Nicol's kallitype

Almost nothing survives of the kallitype’s first decade. Dick Stevens searched the major American collections and found one historic specimen; a handful more have been identified in Britain, and those show faded images and yellow highlights. So the interesting question about the formula below is not how to print with it — the modern sensitiser does that better — but why a process built on sound chemistry, by a professional chemist, with a name meaning beautiful, left so little behind. Everything needed to answer that is in the four lines of the table and the two paragraphs of instructions that follow them.

Ingredient	Quantity	Form the source specifies
Ferric oxalate	167.5 g	Cassell's own metric column, against 750 grains in 10 fluid ounces in the imperial one. No hydrate is named, and none could be: Cassell's own entry for the substance gives Fe2(C2O4)3 and a molecular weight of 376, which is the anhydrous formula unit, while the commercial article is of uncertain and variable water content
Silver nitrate	67 g	300 grains in 10 fluid ounces in the imperial column. Added to the filtered iron solution after it has cooled enough to handle, never before filtering
Water	to make 1000 mL	Distilled water, and a make-up volume rather than an added one: Cassell's line is "Distilled water — 10 oz. — 1,000 ccs." at the foot of the table, which is the nineteenth-century convention for a stated final volume. Distilled matters chemically and not only for tidiness: calcium in hard water precipitates calcium oxalate and drives the iron(III) complex towards hydrolysis, which is the reaction that eventually stains a siderotype yellow.

Purpose

To sensitise plain paper so that a contact print can be developed out in metallic silver, at a fraction of the cost of platinum. W. W. J. Nicol, an academic chemist at Mason College in Birmingham, said so himself in a letter to the British Journal of Photography in 1891, quoted by Mike Ware:

For years I had believed that it ought to be possible to discover a silver printing process, which in point of simplicity and artistic effect would rival the expensive platinotype, and be able to claim for its results a degree of permanence higher than that possessed by ordinary silver prints.

Two ambitions, then: the look of a platinum print, and better permanence than a salted or albumen print. The first was achieved. The second is what this page is really about.

The formula is one bottle that does two jobs. It carries the light-sensitive iron salt and the silver that will become the image in the same solution, so a single coating puts both into the paper and the exposure and the development happen in the same layer. That arrangement is what distinguishes it from the printing-out iron-silver papers — the Van Dyke and Herschel’s argentotype — where the image appears during exposure and there is nothing to develop.

Recommended uses

Understanding what a historical kallitype is made of. If you are looking at a print in a collection, or reading a period manual, or trying to work out why a hundred-year-old brown print has gone yellow in the highlights, this is the composition the answer starts from.

Reconstructing the process at its own strengths. No modern kit is quite this formula. The common contemporary practice — equal parts of a 20 per cent iron stock and a 10 per cent silver stock, mixed in the coating dish — is a quarter weaker in iron for every unit of silver and two thirds as concentrated overall, and everything downstream of the coating has changed as well. Photographers’ Formulary’s traditional kit is the closer relative, being a single bottle carrying both salts, at almost exactly Nicol’s silver strength; the comparison is worked out under Variants. A print made to Nicol’s proportions and Nicol’s workflow is a different object, and if that difference is what you want to see, this is the mixture.

Teaching the whole iron-silver family from one page. Change the iron salt from the oxalate to the citrate and you have the Van Dyke. Change the noble metal from silver to platinum and you have the platinotype, which shares the sensitiser’s iron chemistry exactly. Take the silver out of the paper and apply it after the exposure instead, and you have what Robert Hunt described in 1844 and what Sandy King says Nicol’s original patent claimed.

Not for a print you intend to keep — not, at least, without the clearing and toning the period instructions do not contain. That is not a judgement on Nicol; it is the documented outcome, and the reasoning is under Behaviour and The mechanism.

When another formula is preferable

• For actual printing today, the kallitype sensitiser: two bottles that keep separately, a coating mixture whose proportion can be varied at the moment of use, and suppliers who sell both stocks ready-made. It is the same chemistry at four fifths of the iron per unit of silver, and it never asks you to store a silver solution that also contains oxalate.
• For a print that must survive, any of the noble-metal routes. King’s position is blunt and widely shared among current practitioners: tone a kallitype in gold, platinum or palladium, or accept that you have made something ephemeral. The platinum toner and the gold thiocyanate toner are the course’s entries for that step.
• For a silver process designed around this formula’s central flaw, Mike Ware’s argyrotype. Ware’s reasoning is that nitrate is an oxidising anion which dissolves the colloidal image silver during wet processing, so he replaced it with silver sulphamate and got a sensitiser that works at pH 2 to 3 and washes out cleanly. Nicol’s answer to the same problem was an alkaline developer, and Ware explains under The mechanism why that answer created a second problem.
• For the cheapest and simplest brown print on plain paper, the Van Dyke brown sensitiser, which uses ammonium iron(III) citrate and needs no developer at all. It is slower and it has less shadow depth; Eder’s table puts the citrate at 15 against the oxalate’s 89.
• For an iron process with no silver in it whatsoever, the classic cyanotype sensitiser. Same photochemistry in the first step, no noble metal, and a wash that fixes the print.

Mixing

Cassell’s procedure is four sentences long and every one of them is doing work.

1. Put the ferric oxalate in a bottle with the water and warm it in a saucepan of water until it dissolves. A water bath, not a flame: the salt is the light-sensitive ingredient and heat is one of the things that reduces it. Wall gives a temperature for the same step — “about 110 °F”, which is 43 °C — and modern suppliers put a ceiling near 50 °C on warming ferric oxalate solutions for exactly this reason.
2. If it will not dissolve, add oxalic acid — “not more than from 5 to 10 grains” — to assist solution. This is the fourth line of the table, the one whose quantity Cassell declines to give.
3. Filter the hot solution. Ferric oxalate as sold is not fully soluble and the undissolved fraction would coat as grit.
4. Add the silver nitrate to the clear filtrate. After filtering, never before: filter paper is cellulose, silver nitrate is an oxidiser, and a silver solution poured through a filter leaves silver in the filter.

Then size the paper and coat it. Wall gives the sizing that Cassell only gestures at: Bermuda arrowroot 180 grains to 20 ounces of water — about 2 per cent w/v — made into a thin cream, poured into the rest of the water boiling, boiled another eight or ten minutes, cooled, and brushed or sponged into paper pinned to a drawing board. Fresh paste each time, because it will not keep. The Getty’s account of the platinotype names starch or gelatin sizing for the same purpose, and the reason is the same in both processes: the size holds the sensitiser near the surface instead of letting it sink into the fibre, where the light cannot reach it and the wash cannot get it out again.

🧮Reading Cassell's two columns, and why they do not quite agree

The table prints every quantity twice. Read as a strength, the imperial column says 750 grains of ferric oxalate and 300 grains of silver nitrate in 10 fluid ounces of water; the metric column says 167.5 g and 67 g in 1,000 ccs. Those are supposed to be the same formula. Check it:

750 gr × 0.0648 g/gr ÷ 0.28413 L = 171.0 g/L

Ferric oxalate, imperial column, modern constants

300 gr × 0.0648 g/gr ÷ 0.28413 L = 68.4 g/L

Silver nitrate, imperial column, modern constants

Cassell prints 167.5 and 67.0, about 2 per cent lower. The gap is not a different formula and it is not a misprint. Cassell converted with a fluid ounce of 29 ccs, which was the usual working rounding of the period:

750 gr × 0.0648 g/gr ÷ 0.290 L = 167.6 g/L, and 300 gr the same way gives 67.0 g/L

The conversion Cassell actually used

The same basis reproduces the whole entry. The fixing bath’s 1 ounce of hypo in 20 fluid ounces comes out at 55 g/L on an ounce of about 31.9 g and a fluid ounce of 29 ccs, which is what Cassell prints; so does the borax developer’s 110 g/L. The metric column is therefore a genuine conversion of the imperial one, made consistently, and its 2 per cent error is far smaller than the uncertainty in what a gram of commercial ferric oxalate actually contains. Use either column. Do not mix them.

🧮The oxalic acid: a quantity two sources give differently

Cassell prints q.s. in both columns and then bounds it in the prose: not more than 5 to 10 grains for the ten-ounce batch, added only if the iron will not go into solution. Scaled to a litre on Cassell’s own basis that is 1.1 to 2.3 g/L.

Wall, printing the same two solids at a tenth of the scale, gives oxalic acid 5 grains to the fluid ounce — a fixed ingredient, dissolved with the iron rather than added if needed:

5 gr × 0.0648 g/gr ÷ 0.028413 L = 11.4 g/L, or about 1.14 per cent w/v

Wall's oxalic acid as a strength

That is roughly ten times Cassell’s ceiling, and the temptation is to call one of them a misprint. The course’s reading is that they are measuring different things, and Cassell’s own entry for ferric oxalate is the evidence: the standard platinotype “normal iron solution” of the period was 20 per cent ferric oxalate carrying about 1.2 per cent oxalic acid, that excess being part of how the solution is made in the first place. Wall’s 1.14 per cent is that same built-in excess written into the formula. Cassell’s 5 to 10 grains is a small extra pinch on top of it, for a batch of iron salt that will not co-operate. This is an interpretation, not a statement either book makes, and it is marked as one. What both books agree on is that some free oxalate is present, which matters chemically for a reason set out under The mechanism.

🧮What one sheet of paper carries, and how much of it stays

Neither source publishes a coating volume, so this arithmetic borrows one from the modern process and says so. Bostick & Sullivan give 20 drops of each stock per 8 by 10 inch print, which is about 2 mL of mixed sensitiser. Coat the same 2 mL of Nicol’s solution instead:

2 mL × 167.5 g/L = 335 mg

Ferric oxalate on an 8 by 10 sheet

2 mL × 67 g/L = 134 mg, of which the silver is 134 × 107.87 ÷ 169.87 = 85 mg

Silver nitrate on the same sheet

Now the other end. Ware puts the image coating weight of a plain-paper silver print at about 0.1 g/m², from chemical analyses of surviving salt prints and from an independent calculation using the Nutting equation. An 8 by 10 inch sheet is 0.0516 m²:

0.0516 m² × 0.1 g/m² = 5.2 mg

Image silver actually in the finished print

About six per cent of the silver you coat ends up in the picture, on these figures, and the rest goes down the drain in the developer, the fixer and the wash. Ware’s own figure for salt prints is 2 per cent. Either way the point stands: the finished image is a few milligrams of metal, in particles about 10 nm in radius, sitting in bare paper fibres with no binder over them. That is why a trace of residual iron or thiosulphate is enough to destroy it, and it is the single most useful number to carry away from this page.

Behaviour

It prints out, faintly, and then develops out. Under the negative the sheet goes, in Cassell’s words, to “a bluish brown colour upon a yellow ground”; Wall calls it “a faint brown image on a yellow ground” and says to stop when detail is just visible in the densest parts. The Getty’s account of the platinotype identifies what that faint image is in an iron process — photochemically generated ferrous oxalate — and here there is also some printed-out silver alongside it. Neither is the picture. The picture arrives in the developer.

The developer chooses the colour. This is the property that made the process popular with amateurs, and it is why the course has three separate developer pages rather than one. Cassell prints four baths for the same coated sheet:

Cassell’s developer	Tone	Composition, per litre
(1)	Black	Borax 110 g, Rochelle salt 82.5 g, 1 per cent potassium bichromate 87–112 ccs
(2)	Sepia	Rochelle salt 55 g, 1 per cent potassium bichromate 50–62 ccs
(3)	Purple	Borax 27.5 g, Rochelle salt 110 g, 1 per cent potassium bichromate 87–112 ccs
(4)	Maroon	Rochelle salt 110 g, sodium tungstate 55 g

Wall’s four are nearly the same set, with sodium tungstate alone for maroon in place of Cassell’s tungstate-and-tartrate. Read the table as chemistry rather than as a list and one thing jumps out: the tone tracks the borax, not the tartrate. Sepia has no borax at all; black has more borax than tartrate; purple keeps a quarter of black’s borax while raising the tartrate by a third, and lands between the two. Maroon has no borax and no dichromate and a tungstate instead. The borax and Rochelle salt developer pages work that argument through, and the sodium citrate developer is the modern single-bath replacement for all four.

Three of the four carry a chromium(VI) salt, and this course uses none of them. See Safety. The fourth is the awkward one: sodium tungstate has no entry in this chemical encyclopaedia, so the maroon developer cannot be written up as a formula anywhere in this course even though it is the one bath in the set with no dichromate in it. That is a gap in the encyclopaedia and it is recorded here as one rather than filled by guesswork.

Development is long, and it is long for the iron rather than for the image. Fifteen to thirty minutes in Cassell, ten to fifteen in Wall, and both books say why in almost the same sentence: the print must stay in the bath until the iron salts are completely dissolved. Wall goes further and names the residue as the enemy — “the presence of the iron salts is a most prolific cause of failure” — which is a remarkably accurate diagnosis for 1912, and the whole modern clearing sequence is the answer to it.

There is no clearing bath. This is the largest single difference between the period process and the modern one, and it is the reason to read the two workflows side by side:

Step	Nicol’s process, as Cassell prints it	Contemporary practice
Coat	One bottle, iron and silver together	Usually two bottles mixed in the dish; the Formulary’s traditional kit is still one
Expose	Daylight, to a visible bluish brown	UV box or sun, with almost no visible print-out
Develop	15–30 min, tartrate or borate bath	5–10 min, citrate or tartrate bath
Rinse	none	1–2 min in running water, and King insists it be neutral or slightly acid
Clear	none	citric acid or EDTA; the most thorough published sequence is EDTA, sulphite, EDTA
Tone	an optional after-treatment	treated by current practitioners as necessary
Fix	hypo with ammonia, 10 min	dilute plain hypo, 2–5 min
Wash	20–30 min	15–20 min

King’s reason for the rinse is worth having in full, because it is the step Nicol’s process skips outright: the first rinse must be neutral or slightly acidic, because “if the first rinse is alkaline, ferrous hydroxide compounds may be formed in the paper, making complete clearing difficult or impossible”. Nicol’s print goes from an alkaline borax developer straight into an ammoniacal fixing bath, which is the opposite of that instruction at both ends.

And the print goes into the fixer without a wash. Cassell is explicit: “The print is then passed direct, without washing, into the fixing bath.” Ware names precisely this as the first of three pitfalls of thiosulphate fixation — excess silver nitrate must be washed out first, or the image is seriously stained with brown silver sulphide. In fairness to Cassell, a print that has spent twenty minutes rocking in a litre of developer has had a long wash already; but nothing in the instructions guarantees it, and a short development followed by a direct transfer is a recipe for exactly the stain Ware describes.

Prints bleach in the fixer. Bostick & Sullivan warn about it in the modern process — “kallitype prints will bleach in the fix, so either tone before fixing or overprint by a stop or two. These are the only effective solutions to the bleaching problem” — and Ware gives the chemistry: colloidal silver is easily oxidised by air in the presence of thiosulphate, so too long a fix, too strong a fix or an oxidising impurity in the water costs density. Nicol’s ten minutes in a 5.5 per cent hypo bath is longer and stronger than the two minutes in 5 per cent that the modern kit specifies.

🕰Why Nicol's own fixer was ammonia, and what changed

Cassell records something the modern literature has lost: Nicol “at one time placed the paper ready for use on the market but withdrew it because he was not satisfied with the permanency of the results; his fixer was a weak solution of ammonia, but later experiments proved that the use of ‘hypo’ as a fixer gave more permanent prints.”

An ammonia fixer works by complexing residual silver(I) as the diamminesilver ion rather than by dissolving a halide, and it introduces no sulphur at all. Ware notes the visual consequence from the other end of photography’s history: a few experimental salt prints by John Adamson fixed only with ammonia contain no sulphur and show the yellow-orange colour of pure nanoparticle silver, whereas a hypo bath lays a thin skin of silver sulphide over each particle and deepens the colour to a more satisfying brown. So the change of fixer changed the picture as well as its permanence — and Ware’s caution completes the thought: a little sulphiding enriches a colloidal silver image, complete sulphiding destroys it, because converting the whole of it to silver sulphide drops the optical density by a factor of about thirty.

Image characteristics

Colour: brown to black, chosen after the exposure. Four developers, four tones, one coated sheet. Nothing else in this formulary offers that.

Maximum density: the best of the iron-silver processes. King’s comparison is that a well-made kallitype shows more richness in the shadows than either a Van Dyke or an argyrotype, that the difference is not huge but is visible in side-by-side prints, and that ferric oxalate also permits a wider negative density range. Eder’s numbers, reprinted by Cassell, say why: on his scale of relative light sensitiveness the oxalate scores 89 and ammonium ferric citrate 15.

Surface: none. There is no binder. The image is colloidal silver lying in and on the paper fibres, which is why the sizing changes the result and why every atmospheric pollutant reaches the metal directly.

Resemblance to platinum: close enough to be the whole commercial point. King states that a well-made kallitype toned in platinum or palladium is, for practical purposes, indistinguishable from a true platinum print. Ware is more careful about the historical claim and says the process could claim all the characteristics of platinotype except the two most important — simplicity of operation and permanence.

Permanence: the documented weakness, and the reason this page exists. Paul Anderson wrote in 1913 that “the image is so unstable that the process should be used for only the most ephemeral work”, adding that the statement would provoke violent protest from enthusiasts and was true nevertheless. Roger Child Bayley, whom Anderson quotes inaccurately and Ware quotes correctly, described a batch of kallitypes turned out of a drawer that “bore no sign to distinguish the front of the paper from the back. The image which once had been vigorous enough, had folded its tents like the Arab and had silently stolen away.”

⁇Was the historical kallitype really that bad?

The condemnations are contemporary and severe. Anderson in 1913 and Child Bayley in the same period both write the process off, and their verdicts have been repeated ever since.

Dick Stevens disputes the evidence rather than the conclusion. Ware reports Stevens’s position: Child Bayley’s experience of kallitype may have been limited and his practice uncontrolled, and Anderson reported no laboratory tests of permanence at all, so his condemnation needs substantiation.

A second dispute concerns the missing prints. Stevens surveyed many major American collections and found one historic kallitype, and concluded that the collectable photographers preferred platinum. Some promoters of the process argue instead that historic kallitypes are plentiful in collections but misidentified as platinotypes. Ware says no substantive evidence has been offered for that, traces the idea to letters in Photographic News around 1901 to 1906 complaining that a few commercial photographers sold matte silver bromide prints as platinum — not kallitypes — and calls the wider claim a canard.

What the physical evidence says. Ware reports that the very few hundred-year-old specimens now identified do show pronounced fading and seriously yellowed highlights, consistent with residual iron(III), and that the probable kallitypes he has found in early twentieth-century amateur archives are mostly very deteriorated, with severe iron stains, fogging and fading.

What this course concludes. The mechanism of the deterioration is well understood and is set out in the next section; the formula and the workflow on this page contain no step that would prevent it; and current practitioners, including King, treat noble-metal toning as necessary rather than optional. The course therefore treats the poor permanence of the historical process as established and the severity of the individual verdicts as contested.

The mechanism

⚗Light makes iron(II); the developer makes it useful; the iron(II) makes the silver

Step one, and the only step that needs light. Ultraviolet and blue light absorbed by the iron(III) oxalato complex drives an internal electron transfer from a coordinated oxalate to the iron. Ware writes the overall change for ferric oxalate in solution as:

Fe₂(C₂O₄)₃ → 2 FeC₂O₄ + 2 CO₂

Döbereiner's photolysis, 1831 — the reaction the whole siderotype family runs on

Carbon dioxide is lost and iron(II) oxalate is left behind. Abney’s finding, as Cassell reports it, is that the iron salts respond chiefly in the indigo blue and well into the ultraviolet — which is why this is a sunlight or UV-box process and why an ordinary tungsten lamp does nothing to it.

Step two: the photoproduct is insoluble, which is a problem. Iron(II) oxalate dissolves to the extent of 0.022 g per 100 cc of water. A solid sitting in a paper fibre cannot easily reduce anything. Free oxalate ion solubilises it as a complex:

FeC₂O₄ + C₂O₄²⁻ → Fe(C₂O₄)₂²⁻

Why a developer is needed at all

In platinum printing that job is done by a strong potassium oxalate bath. In the kallitype it is done by tartrate, borate or citrate, which are weaker but sufficient, and this is the reason a kallitype developer contains nothing that could reduce silver on its own.

Step three: the iron(II) reduces the silver. The redox potentials Ware gives make the direction inevitable:

E(Fe(C2O4)3³⁻ / Fe(C2O4)2²⁻) = +0.02 V

The reducing couple

E°(Ag⁺ / Ag) = +0.80 V

The couple being reduced

Fe(C₂O₄)₂²⁻ + Ag⁺ → Fe(C₂O₄)₂⁻ + Ag

The image forms: one electron, one atom of silver

Nearly eight tenths of a volt of driving force, which is why the image appears the moment the print enters the developer. For comparison Ware gives the citrato couple at +0.372 V: still enough to reduce silver(I) and gold(III), not enough for platinum(II) or palladium(II), which is exactly why the Van Dyke works and a citrate platinotype does not.

What Cassell gets wrong. The entry says the exposed paper holds “an image in ferrous oxalate and silver oxide”. Ware’s glossary defines the process without any such intermediate — “the photoproduct from exposure of ferric oxalate reduces silver nitrate to give a silver image” — and Wall states the same in 1912 and again in 1924. The course follows Ware and Wall, and the reason it does is worth stating as reasoning rather than as authority: silver oxide precipitates from a silver(I) solution when it is made alkaline, and this sensitiser is acidic, carrying free oxalic acid by design. What may well be present instead is a little silver oxalate, formed from that free oxalate and the silver in the coating — see Incompatibilities, because that compound has a hazard attached.

Deeper: the two design faults, and how each was eventually fixed

The chemistry above works. The permanence problem comes from two decisions that had nothing to do with the image-forming reaction, and Ware sets out both.

Fault one: the anion of the silver salt. Every historical iron-silver process uses silver nitrate, because it was the only cheap soluble silver salt. But nitrate is an oxidising anion, and Ware’s finding is that it tends to dissolve the colloidal image silver during wet processing, especially under acidic conditions. The image you have just made is being attacked by the counter-ion of the salt that made it.

Nicol’s answer was an alkaline developer, and it works: borax buffers the bath well above neutral, the acidic condition goes away, and the loss of image silver is minimised. Ware names borax by name as the example. It is at least suggestive that the bath carrying the most borax is the one Cassell and Wall both print for pure black, and the least borax the one they print for sepia — but no source read for this page says that the pH is what sets the colour, and the borax developer page takes that question up properly. The observation is recorded here as an observation.

Fault two: what an alkaline bath does to the excess iron. Ware’s next sentence is the sting: those high-pH developers “cause hydrolysis of the excess iron(III) in the sensitizer and the deposition of insoluble ferric hydroxide in the image, which ultimately causes it to fade.” Above about pH 4, iron(III) hydrolyses to polymeric colloidal iron(III) hydroxide. Freshly formed, it can be redissolved in dilute acid. Left in the paper and allowed to dry, it transforms irreversibly to iron(III) oxyhydroxide — the mineral goethite — which dilute acids will not touch:

[Fe(H₂O)₆]³⁺ → … → Fe(OH)₃ ↓ → FeO(OH) ↓

Iron(III) hydrolysis, as Ware writes it — a scheme, not an equation

Ware prints that chain with “etc.” between the arrows and no coefficients, because the intermediates are a series of polymeric colloidal species rather than a single compound. This course does not print it as a balanced equation, for the same reason it declines to balance the citrate photochemistry on the cyanotype pages: completing the stoichiometry would mean inventing species the source deliberately leaves unnamed. What is established is the sequence and its irreversibility.

Some of the iron(III) is also chemisorbed directly onto the hydroxyl groups of the cellulose, and any calcium — from hard water, or from the chalk buffer in the paper itself — precipitates calcium oxalate and pushes the iron complex further along the same road:

Ca²⁺ + [Fe(C₂O₄)₃]³⁻ + 2 H₂O → CaC₂O₄ + [Fe(C₂O₄)₂(H₂O)₂]⁻

Why the water has to be distilled

That is the whole explanation of the yellowed highlights. Residual iron in the paper is a yellow stain in its own right, and the shadows fade at the same time because the image silver is being oxidised. The condition of the surviving specimens Ware describes is precisely what this chemistry predicts.

Two authorities name different culprits, and the difference is worth keeping. Ware’s account throughout is about residual iron(III): chemisorbed to cellulose, hydrolysed above pH 4, ending as goethite, and yellow. King’s account of permanence names residual ferrous iron — “even very small quantities of residual ferrous iron will eventually oxidize the silver, and the image will fade” — and prescribes direct toning as the answer. The redox argument favours Ware’s reading, because it is iron(III) that can oxidise silver and iron(II) that reduces it; iron(II) left in a print is also oxidised to iron(III) by air in time, which reconciles the two if one reads King’s sentence as naming the reservoir rather than the oxidant. The course reports both attributions and does not merge them. What they agree on is the practical instruction, which is the part that matters at the sink: get every trace of iron out of the paper before it dries, and tone.

The fix arrived seventy years late. The modern three-bath clearing sequence — disodium EDTA at pH 3 to 4 to chelate the iron(III) before it can hydrolyse, then sodium sulphite to reduce whatever is still bound to the cellulose from iron(III) to the more easily removed iron(II), then tetrasodium EDTA at pH 9 to 10 to take out the last of it — was tested by Clarke and Hemmenway using X-ray fluorescence on platinum and palladium prints and found to leave less iron in the print than the uncoated paper started with. Nicol’s process has no equivalent step at any point. The iron-silver clearing sequence is the course’s entry for it, and dropping it into this workflow between development and fixing is the single change that turns a period kallitype into one that might last.

Deeper: how much iron, how much silver, and why the ratio was chosen

🧮Nicol's proportions against the modern ones

The two solids are in a fixed mass ratio in both formulas, and the ratio is the interesting number because it survives every uncertainty about hydrates:

167.5 g ferric oxalate ÷ 67 g silver nitrate = 2.50 : 1 by mass

Nicol, per Cassell

100 g/L ÷ 50 g/L = 2.00 : 1 by mass

Equal parts of 20 % iron and 10 % silver stocks

6 g ferric oxalate ÷ 2 g silver nitrate = 3.00 : 1 by mass

Photographers' Formulary, traditional one-bottle kit

Nicol’s sensitiser carries 25 per cent more iron for every unit of silver than the equal-parts mixture, and it is 1.56 times as concentrated overall — 23.5 per cent total solids against 15 per cent. The Formulary’s surviving one-bottle kit goes further still at 3 to 1. All three figures follow from the published strengths and none of them depends on knowing the water of crystallisation, because the same molar mass divides both sides of every ratio.

The silver strengths are the striking part. The Formulary’s sheet gives 30 mL of 20 per cent ferric oxalate carrying 2 g of solid silver nitrate and states no final volume, so the silver works out at about 65 to 67 g/L depending on how much the dissolved solid adds to the volume — which is Cassell’s 67 g/L to within the width of the uncertainty. A modern supplier’s traditional kallitype sensitiser and the strength printed in 1911 as Nicol’s are, on the silver, the same solution.

Going to moles needs an assumption, and the course flags it rather than hiding it: Cassell’s own entry gives ferric oxalate as Fe₂(C₂O₄)₃ with a molecular weight of 376, which is the anhydrous formula unit, while the commercial article is of uncertain and variable hydration. On the anhydrous figure:

167.5 g/L ÷ 375.75 g/mol × 2 Fe = 0.89 mol/L of Fe(III)

Iron(III) in Nicol's sensitiser

67 g/L ÷ 169.87 g/mol = 0.39 mol/L of Ag⁺

Silver in the same litre

About 2.3 iron atoms for every silver atom, where the reduction needs one electron and therefore one iron(II) per silver. Take a hexahydrate instead and the figure falls to about 1.8, and the modern mixture’s equivalent falls in the same proportion — the comparison is stable even though the absolute number is not.

Why so much excess iron? Because only a small fraction of the iron(III) on a sheet is ever photoreduced: exposure is stopped when a faint image appears, not when the iron is exhausted. The excess is what buys shadow density. It is also, exactly and unavoidably, the residual iron(III) that the clearing bath was later invented to remove — so the same decision that gives a kallitype its depth is the decision that makes it fade. That is not a flaw in Nicol’s arithmetic; it is the central trade of the whole siderotype family, and platinum printing lives with it too.

Function of every ingredient

Ferric oxalate — 167.5 g/L, the light-sensitive half. Cassell’s own entry describes it as greenish glistening scales, extremely sensitive to light, “the most light-sensitive of any of the iron salts”, and gives Fe₂(C₂O₄)₃ with a molecular weight of 376. Its job here is to absorb ultraviolet and blue light and hand an electron from a coordinated oxalate to the iron, giving iron(II) oxalate and carbon dioxide. Everything else in the process is a consequence of that one reaction.

More of it raises the amount of iron(II) available at any exposure and so raises maximum density, which is the obvious reason to pitch a sensitiser a quarter richer in iron than the equal-parts mixture — though no source read for this page records Nicol’s own reasoning; it also raises the residual iron(III) that has to be got out of the paper afterwards, which is the failure mode. Less of it costs shadow density and, past a point, leaves silver in the paper with nothing to reduce it — which fixes out and is simply wasted. Its interactions are with everything: with the silver, whose reduction it drives; with the free oxalate, which keeps it in solution before exposure and mobilises its photoproduct after; with the developer, which must complex iron(II) without reducing silver on its own; and with the paper, whose cellulose hydroxyls chemisorb the iron(III) it leaves behind.

A caution the encyclopaedia entry carries and this page repeats. Ware calls ferric oxalate “a chemists’ nightmare — ill-characterised, evidently polymorphic, apparently uncrystallisable”, and sources differ on its water content. Two bottles labelled the same thing may not carry the same weight of iron per gram. That uncertainty sits underneath every molar calculation on this page.

Silver nitrate — 67 g/L, the image. It is not light-sensitive in this formula; Ware states flatly that silver nitrate is not light-sensitive per se. It waits, as Ag⁺ in the paper, until the developer mobilises the iron(II) that light made, and is then reduced to metallic silver at exactly those points. Every visible tone in the finished print is that metal.

More of it gives more image silver only while there is iron(II) to reduce it, and beyond that adds nothing but fixer load, cost and stain risk. Less of it thins the image and, at the extreme, makes a print that the fixer can bleach away entirely. Its interaction that matters most is not with the iron at all but with its own anion: nitrate is an oxidising anion which, on Ware’s account, dissolves colloidal image silver during wet processing under acidic conditions. That single property drove the choice of alkaline developer, which in turn drove the iron-hydrolysis problem, which is what killed the prints. The whole causal chain of this process’s reputation starts at the nitrate ion.

Oxalic acid — the line the table cannot price. Cassell gives it as q.s., to be added “if there is need” and never more than 5 to 10 grains to the ten-ounce batch; Wall makes it a fixed 5 grains to the ounce, about 1.14 per cent w/v. It has three jobs and it is worth separating them.

• It gets the ferric oxalate into solution and keeps it there. A slight excess of oxalate suppresses the hydrolysis that would otherwise throw iron(III) hydroxide out of an aqueous iron(III) solution. Cassell’s platinotype “normal iron solution” carries about 1.2 per cent of it for this reason.
• It mobilises the photoproduct. Free oxalate converts insoluble iron(II) oxalate — 0.022 g per 100 cc — into the soluble bis-oxalato complex, which is the species that actually reduces silver. A sensitiser with some free oxalate in it therefore develops more readily than one without.
• It acidifies the coating, which is where Ware’s warning about nitrate and colloidal silver bites hardest, and which is one reason the developer has to be alkaline.

More of it improves keeping and solubility and pushes the coating further into the acidic region where image silver is most vulnerable during processing; less of it risks a sensitiser that throws a precipitate, and a print that develops sluggishly because its iron(II) will not dissolve. It also raises the amount of oxalate available to meet the silver, which has a hazard consequence set out under Incompatibilities.

Distilled water — the make-up volume, and not a neutral ingredient. Cassell specifies distilled and the chemistry justifies it: calcium in hard water precipitates calcium oxalate and simultaneously promotes aquation and hydrolysis of the iron(III) complex, which is the first step on the road to goethite in the paper. The line “water to make 1,000 ccs.” also fixes the strength of everything above it — the period convention is that a make-up volume, as opposed to a stated volume of water added, is what makes a formula reproducible.

What is deliberately not in it. No developing agent, because the developer is a separate bath and contains nothing that reduces silver. No restrainer and no contrast agent: the historical contrast control is a dichromate and it lives in the developer, not here. No humectant and no surfactant — Tween 20 as a spreading agent is a late twentieth-century addition, and the period answer to a badly-wetting paper was arrowroot size and a Blanchard brush. And no clearing agent anywhere in the process, which is the omission the whole of The mechanism has been about.

Interactions

Iron against silver. The formula is 2.5 parts of ferric oxalate to 1 of silver nitrate by mass, where the equal-parts coating mixture is 2 to 1 and the Formulary’s one-bottle kit is 3 to 1. The extra iron buys density and costs permanence, and because only a fraction of the iron is ever photoreduced, the surplus is not a safety margin but a stain in waiting.

Acid sensitiser against alkaline developer. The coating is acidic; the developers are borate- or tartrate-buffered and alkaline. That mismatch is not an oversight — it is Nicol’s defence against the nitrate ion — but the transition from acid to alkali is precisely when the excess iron(III) hydrolyses. The modern process breaks the conflict by clearing at pH 3 to 4 immediately after development, before anything dries.

Silver against oxalate, in the bottle. Free oxalate and silver ions in the same solution can give silver oxalate, and Photographers’ Formulary — which still sells a one-bottle sensitiser — says so plainly: “It is very common for a precipitate of silver oxalate to form. This precipitate does no harm. If your sensitizer does form a precipitate, just be sure you do not transfer it to the paper you will be coating.” That is a manufacturer’s statement about a manufacturer’s product and it is evidence. It is also the reason this formula is an exception to a rule the rest of the encyclopaedia enforces. See Incompatibilities.

Silver against ammonia, in the fixer. Nicol’s own fixer was weak ammonia and Cassell’s is hypo with 12.5 ccs of 0.880 ammonia per litre. Ammoniacal silver solutions are a documented explosion hazard when they are allowed to stand or are made alkaline further; the course has a dedicated quench SOP for them and does not reproduce the period’s ammoniacal sensitisers.

Dichromate against the highlights. Wall’s 1924 description of the bichromate in Thomson’s developer is that it “keeps the whites pure and acts as a restrainer”, which is a functional description worth having even though the course will not use the substance: it is a contrast and fog control acting in the development bath, not a sensitiser additive. The Getty records potassium chlorate and dichromate salts as the two period contrast controls for the platinotype, the same pair.

Size against sensitiser. Arrowroot, starch or gelatin all hold the coating nearer the surface. Different sizes give different colours in the same developer, which is one of the variables that made period results so inconsistent between workers who thought they were following the same formula.

Calcium against everything. Hard water, or a chalk-buffered paper, precipitates calcium oxalate in the sheet and accelerates the hydrolysis of the iron. It is a slow interaction and its effects appear in the highlights over decades.

Variants

⁇Where was the silver in Nicol's original process?

Sandy King’s account is that “in Nicol’s original patent, the print was developed in a silver nitrate bath. He patented several revisions in the early 1890s and in one of the formulas recommends making silver nitrate part of the sensitizing bath rather than adding it to the developer. This last revision is the method used by most contemporary kallitype printers.”

Cassell’s account is that the instructions it prints — a single sensitiser carrying both the iron and the silver — are “practically those originally published by Nichol”.

These need not contradict each other. King is describing the 1889 patent; Cassell says “originally published”, which could as easily mean Nicol’s published working instructions of the early 1890s, after the revision King describes. Ware’s footnotes record both a patent, No. 5,374 of 29 March 1889, and two papers by Nicol in the British Journal of Photography, on 14 March 1890 and 24 July 1891; his history of the chrysotype cites British Patents 5374 and 7312 with a date of 1892 attached, which is a third reading of the same patent record.

What the course does. It publishes the composition Cassell prints, because that is the only complete set of quantities read for this page that any source ties to Nicol, and it says on this page that the arrangement of the silver may not be the arrangement of the 1889 patent. No printing of Nicol’s own words was read. Until the patent specification or one of the two BJP papers is read directly, the question of which arrangement he claimed first stays open, and this page will not settle it by preferring whichever secondary source is more convenient.

🕰Robert Hunt described both arrangements in 1844

Forty-five years before the patent, Robert Hunt published both versions of the process in Researches on Light, in a paragraph explicitly offered as a gap-filler in Herschel’s work:

OXALATE OF IRON in solution, to which an excess of oxalic acid has been added, affords, after a few minutes’ exposure, when washed with nitrate of silver, a very intense black picture, which slowly fades into a dingy grey. If the oxalate of iron and silver be combined in the paper, and exposed, so powerful a picture results, that it is difficult to tell the right from the wrong side, the impression penetrating quite through the paper.

The first sentence is the silver-in-the-developer arrangement; the second is the silver-in-the-paper arrangement, which is the formula on this page. Hunt also recorded the fading — “slowly fades into a dingy grey” — in 1844. Ware’s judgement is that “in 1844, Hunt essentially invented the process that in 1889 was announced as Kallitype, and is generally attributed to Dr. W.W.J. Nicol.” The course records that as Ware’s assessment, not as a claim that Nicol copied anyone: Hunt published an observation, Nicol built a process, marketed a paper and patented it, and those are different achievements. Hunt’s own paragraph 231 states the general principle the whole family rests on — that all the persalts of iron are converted into protosalts by sunshine when in combination with organic matter.

Wall’s 1912 printing, at a tenth the scale with the oxalic acid fixed at 5 grains to the ounce. The two solids agree exactly with Cassell’s. It is not entered as a separate formula because it is the same formula; the disagreement over the oxalic acid is recorded under Mixing rather than resolved.

Photographers’ Formulary’s traditional kallitype kit, which is the one-bottle arrangement still in commerce: 30 mL of a 20 per cent ferric oxalate solution with 2 g of solid silver nitrate stirred into it, and ripened two or three days in the darkroom before use. It is 3 parts of iron salt to 1 of silver by mass against Nicol’s 2.5, and its silver strength is Cassell’s to within the uncertainty of the unstated final volume. If you want to know how Nicol’s coating solution behaves without mixing one yourself, this is the nearest thing anyone sells.

The two-bottle kallitype sensitiser, 20 per cent ferric oxalate and 10 per cent silver nitrate mixed in equal parts in the coating dish, at its own page. Four fifths of Nicol’s iron per unit of silver, two thirds of his total solids, nothing kept in contact that need not be, and a supplier on both sides of the Atlantic — the same maker’s New Kallitype kit is built this way, and so are Bostick & Sullivan’s and Sandy King’s methods. This is the descendant, and it is what the course recommends for printing.

Cassell’s “Modified Kallitype”, the water-developing paper. Four stocks — green ferric ammonio-citrate 252 g/L, tartaric acid 41 g/L, silver nitrate 106 g/L and gelatine 69 g/L — mixed in equal parts, so that the coating solution is a quarter of each strength; coated warm, printed to a vigorous image, developed in plain water for two minutes and fixed in a very weak hypo. Cassell’s own verdict is that the warm brown results “cannot, as a rule, be said to equal those given by the proper kallitype process”. Chemically this is not a kallitype at all: the iron salt is a citrate, there is a binder, and the image prints out. It is a Van Dyke in everything but name, and the comparison is the clearest illustration in the period literature of what changing the iron salt costs.

The American formulas of Hall and Thomson, printed by Wall in 1924 and by Cassell in 1911. They are the “needlessly complex elaborations” Ware complains about: Thomson’s later sensitisers carry ferric ammonio-citrate, ferric oxalate, potassium oxalate, cupric chloride, gum arabic, oxalic acid, silver nitrate and a dichromate solution in one bottle, and his developer is itself a silver nitrate solution with citric and oxalic acids. Two of the variants in that family are excluded by course policy rather than by taste: one uses uranium nitrate, which is Level D under the historical-study ruling, and the dichromates throughout are excluded by the chromium policy.

George E. Brown’s single-solution sensitiser, which Cassell calls one of the simplest and most widely used for menu cards and postcards: silver nitrate dissolved in a few drachms of water, ammonia added drop by drop until the silver oxide precipitate just redissolves, weak sulphuric acid added until the ammonia smell goes, then green ferric ammonium citrate. The course does not reproduce it. Making an ammoniacal silver solution and then adjusting its pH is the exact procedure behind the documented silver nitride incidents; see Incompatibilities.

Ware’s argyrotype, 1990, which is the considered modern answer to this formula rather than a variant of it. Ware went back to the underlying chemistry, identified the nitrate anion as the problem, and replaced it with silver sulphamate made in situ, giving a sensitiser that works at pH 2 to 3 and washes out without hydrolysing the iron.

Safety

Level B, and the sensitiser earns it twice over. The course’s classification rubric names silver nitrate explicitly at Level B, and ferric oxalate is classified Level B on its own page — its notified GHS classification rests on a single notifier reporting H302 and H312, which is too thin a basis to classify downward, so the course applies the precautions established for oxalic acid and the oxalate salts. This formula is half again as concentrated as the equal-parts coating mixture, which changes nothing about the controls and everything about how much of each substance is on the bench.

• Silver nitrate is corrosive and an oxidiser. Solid on skin can burn; solution stains brown-black because the silver binds to skin protein, and the stain wears off rather than washing off. Never put the solid in a wastepaper basket. The handling SOP and the spill SOP carry the procedure.
• Oxalates are systemic poisons. International Chemical Safety Card 0529 records that exposure to oxalic acid may affect the calcium balance, that repeated exposure can produce kidney stones and slow-healing ulcers, and that the substance is corrosive to eyes, skin and respiratory tract. Weigh without raising dust, wear nitrile gloves and splash goggles, keep an eyewash within reach, and eat nothing at the bench.
• Warming the solution is a step, so treat it as one. A water bath on a hotplate, a bottle that can take the heat, and no direct flame. The point of the water bath is temperature control, not ceremony: too hot and the ferric oxalate reduces itself.
• This is an ultraviolet process. Sunlight or a UV box, both of which are skin and eye exposures that do not feel like exposures. Enclose the unit and use a timer; the UV unit SOP covers it.

☠The historical developers, the historical fixer, and what this course will not do with either

Three of Nicol’s four developers carry potassium bichromate, at 87 to 112 ccs of a 1 per cent solution per litre in the black and purple baths and 50 to 62 ccs in the sepia one. The chromium policy settles this without argument: chromium(VI) is never used in this course at any level. The ECHA aggregation for potassium dichromate carries H350 for cancer, H340 for genetic defects and H360 for reproductive toxicity, and the UK workplace exposure limit for chromium(VI) is fifty times lower than for chromium(III). The three developer pages in this formulary give the dichromate-free baths, and the historical dichromate line is printed on this page as a record of what was published, not as an instruction. If a kit you own contains a dichromate packet it still has to be labelled, stored away from anything reducing, and disposed of through a licensed route.

The uranium variant in the American family is Level D and is covered by the historical-study ruling: described, never performed.

Brown’s ammoniacal single-solution sensitiser is not reproduced here. CAMEO records a case in which 28 per cent ammonium hydroxide and silver nitrate solution, treated with sodium hydroxide, gave a black precipitate of silver nitride that exploded on stirring, and NIOSH lists ammonia second among the incompatibilities of soluble silver compounds. Cassell’s procedure involves making exactly such a solution and then adjusting it with acid. The course teaches what it is and stops there; the ammoniacal silver quench SOP exists for the case where one has been made by accident.

⚠Silver and oxalate in one bottle: the exception, and its limits

Every oxalate page in this encyclopaedia carries the same incompatibility: the NIOSH pocket guide lists silver compounds among the incompatibilities of oxalic acid, and Chemical Safety Card 0529 names explosive silver oxalate as the reason. PubChem’s GHS record for silver oxalate carries the signal word Danger with the explosive pictogram, and 104 of 109 company notifications give H201, mass explosion hazard.

This formula puts them in the same bottle on purpose. What makes that a manageable situation rather than a reckless one is that the mixture is dilute and aqueous, made in millilitres, and — in the equal-parts practice — used within minutes of being mixed. Two things are different here and both deserve stating plainly. Cassell says the mixed sensitiser keeps for several months in the dark, so a bottle of it sits on a shelf in a way the two-bottle arrangement never allows. And a modern maker does the same thing on a shorter timescale: Photographers’ Formulary’s traditional kit is one bottle, ripened two or three days, and the sheet’s own account of the silver oxalate that forms in it is that the precipitate does no harm provided it is not carried onto the paper.

No source read for this page evaluates the risk of a dried silver-oxalate residue at photographic scale, and the course does not invent one. The precaution that follows from the uncertainty is the same as on the modern page and matters more here: do not let this mixture dry to a solid anywhere except in paper. Rinse the coating dish, the pipette and the brush immediately; wipe spills before they evaporate; never let a crust build in a bottle neck or a cap thread; and do not scale the mixing up beyond what a session will use.

⚠What is not a hazard here, and why

Nothing in the sensitiser is volatile, and no step evolves a gas you could smell. The only gas the chemistry produces is carbon dioxide, liberated inside the paper during exposure and bounded by the iron on the sheet. Ventilation is therefore not the governing control at the mixing bench; the reason to have ordinary extraction or an open window is dust while weighing, not vapour while working.

There is no strong acid or alkali in the sensitiser as mixed. The oxalic acid is present at around 1 per cent at most, and the borates and tartrates are in the developer, not here. The one genuinely respiratory hazard in Nicol’s process is the ammonia in the fixing bath, and it belongs to the fixer rather than to this page.

There is no cyanide chemistry, which separates this bench from a cyanotype’s, and no chromium anywhere in the sensitiser — the dichromate problem is entirely a developer problem.

The warming step is not a fire risk. A water bath cannot exceed 100 °C, nothing here is flammable, and the only reason the temperature matters is photochemical.

Storage

Cassell’s claim is that the mixed sensitiser keeps for several months in the dark, and it is worth treating carefully rather than repeating. It is a period statement about a period product, in a reference work that also gets the inventor’s name and date wrong. It is not, however, absurd on its face: Photographers’ Formulary’s traditional kit still mixes silver nitrate into ferric oxalate solution in one bottle and instructs the user to let it ripen two or three days in the darkroom before coating, stirring occasionally. A modern maker keeping that mixture for days makes months look like a difference of degree rather than of kind. The equal-parts practice supplies two bottles precisely so the mixture never stands at all, and is the safer default. What is not in dispute is the direction of the risk: a solution holding silver(I) and free oxalate together is a solution in which slow reactions are happening whether or not anybody has measured them.

The course’s practice, in the absence of evidence either way: mix what a session will use, keep it in a stoppered brown bottle in the dark and cool, label it with the strength, the date and the supplier of the ferric oxalate, and discard rather than accumulate. If you want to test Cassell’s claim, that is an experiment and it is set out under Experiments — with the residue precaution from Safety attached to it.

Ferric oxalate powder keeps and its solutions do not. King’s figure for the modern process is that the powder “lasts indefinitely, but once mixed with water will slowly degrade, with a resulting increase in print fog”, and he advises mixing no more than two to three months’ worth. That is the only keeping figure any source read gives for the iron half.

Silver nitrate goes in dark glass, tightly closed, away from organic material, away from reducing agents and away from ammonia. The storage rotation SOP and labelling cover the general practice.

Coated paper: no keeping time is published. Wall says printing may proceed at once when the sheets are dry, and neither book says what happens if it does not. The modern instruction for the equivalent sensitiser is an hour or two. Treat the historical silence as an absence of evidence, not as a licence.

Incompatibilities

• Oxalates and silver salts, as a storage and waste rule. Bottles apart, wastes apart, and never the two in a shared container or a shared spill tray. The exception this formula makes is deliberate, dilute and momentary, and it does not extend to the stock bottles.
• Silver nitrate and ammonia. Documented above. Never allow an ammoniacal silver solution to stand or dry, and never add alkali to one.
• Silver nitrate and alcohols, with which CAMEO records that it may produce silver fulminate, and acetylene, with which it gives silver acetylide in the presence of ammonia. Neither belongs anywhere near this bench.
• Silver nitrate and paper, in the solid state. It is an oxidiser; solid silver nitrate in a wastepaper bin is a fire.
• Ferric oxalate and reducing agents, including light and heat. The sensitiser’s whole function is to be reduced; anything that reduces it before the exposure is fog.
• Dichromate and anything reducing, if a period kit or a modern one has put one on your shelf. The chromium(III) rinse route is for chrome alum, not for dichromates, which go out through a licensed route.
• Hard water, at every stage. Distilled for the sensitiser, and — on the modern evidence about clearing — for the processing baths too wherever it can be managed.

Waste

Every bath in this process is silver-bearing except the sizing. The developer holds the silver that never got reduced, the fixer holds the silver-thiosulphate complex, and the first wash holds both. Kodak’s disposal guidance treats silver as one of the most frequently regulated parameters in photographic effluent, and Ilford’s advice for domestic users is to collect rather than to pour. Route them through the silver-bearing waste SOP.

Iron and oxalate. Iron(III) is not the environmental problem silver is, but oxalate is a systemic poison and the volumes here are small enough to collect without inconvenience. Collect the spent developer with the silver-bearing stream rather than sorting it.

Chromium(VI), if you have any. Licensed route only. It is not to be neutralised, diluted, or put down a drain, and it is not covered by the chromium(III) rinse route.

Ammonia in the fixing bath raises the pH of whatever it joins. Kodak’s sewer guidance works to a pH window, and an ammoniacal hypo bath tipped into a general container can take the whole container out of it.

Local regulation governs, everywhere. The course states chemistry and general practice; what is permitted for a household, a school and a business differs by jurisdiction, and the disposal page says so at more length.

Troubleshooting

Yellow highlights on a dry print, worsening over time. Residual iron(III), hydrolysed to iron(III) hydroxide and then to goethite. It is the signature defect of this workflow because the workflow has no clearing step. Once the print has dried, dilute acid will not shift it. The prevention is to clear before drying, which means inserting a modern clearing sequence into a historical process.

Shadows thinning over months or years while the highlights yellow. The same residual iron, doing the other half of its work: King’s account is that residual ferrous iron will eventually oxidise the image silver, and Ware’s is that the residual iron(III) is what stains and what the clearing sequence exists to remove — the two attributions are set out under The mechanism and both point at the same bottle of EDTA. If the print matters, tone it; if it has already faded, nothing on this page will bring it back.

A brown stain over the whole image after fixing. Silver sulphide from silver nitrate meeting thiosulphate — Ware’s first pitfall, and Cassell’s instruction to pass into the fixer “direct, without washing” is what invites it. The fix is a water rinse between the developer and the fixer, which costs nothing and is what every modern instruction specifies.

The image bleaching visibly in the fixer. Colloidal silver oxidised in the presence of thiosulphate. Shorten the fix, dilute it, keep the temperature down — the modern kit warns specifically that warm water accelerates the bleaching — and either tone before fixing or overprint to compensate. Bostick & Sullivan’s judgement is that those are the only two effective answers.

A weak image everywhere, with normal exposure. Either the development was cut short, in which case the iron(II) never fully dissolved and the reduction never finished, or the ferric oxalate solution has aged. Both books give long development for the first reason; King’s fog warning covers the second.

Fog, or a grey ground. Old ferric oxalate solution, light leaking onto the coated sheet, or heat during drying. Cassell’s warning in the American section is exact and applies here too: excess heat will convert the ferric salt to ferrous before you ever expose it.

Grit or specks in the coating. The solution was not filtered, or was filtered before the silver went in but stood long enough afterwards to throw a precipitate. Filter hot, add the silver to the clear filtrate, and coat from a freshly made bottle.

Inconsistency between sheets that ought to match. Look first at the paper and the size, not at the chemistry. Different sizings give different colours from the same developer, and a paper with a chalk buffer will behave differently from one without.

Experiments

1. Nicol’s proportions against the modern ones, on one negative. Coat one sheet with this formula and one with the modern equal-parts mixture, expose both under the same negative for the same time, develop both in the same bath, and read the maximum density and the highlight separation. The prediction from the arithmetic is that Nicol’s sheet is faster and deeper and that its highlights are harder to clear. Record the coating volume, because it is the variable most likely to spoil the comparison.

2. Put the clearing sequence back in. Two prints from one coated batch: process one exactly as Cassell prints it — develop, straight into the fixer, wash — and give the other a water rinse, the three-bath EDTA and sulphite clearing sequence, and then the same fixer. Compare the highlights wet, dry, and again after a week on a windowsill. This is the single most informative experiment on the page because it isolates the one difference that the whole permanence argument turns on.

3. Test Eder’s ranking. Coat one sheet with this sensitiser and one with a Van Dyke at its own published strength; expose a step wedge on each under identical light; find the exposure that gives matched mid-tone density. Eder’s table puts the oxalate at 89 and the citrate at 15. See whether the ratio you measure is anywhere near six to one, and think about why a period ranking of dry sensitised papers might not transfer to your materials.

4. Hunt’s two arrangements. Coat one sheet with iron and silver together, as here; coat another with the iron alone at the same strength; expose both identically; then, in dim light, brush the second with a silver nitrate solution before developing it. Hunt reported in 1844 that the first gives an image that penetrates the paper and the second a very intense black that fades. This is the experiment that would settle, empirically if not historically, why Nicol might have moved the silver from one place to the other.

5. Does Cassell’s several months stand up? Mix a small quantity, split it, keep half in a stoppered brown bottle in the dark and coat from the other half at once; coat from the stored half at one week, one month and three months, printing the same step wedge each time and measuring the base fog and the maximum density. Attach the residue precaution from Safety to this one: keep the volume small, keep the bottle capped, and rinse everything the moment it is finished with. Record which supplier’s ferric oxalate you used, because the answer may well belong to the bottle rather than to the formula.