Herschel's argentotype

The first process in which light acting on iron put a silver image into paper is four sentences long, sits in a postscript, and has no name in the document that contains it. Herschel had spent the summer of 1842 finding out what the light was doing to his ammonio-citrate of iron papers, had answered the question with gold, and then tried silver in the same place. It worked, more slowly and more strongly, and the Van Dyke, the kallitype and Ware’s argyrotype are all descended from those four sentences.

This formula's source publishes no quantities. The paper is "washed with a moderately concentrated solution of ammonio-citrate of iron, and dried. The strength of the solution should be such as to dry into a good yellow colour, not at all brown." The development is the whole of "be washed with nitrate of silver instead of a solution of gold", with no strength, no volume and no temperature; the fixing is "may be fixed by the hyposulphite of soda", with no strength and no time. Only the paper has ever been given a number, once, in Herschel's letter to The Athenaeum of 10 August 1842.

The composition is fixed by the effect — the preparation is carried until the source’s own end point appears. By an effect, and only the first of the three baths has one. The coating is carried until the dried sheet is a good yellow and not at all brown, which is a real and repeatable end point that anybody can hit by eye and which the salt's own variability makes necessary: ammonium iron(III) citrate is a family of complexes, not a compound, and its iron content runs from 14 to 28 per cent, so Herschel's one published make-up of 100 grains in 900 grains of water fixes the weight of salt and leaves the weight of iron uncertain by a factor of two. The silver bath is fixed by nothing whatever except the outcome it has to produce — a few moments with no apparent effect, the dark shades touched in first, the maximum of distinctness in two or three minutes — and the fixing bath by nothing at all. What is left undetermined is therefore the whole of the silver concentration, which in this family sets the image colour, the amount of silver left to be washed out and the risk that the nitrate dissolves the image it has just made. A reader can reproduce Herschel's paper and cannot reproduce his development.

The amounts below are the source's own words, not measurements, and the table is not a recipe. 8 documents were read for a published quantity and none states one; they are listed in this page's sources.

The ammonio-citrate of iron paper, as Herschel published its make-up — the only light-sensitive part of the process

Ingredient	As the source states it	Form the source specifies
Ammonium iron(III) citrate	6.48 g	Crystallised, and the brown salt: the green form did not exist until Valenta made it in 1897. Herschel's own figure is 100 grains, from his letter in The Athenaeum of 20 August 1842; Article 218 itself gives no strength and sends the reader to Article 212, which specifies the coating by its dried colour instead
Water	58.3 mL, added	Herschel's 900 grains of water, which is 58.3 g, entered as a volume at 1.00 g/mL. It is an addition and not a make-up volume, so no per cent w/v can honestly be stated for it; what can be stated is one part of the salt to nine of water by weight, which is 10 per cent by weight of the finished solution

The silver wash, given no strength anywhere — the bath that makes the image, applied after the exposure

Ingredient	As the source states it	Form the source specifies
Silver nitrate	“nitrate of silver, in no stated strength”	The whole of Herschel's instruction is "be washed with nitrate of silver instead of a solution of gold". Where the parallel gold bath of Article 212 at least carries a colour test — "of such strength as to have about the colour of sherry wine" — the silver bath carries nothing
Water	“implied by "washed with" and never measured”	Herschel washes the sheet with a solution and neither states its strength nor its volume nor its temperature. The dilution is the largest single unknown on this page, because in every descendant of this formula the silver concentration sets the image colour and the amount of free nitrate that later has to be washed out

The hyposulphite of soda, offered and not specified — the fixing bath

Ingredient	As the source states it	Form the source specifies
Sodium thiosulfate pentahydrate	“the hyposulphite of soda, in no stated strength”	"The picture may be fixed by the hyposulphite of soda, which alone, I believe, can be fully depended on for fixing argentine photographs." The permissive "may" is Herschel's, and so is the claim about hypo, which he had discovered the solvent action of in 1819
Water	“a bath, of unstated strength, volume, temperature and duration”	Nothing is published about this bath except the salt it is made of. Article 217, three paragraphs earlier, gives the chrysotype's rinsing regime in careful detail — spring water three times renewed, five or ten minutes in the third — which makes the omission here look like haste rather than reticence

Used in this order — Herschel's own order in Article 218

1. The ammonio-citrate of iron paper, as Herschel published its make-up — Washed over the paper with a soft brush and dried, then exposed in the camera or under an engraving until a latent picture is impressed, which is "very faint, and sometimes hardly perceptible"
2. The silver wash, given no strength anywhere — A few moments elapse without apparent effect; in two or three minutes the maximum of distinctness will not fail to be attained — The dark shades are touched in first and the details appear by degrees
3. The hyposulphite of soda, offered and not specified — Herschel gives no time, no strength and no washing instruction for this step

If paper prepared as above recommended for the chrysotype, either with the ammonio-citrate or ammonio-tartrate of iron, and impressed, as in that process, with a latent picture, be washed with nitrate of silver instead of a solution of gold, a very sharp and beautiful picture is developed, of great intensity ... The picture may be fixed by the hyposulphite of soda.

The order is the whole of what makes this a developed-out process rather than a printed-out one. The silver is not in the coating and is not present during the exposure; it meets the paper only after the light has finished acting on the iron.

!What the banner above means on this particular page

The standard heading speaks for Article 218, which is this formula’s source and which publishes no quantity for anything. Two of the three baths below are therefore Herschel’s own words in place of numbers.

The first bath is not. The 6.48 g is a real published quantity, converted from the 100 grains of crystallised ammonio-citrate of iron that Herschel gave in a letter to The Athenaeum on 10 August 1842, nineteen days before he wrote the Postscript that contains Article 218. The course found it by reading that letter, and it changes what this page can say: docs/FORMULARY-SCHEMA.md records this sensitiser among the amounts no source states, which is true of Article 218 and false of Herschel.

The declaration stays anyway, and not as a formality. The silver bath and the fixing bath have no strengths in any document anybody has produced, and even the paper’s number leaves the coating open: ammonium iron(III) citrate is a family of complexes rather than a compound, and its own encyclopaedia entry records an iron content running from 14 to 28 per cent by weight. Ten per cent by weight of a substance that variable is not a fixed weight of iron. Herschel’s dried-colour test is what really fixes the coating, which is why fixedBy is to-effect.

The Part I lesson owns Herschel’s summer of 1842 as history, and the process entry owns the iron-silver family. This page owns Article 218 as a formula: what is in it, what is not, and why the boundary between the two is not quite where this course’s own notes had put it.

Purpose

To develop a silver image out of a latent image made in an iron salt. That sentence is the invention. Before it, a silver photograph was made by exposing a silver salt; after it, the silver could be kept out of the paper until the exposure was over and then brought in to build the image where the light had already done its work on something else.

Herschel’s own purpose was narrower and is worth keeping in view, because it explains the shape of the source. He was not designing a printing process. Article 210 had shown that ammonio-citrate paper with no precipitant in it at all is “apparently little, but in reality highly sensitive”, and Article 212 used a gold solution to prove that the iron really had been reduced to the lower oxidation state. Silver was the second reagent tried on the same question. The process fell out of the answer.

Recommended uses

None as a procedure. This is a historical-study entry, and two of its three baths were never given a strength by anybody, so there is nothing here to mix.

Three uses it does serve.

• Reading Article 218. Anybody who meets “argentotype” in a history, a catalogue or a conservation report can see the four sentences it refers to and what they do and do not contain.
• Understanding the difference between developing out and printing out in this family. It is easy to get backwards, and Herschel’s original is the clean case: the silver is applied after the exposure, so there is a latent image and a development. The Van Dyke puts the silver into the coating and prints out; the kallitype puts the silver into the coating and develops out. All three are often described in the same words and they are three different arrangements.
• Seeing what a partly published formula looks like. One of the three baths on this page has a number and two do not, and the number itself does not settle the composition. That is a more common historical situation than either a full recipe or a total silence, and it is worth meeting once.

When another formula is preferable

Always, for anything a reader intends to print.

For the same chemistry with every quantity sourced: the Van Dyke Brown sensitiser, which is this formula with the silver moved into the coating and tartaric acid added, followed by the iron-silver clearing sequence and the alkaline thiosulfate fixer.

For the developed-out version: Nicol’s kallitype with the kallitype sensitiser and a borax developer, which swaps the citrate for ferric oxalate and gets a considerably more energetic iron salt.

For the version designed against this one’s known defects: Ware’s argyrotype, which replaces the nitrate anion for the reason given under Interactions below.

For the same paper in blue, with Herschel’s own proportions: Herschel’s 1842 cyanotype, whose quantities survive in his manuscript Memoranda; and for the modern working version, the classic cyanotype sensitiser.

For a plain silver print with no iron in it at all: the salted paper sensitiser and a plain hypo bath.

Mixing

Three baths, in Herschel’s order, and only the first of them can be mixed from what he published.

The paper. Dissolve the crystallised ammonio-citrate of iron in water — Herschel’s letter gives 100 grains in 900 grains, which is one part of the salt to nine of water by weight — and wash it over “any thin, smooth, even-textured paper” with a soft brush. Dry it. The test is the dried colour: a good yellow, and not at all brown.

The silver wash. Herschel says “nitrate of silver” and stops. He does not say how strong, how much, at what temperature, or whether the sheet is flooded, brushed or immersed. The parallel gold bath of Article 212 at least has a colour test attached to it — the strength “to have about the colour of sherry wine” — and the silver has nothing.

The fixing bath. “The hyposulphite of soda”, and nothing else.

!What fixes the composition, since two thirds of the numbers do not

The paper is fixed by an effect and it is a good one. Coat and dry until the sheet is a good yellow, not at all brown, and you have hit the specification. That end point exists because the salt demands it: ammonium iron(III) citrate is sold under one name and is a family of iron-citrate-ammonium complexes with an iron content anywhere between 14 and 28 per cent by weight. Herschel had the brown salt, which sits at the top of that range. Weighing 10 per cent of a substance that variable does not give you a fixed amount of iron, and looking at the dried sheet does.

The silver bath is fixed by nothing at all. The only bound on it in the source is the outcome: a few moments with no apparent effect, dark shades touched in first, maximum distinctness in two or three minutes. A bath much weaker than Herschel’s would fail that test by being slow; a bath much stronger would fail it by fogging or by leaving more free silver in the paper than the wash can remove. That is a range and not a value, and the range is wide.

What is left undetermined is therefore the silver concentration, and in every descendant of this process the silver concentration is what sets the image colour, the quantity of free nitrate to be cleared afterwards, and the risk described under Interactions that the bath dissolves the image it has just made.

The order matters and is Herschel’s. Iron first and alone; exposure; silver afterwards. Putting the silver in the coating gives a different process — a good one, and two of them are in this formulary — but not this one.

Behaviour

It is slow to disclose and it does not hurry. Herschel is unusually precise about the one thing he could watch: “Its disclosure is not instantaneous; a few moments elapse without apparent effect; the dark shades are then first touched in, and by degrees the details appear, but much more slowly than in the case of gold. In two or three minutes, however, the maximum of distinctness will not fail to be attained.”

It is stronger than the gold version and slower. He says so twice, and the second time is a footnote he attached to the Philosophical Magazine’s report of the meeting: “A solution of silver produces a like effect, and with greater intensity, but much more slowly. Consequently the name Chrysotype would seem less appropriate than Siderotype.” That footnote is where the word siderotype comes from, and the silver process is what prompted it.

Hunt did not like it. Robert Hunt repeated all Herschel’s published experiments and published the result in three separate books; his whole verdict on this one is that the picture “is brought out somewhat more slowly, and, as far as my own experience goes, with much less beauty, whether we consider colour or detail”. It is the only contemporary judgement on record from somebody who actually made one.

The latent image is faint and the process is not. Article 212 describes the impressed picture as “very faint, and sometimes hardly perceptible” before development, which is the whole point of a developed-out process and is why Herschel could work in “moderate daylight” without spoiling it.

Image characteristics

Herschel gives one adjective and no colour. “A very sharp and beautiful picture is developed, of great intensity.” Nothing in Article 218 says what colour it is.

The colour is brown, and modern work says why. Ware describes the argentotype image as a brown silver image consisting of nanoparticles, and elsewhere gives the colloidal dimensions of brown silver images generally as about 20 nm — far smaller than the wavelengths of visible light, about 500 nm — with the colour arising from a size-dependent absorption rather than from any added substance. That is the silver page’s argument, and it is the same reason a salt print is not black.

Permanence: poor, and known to be poor. Ware’s survey is blunt about it. The particles are small, their surface area is large, they sit in the paper with no binder over them, and they are attacked by sulphur compounds and by oxidising acids. Very few identifiable specimens survive from the early years and most of those are badly faded. Toning with gold changes that; nothing in Herschel’s four sentences does.

Sharpness: the two witnesses disagree. Herschel calls the picture “very sharp and beautiful”; Hunt, who compared it directly against the gold version, thought it worse “whether we consider colour or detail”. Nothing else about it is on record, and the course does not adjudicate between them. What can be said structurally is that there is no binder to scatter in, and the silver is laid down where the iron(II) was.

The mechanism

Three steps, and Herschel established the first two experimentally without being able to write either of them.

Light reduces iron(III) to iron(II). The sensitive body is the citrato-iron(III) complex, and the photochemistry is an internal electron transfer from the carboxylate ligand to the metal. Ware writes the general case for the oxalate and notes that the citrate is the same in principle and more complicated in detail, so the equation below is the oxalate form and is offered as the model rather than as the citrate’s own stoichiometry.

2 Fe³⁺ + C₂O₄²⁻ → 2 Fe²⁺ + 2 CO₂

Ware's dissected form of the light reaction in an iron(III) carboxylate, given here as the model for the citrate

Iron(II) reduces silver(I) to metal. One electron, so one iron for one silver — where platinum would need two.

Fe²⁺ + Ag⁺ → Fe³⁺ + Ag

The image: the whole of the argentotype in one line

Thiosulfate takes away what is left. Herschel had discovered in 1819 that the hyposulphites dissolve the chlorides of silver, and Article 218 is him reaching for his own twenty-three-year-old result.

Ag⁺ + 2 S₂O₃²⁻ → [Ag(S₂O₃)₂]³⁻

Unreduced silver leaving as a soluble complex

⚗Why silver and gold, and not platinum

Ware gives the redox potential of the citrato-iron(III)/iron(II) couple as +0.372 V, and the metals it has to reduce as silver at +0.80 V, gold at +1.00 V, platinum at +0.73 V and palladium at +0.62 V. On thermodynamics alone the citrate couple should reach all four, and Ware reports that under printing conditions it reduces gold and silver and does not reduce platinum(II) or palladium(II); he attributes the failures to the kinetics of the particular complexes rather than to the potentials.

So the numbers do not by themselves explain the division, and this page reports the observed outcome rather than offering the potentials as a complete account of it — which is the same position the process entry and the citrate page take.

What the division does explain is the history. With the salt he had, Herschel could print in gold, silver, mercury and Prussian blue, and he did, all in one summer. Platinum printing had to wait thirty-one years and a more energetic iron salt, ferric oxalate.

Function of every ingredient

Ammonium iron(III) citrate — Herschel’s “crystallized ammonio-citrate of iron”, and the only light-sensitive thing on the page. Its job is to record the exposure as a distribution of iron(II), and it does the entire photographic part of the process by itself: Article 210 is Herschel proving exactly that, by leaving the cyanide salt out of a cyanotype coating and finding that four or five seconds of sun produced a change “quite imperceptible to the eye” which a later wash could still develop.

Three things about the substance bear on the number in the table. It is not a compound, and its iron content varies from 14 to 28 per cent by weight, so the same weight of salt is not the same weight of iron. Herschel had the brown form, which Eder confirms and which carries 19 to 28 per cent iron on the citrate page’s own figures; the green form that most modern recipes call for was made by Valenta in 1897, and Eder records that it greatly reduced the printing time. And Herschel’s alternative was the ammonio-tartrate, which Article 218 offers in the same breath and for which no strength has ever been published by anybody.

More of it, or a stronger solution, dries browner — which is precisely the fault Article 212 names. Less of it dries pale and holds too little iron to build an image.

Silver nitrate — the image, and not the sensitiser. It supplies the silver(I) ions that iron(II) reduces, and it is not itself sensitive to light in this process because it is not present while the light is acting. That is the sentence to hold on to: a fault in an iron-silver print that looks like a speed problem is an iron problem.

Its strength is the great unpublished quantity of this page. What can be said about the consequences of getting it wrong is said by the descendants rather than by Herschel: too little silver and the image is thin and the development slow; too much and there is free nitrate left in the paper that must be washed out before any thiosulfate touches it, on pain of brown silver sulfide staining.

Sodium thiosulfate — “the hyposulphite of soda”, and Herschel’s own discovery coming back to him. Its job is to dissolve the silver that was never reduced, as a soluble bis-thiosulfato complex, so that it can leave the paper in the wash. Herschel had published the solvent action of the hyposulphites on silver chloride in 1819 and had confirmed its use as a photographic fixer in January 1839, so his “which alone, I believe, can be fully depended on for fixing argentine photographs” is a statement of long standing rather than a guess.

Its strength is unpublished here too, and Ware’s account of thiosulfate fixation of print-out silver names the three ways of getting it wrong: excess silver nitrate not washed out first, giving brown staining; over-long immersion or too concentrated a bath, oxidising the colloidal image away in air; and insufficient washing afterwards, leaving thiosulfate to sulphide the image slowly for the next century.

Water is not an encyclopaedia ingredient here but it is part of all three baths and it is measured in only one of them. Herschel weighs it for the paper — 900 grains — and says nothing at all about it for the silver or the hypo. Its amount in those two is the dilution, and the dilution is the strength.

Interactions

Nitrate against the image it has just made. This is the design fault Ware identifies in the whole family and the reason his argyrotype exists: nitrate is an oxidising anion, and it tends to dissolve colloidal image silver during wet processing, particularly under acid conditions. Herschel’s process puts the nitrate into the paper after the exposure, at an unstated strength, and then gives no washing instruction at all before the fixing bath.

Iron(III) against the image, afterwards. Any citrate iron left in the paper after processing goes on oxidising the silver, which is residual iron and is the family’s characteristic long-term failure. It is a clearing problem, not a fixing one, and the modern answer is the iron-silver clearing sequence with EDTA and citric acid in it. Herschel washes with nothing.

Thiosulfate against unwashed silver nitrate. Bring the two together in the paper and the result is brown silver sulfide in the highlights rather than clear paper. Ware records that Herschel recognised the importance of washing the free silver out first, which is the course’s reason for reading the omission from Article 218 as a gap in the text rather than in the practice.

Alkali against iron(III). Raising the pH hydrolyses the excess iron(III) and deposits ferric hydroxide in the image. It is why the kallitype’s alkaline developers are a compromise and why Ware’s redesign runs the whole process at about pH 3.5, using sulfamic acid as the acid you can weigh out.

Variants

Herschel’s own, all from the same coated sheet. Ware’s reading of the manuscript List of Prepared Papers makes the point better than any description: a single sheet of ammonio-citrate paper was cut into a dozen or sixteen pieces, exposed, and then treated variously to make a cyanotype, a chrysotype or an argentotype. The variants are not different formulas; they are different second baths.

• The gold version, the chrysotype of Article 212, which came first and which is why Article 218 begins “instead of a solution of gold”.
• The blue version, the cyanotype of Article 219, developed with potassium ferricyanide — and its mercury-modified relative.
• The mercury version, the kelainotype, developed with a mercury salt. Ware’s table of the siderotypes gives its iron salt as the tartrate rather than the citrate, and it failed for a reason none of the others had: the metallic mercury of the image is volatile and evaporates within a few days.
• The ammonio-tartrate paper, which Article 218 names alongside the citrate and which Article 210 says “fully possess” the same property. No strength for it exists.

The descendants, which are what the process became. Eder traces the line: the argentotype “came up again with small changes in England as the kallitype process” in 1889 and then in the sepia papers of 1895, and Valenta’s substitution of the green ammonium ferri-citrate for the brown in 1897 “greatly reduced” the printing time for argentotypes. Ware puts the same family in one sentence — Van Dyke Brown, kallitype, sepiaprint and brownprint are “no more than re-inventions of Herschel’s argentotype” — and adds his own argyrotype of 1991 as a deliberate redesign rather than a variation.

✕Two things called argentotype that are not this

Cassell’s Cyclopaedia of Photography of 1911 defines ARGENTOTYPE as “a name for bromide paper, and widely used in the early days of the bromide process”, and carries a BROMO-ARGENTOTYPE entry beside it. A gelatin silver bromide enlarging paper has nothing to do with iron.

Ware’s own Argyronomicon captions an 1839 print — the Advocate’s Library in Edinburgh, the earliest dated photographic print by Herschel — as “Herschel’s argentotype of 1839”, while his running text calls the same object “a silver negative print” and his tables date the argentotype process to 1842. The 1839 print is from Herschel’s silver work of that year and predates the iron-silver process on this page by three years.

And the name itself is not in Article 218. The Postscript gives the process no name at all; Herschel names the cyanotype one article later and had named the chrysotype in Article 213. Ware attributes “argentotype” to Herschel and lists it in his table of Herschel’s etymologies; Eder writes it in quotation marks. This page follows the usage without asserting where in Herschel’s hand it first appears.

Safety

Silver nitrate sets the level and it is Level B. The encyclopaedia entry carries the GHS classification, the exposure controls and the handling rules, and the rubric names the substance explicitly at that level. It is corrosive to eyes and skin, and it stains skin, clothing and bench permanently and blackly.

Ammonium iron(III) citrate is a Level A irritant that is deliquescent and stains a yellow-brown which turns blue on washing. Sodium thiosulfate is Level A. Nothing on this page is above Level B, and no procedure for making any of it appears here, because two of the three baths have no published strength to give.

Three specific points, since a reader may be tempted to reconstruct.

• Thiosulfate meeting an acid gives sulfur dioxide and a cloud of colloidal sulfur, and this formula prescribes no rinse whatever between the silver bath and the hypo. Anybody reconstructing it puts one there; the incompatibilities page carries the pair and the reason.
• The rinse water and the fixer are silver-bearing and belong to recovery, not to the drain.
• Choosing your own strength is choosing your own hazard. A concentrated silver bath is a different exposure from a dilute one, and this page cannot tell you which one Herschel used.

Storage

No keeping figures are published for any of the three baths, and none can honestly be derived, so this entry carries none.

What the sources do support is worth having in its place. The coated paper is dried and, in the Athenaeum letter, is simply “ready for use”; Herschel says nothing about how long it stays that way. The salt itself is highly deliquescent and will go sticky and then solid in a humid room, so its tub is opened briefly and closed hard; the citrate page carries that and the rest of its storage, including Ware’s observation that a stock solution of the salt is an excellent nutrient medium for moulds and grows a furry surface within a week or two in many rooms.

A silver nitrate solution keeps in the dark and darkens in the light, and this is the one place where the missing strength has a storage consequence: without a stated concentration there is no stated shelf life either, because the two are not independent.

Incompatibilities

Alkali, in any of the three baths. It hydrolyses iron(III) to ferric hydroxide and puts it in the image; see Interactions.

Chloride, in the water or the paper. The silver nitrate page names soluble chlorides, bromides and iodides, “including those in tap water”, as precipitating the silver halide immediately. Wherever they meet this silver wash they make silver chloride, which is a different and light-sensitive material — it is exactly what a salted paper print is made of, deliberately — and in an argentotype it is contamination.

Acid meeting the thiosulfate bath. Sulfur dioxide; see Safety.

Reducing agents generally, with the silver bath. The silver nitrate page lists them, and the sensitiser’s own photochemical iron(II) is one of them — which is the process working rather than failing, and is exactly why the silver arrives last and not sooner.

Combustible material, once the sheet is dry. Silver nitrate is an oxidising agent, and CAMEO’s warning, carried on the same page, is that paper, cloth and wood loaded with it are a fire risk as they dry. This is a process that loads paper with it deliberately.

Sulfur, afterwards. The finished image is nanoparticle silver with nothing over it, and Ware identifies sulphur-containing compounds and oxidising acids as its chief enemies. Storage boards, adhesives and enclosures are part of the process for this family in a way they are not for a gelatin print.

Waste

Everything downstream of the silver wash is silver-bearing and is recovered rather than poured away: the wash water, the spent fixer and any unused developing solution. The silver and silver nitrate entries carry the recovery routes, and the disposal page carries the rules.

The spent sensitiser is an iron(III) citrate solution and is not silver-bearing. CAMEO’s record, as the citrate page sets out, is that the primary hazard of ferric ammonium citrate is the threat to the environment, so none of it goes onto a garden, into a soakaway or into a watercourse, and local regulation governs the rest.

Troubleshooting

Every fault below is drawn from a source rather than from reconstruction, and the source is named, because nobody now works this formula as published.

The paper dries brown instead of yellow. The solution is too strong. This is Herschel’s own specification failing, in his own words — “not at all brown” — and it is the only coating fault he names.

The picture is very faint before development. Expected, and not a fault. Article 212 says the impressed image is “very faint, and sometimes hardly perceptible”; that is what a developed-out process looks like before the second bath.

The image comes up and then stops short. Herschel’s bound is two or three minutes to maximum distinctness, and beyond that he reports nothing. Hunt, who made one, judged the result “much less beauty, whether we consider colour or detail” than the gold version. That the ceiling is in the process rather than in the operator’s technique is the course’s reading of those two statements together, and neither author says it.

The highlights go brown in the fixer. Free silver nitrate was not washed out before the thiosulfate, and the brown is silver sulfide. Ware names this as the first of the three pitfalls of thiosulfate fixation for print-out silver, and records that Herschel recognised its importance even though Article 218 does not say so.

The image loses density in the fixer. Over-long immersion, too concentrated a bath, or oxidising impurities in the water, with the print exposed to air: Ware’s second pitfall, and the one that bites hardest on colloidal silver.

The print fades over months or years. Ware’s third pitfall is insufficient washing after fixation, and his separate diagnosis of the family is residual iron going on oxidising the silver. Both are post-processing failures rather than chemistry failures, which is the course’s general position on this family.

The paper yellows. Residual iron again. The confirming test and the fault are in the troubleshooting atlas.

Experiments

Everything here is reading and comparison. The course does not prepare this formula, and every comparison below can be run on a Van Dyke or an argyrotype at Level B with published numbers and a better print at the end.

• Read Article 218 with Article 212 open beside it. Four sentences, and every one of them depends on the earlier article. Then read the footnote to Article 214 and satisfy yourself that its forty grains of muriate of ammonia and sixty grains of nitrate of silver belong to Robert Hunt’s salted paper and not to this process. That is the trap in this document, and it has very nearly been walked into while this page was being written.
• Read Herschel’s letter in The Athenaeum of 20 August 1842 beside Article 212. The same author describes the same paper twice within a month, once by its dried colour and once by weight, and the two descriptions are for two different audiences. It is the clearest small lesson available in why a formula’s silence is sometimes a matter of where it was printed.
• Follow the one number through the salt’s variability. Ten per cent by weight of a salt whose iron content runs from 14 to 28 per cent gives an iron concentration between about 1.4 and 2.8 per cent of the solution. Herschel’s colour test collapses that range; the weight does not.
• Compare the three second baths on paper. Gold brings the picture up “instantly”, though Herschel’s letter puts full intensity at “about a minute or a minute and a half”; silver takes two or three minutes and comes out stronger; ferricyanide gives a cyanotype. Same sheet, same exposure, three chemistries — which is literally what Herschel did with his cut-up sheets.

🧪The one this course would set instead

Coat a sheet with the Van Dyke Brown sensitiser, print it out, clear it with the iron-silver clearing sequence and fix it in the alkaline thiosulfate bath. Then deliberately skip the clearing step on one strip and keep both in the light for a month.

You will have made Herschel’s picture by a route with every quantity published, and the uncleared strip will show you, in a month, the residual-iron failure that gave this whole family its reputation for impermanence.