of a fire”. No keeping time is published for it, by anyone.

The finished photograph, which is the storage question that actually has evidence behind it. This is a conservation problem and the numbers are Ware’s.

A chloride-fixed photogenic drawing has a threshold exposure of the order of 200 lux hours, from three independent estimates that agree — about four hours at 50 lux to inflict one just-noticeable difference. Worse, Ware argues that the Becquerel effect makes such sheets vulnerable even under ultraviolet-free light. He will not predict a threshold for a sheet already “sunned”, saying only that it is likely to be shorter still, and shows that a lifetime of about three hours would require only some 1.5 per cent of the light between 450 and 650 nm to be actively absorbed. He adds that such prints may even be at risk under yellow or red photographic safelights. His conclusion is that the hope of displaying chloride-fixed photogenic drawings safely under filtered light is effectively destroyed, and that the only protection is to keep light levels low at all wavelengths and exposure durations strictly limited.

A thiosulphate-fixed sheet is in a different world: no significant change measured after 30 kilolux hours, and Ware suggests 100 klx hr is an acceptable risk. That is the single strongest practical argument for Herschel’s fixer over Talbot’s, and it is measured rather than asserted.

Incompatibilities

Silver nitrate and any soluble chloride, bromide or iodide, absolutely and by design. The two halves of this formula must never meet outside the paper. A splash of the salting solution into the silver bath throws silver chloride and destroys both; the same is true of tap water, which is why the modern sensitiser specifies distilled water and Reilly gives the reason. Separate vessels, separate trays, separate everything. See incompatibilities.

Silver nitrate and ammonia. Covered in the callout under Variants. The hazard is silver nitride, and it is a real one, not a historical curiosity.

Silver nitrate and combustible material. Paper, cloth and wood loaded with it and allowed to dry are a fire risk, because it is an oxidiser. Talbot’s fire-drying is the historical instance.

Silver nitrate and every reducing agent, which in a darkroom means every developer. A trace of metol or hydroquinone in a tray is enough to reduce silver out of solution.

Thiosulphate contamination running backwards. A trace of hypo carried on a wet hand into a sensitising bath does nothing visible at the time and produces a dead patch on a print later. In a printing-out workflow contamination travels upstream at least as often as downstream.

Iodide and the image itself, which is the odd one on this list because it is an incompatibility within the process. Excess iodide oxidises image silver back to silver iodide. Talbot’s fixer is incompatible with Talbot’s image, on a timescale of days to years.

Waste

Everything in this formula except the salting bath is a silver-bearing stream, and knowing which tray is which is the point.

The sensitiser, the first wash after exposure, the strong salt fixer and the iodide fixer all carry silver — the fixers particularly, because they convert the free silver rather than removing it and whatever runs off carries it. None of it goes to a drain. It is collected into one labelled container under the silver-bearing waste SOP. Silver’s classification includes H400 and H410, very toxic to aquatic life with long-lasting effects, which is the reason and not a formality.

The salting solution alone is dilute sodium chloride with a little paper fibre in it and carries no silver, no heavy metal and no thiosulphate.

Local regulation decides what may be discharged and this course cannot tell you what it says where you are. See disposal and the general chemical waste SOP.

Troubleshooting

The first five faults below are ones Talbot reported himself, in print, in 1839 or 1844; the last three are Ware’s, from the conservation side. Where a modern reading is given it is the course’s, drawn from Ware, and it is marked as an interpretation rather than as something Talbot said.

Large white spots of very definite outline, on a sheet that was supposed to be uniform, where the rest turns black as fast as possible. Talbot’s most puzzling defect and the one he devoted a whole closing section to: “the paper sometimes, although intended to be prepared of the most sensitive quality, turns out on trial to be wholly insensible to light”. Sometimes the spots were a pale cerulean blue with perfectly white outlines. His theory was unstable equilibrium between two definite chemical compounds. Modern reading: local excess of chloride. Where the salting was heavier or the silver thinner, the crystals ended up halide-jacketed rather than silver-jacketed, which is precisely Ware’s “fixed” silver halide, and a fixed halide does not print out. It is the same mechanism as Talbot’s own salt fixer, occurring by accident in patches.

The sheet darkens by itself, in the dark. Talbot: “this shows that the attempt to give it sensibility has been carried too far.” His remedy was to work as close to that edge as possible without crossing it, and to test rather than assume. Modern reading: the free silver is now high enough that the paper’s own oxidisable organic matter — cellulose, and whatever the mill sized it with — reduces some of it without any light at all. Ware’s account of why a silver nitrate paper is sensitive in the first place names those same substances as the oxidisable partner the nitrate cannot supply for itself; extending that to spontaneous darkening in the dark is this page’s reading rather than a sentence of his.

An uneven ground. Talbot’s own check is worth having: before using a sheet, hold it to the light briefly on purpose to give it a slight tint, and see whether the tint is even. If it is not, reject the sheet, because it will print with white patches where the preparation failed. That is a pre-flight test, published in 1839, and it costs one corner of a sheet.

The dark parts of the picture attacked during fixing. Talbot names the cause exactly: the iodide solution was too strong. “It is requisite, therefore, to find by trial the proper proportions.”

The whole picture fades to a uniform pale yellow over days. Iodide fixation, working as designed and then continuing. Not recoverable, and the reason the method was abandoned.

A lilac or mauve veil creeping over the highlights of a salt-fixed print. Residual silver chloride printing out slowly under display light. Ware’s estimates put one just-noticeable difference at roughly four hours at 50 lux, and the Becquerel effect means filtering the ultraviolet out does not save you. The only answer is dark storage and short, dim exhibition.

A brown stain over a hypo-fixed print. The free silver nitrate was not washed out before the thiosulphate went on, and thiosulphate reacted with silver nitrate to give silver sulphide. Ware lists it first among the three pitfalls of hypo on a print-out paper, and Herschel knew it in 1839.

A camera negative that is an outline with blank shadows. Not a fault. That is what this paper does in a camera, and Talbot published the fact in the Pencil of Nature rather than hiding it.

Experiments

None of these is performed with the formula above, because the formula above is not a procedure. Each is run with the course’s own salting solution and sensitiser, at Level B, and each tests a specific claim made on this page.

The equivalence experiment, which is the one that matters. Salt three sheets of the same stock at 2, 4 and 6 per cent sodium chloride, and sensitise all three on the same 12 per cent silver bath. The middle sheet is the interesting one: 120 g/L of silver nitrate is 0.71 mol/L, which would exactly consume 0.71 mol/L of chloride, and that is 41 g/L — a 4.1 per cent salting solution. So the 2 per cent sheet is silver-rich, the 4 per cent sheet is at about chemical equivalence, and the 6 per cent sheet is halide-rich. Print all three side by side under the same negative in the same light and watch rather than time them.

The prediction from Ware is not subtle. The 2 per cent sheet should keep building to a full brown-black; the 4 per cent sheet should go lilac and then stop, at a density you can barely measure; the 6 per cent sheet should be worse still. Two honest caveats. The crossover happens inside the paper and not in the trays, so where it falls depends on how much of each solution the stock carries away — the assumption flagged in the Mixing callout, and exactly why Talbot said the proportion has to be found by trial. And this is a Level B silver-nitrate exercise: read the silver nitrate handling SOP first. If it works, you will have reproduced in an afternoon, with two bottles, the single observation on which photography on paper rests.

Measure the ceiling. Same experiment, with a densitometer and a step tablet, printing each sheet to exhaustion — that is, until further exposure changes nothing. Record maximum density against salting strength. Ware’s figure for a pure halide with no acceptor is about 0.02; a well-made salted paper reaches well over 1.0. The curve between them is the halogen acceptor’s contribution and, so far as the course knows, nobody has published it for a modern salted paper.

Test the water half of the acceptor. Two sheets from the same batch, identically exposed: one dried hard in a warm cupboard and printed immediately, one equilibrated overnight in a humid room, or used frankly damp as Talbot used his. If Ware’s mechanism is right the moist sheet should print out faster and further. This is the cheapest test of a claim that is central to the whole page, and Talbot’s own 1835 result is the historical precedent for the answer.

Reproduce Talbot’s chloride fixer, and watch it fail. Fix one small print in saturated salt as Talbot describes, and its twin in a thiosulphate fixer. Put both in a north-facing window and photograph them weekly against a grey card. Ware’s threshold figures predict a visible lilac veil on the salt-fixed one within days of accumulated light and no measurable change on the other. Do not use a print you care about, and note that this experiment is also a demonstration of why museums keep these objects in the dark.

Talbot’s numbered strips. Prepare five sheets at salting strengths spanning the range, cut a strip from each, number the backs in pencil, and expose all five side by side in weak diffused light for a quarter of an hour. Rank them by eye, then check the ranking against a proper print from each. The question worth answering is whether a fifteen-minute ranking in weak diffused light predicts the ranking of finished prints. Talbot acted as though it did, and staked his camera work on it.

Read the ratio out of a print instead of into one. For anyone with access to X-ray fluorescence, the Getty atlas’s identification methods point the way: the silver-to-chlorine ratio in the highlights of a finished print is a measurement of how much unconverted halide the fixing left behind, and it distinguishes a halide-stabilised sheet from a thiosulphate-fixed one. That is a conservation question with a chemical answer, and it starts from the same arithmetic as everything else here.