The Additives We Study and Do Not Reproduce

Every process in Parts XXI to XXV has a bottle beside it that this course does not open. Not a different process — the same one, with something added to the sensitiser, the developer or the toning bath. Herschel put a mercury salt on a finished cyanotype. Willis put lead in his platinum coating and mercury in his sepia papers. Photographers added uranyl nitrate to a cyanotype sensitiser to make it faster, and a dichromate to a palladium developer to make the print brighter.

They are collected here rather than scattered across five parts because they are one body of chemistry with one classification, and because a warning attached to each process separately teaches nothing about why any of them were used. This page gives the chemistry, the history, the mechanism, the image effect and the hazard record. It gives no procedure and no quantity that could be used as one.

The boundary, stated first

Three pages in this course own these substances and they divide by what the substance is acting on, not by which element it is. Read this before you go looking for something you expected to find here.

Three pages, three questions, no overlap

1. 1This page — the addition inside a process this cluster printsA metal salt added to an iron or noble-metal sensitiser, to its developer, or to a toning bath applied to a Prussian blue or noble-metal image. What the addition did chemically, what it did to the picture, and what the course teaches instead of it
2. 2Part XX — the same chemistries acting on a developed silver gelatin imageUranium, mercury, cyanide and chromium(VI) as toners on a silver print. Different substrate, different mechanism, different page: /part-20-toning-chemistry/the-toners-we-study-and-do-not-use/
3. 3Part XXVI — the toxicology in depth, the conservation of objects containing them, and whole processes the course refuses to reproduceThe full hazard argument for uranium and mercury, what a conservator finds in the objects, what an owner should do, and the processes that are excluded entire rather than as an addition: /part-26-processes-we-do-not-reproduce/uranium-mercury-and-the-heavy-metal-treatments/

Each of the three names the other two and restates none of them. Where this page needs a fact that belongs to another, it links rather than repeating.

Two definitions live once, elsewhere, and every page in this cluster cites them rather than restating them: Level D at the Level D policy, and the oxidation-state rule at the chromium policy.

And one subject is not here at all. Potassium cyanide as a fixer and as a component of toning baths belongs to the printing-out and silver-gelatin pages; Part XXVI’s own lesson owns it. It is named in this sentence and nowhere else on this page. Note that this is not about the hexacyanoferrates: the ferricyanide in a cyanotype is a stable complex, not a cyanide salt, and Part XXI settles that question properly.

The rule being applied

Safety overrides historical authenticity. That is Rule 5 of the authoring standard, and this page is what it looks like in practice.

What Level D permits, and what this page therefore does: the chemistry as reactions with mechanism; the historical workflow in outline, told as what people did rather than as what you do; the image characteristics and why the chemistry produces them; the hazard record with named sources and dates; the reason each substance is classified here rather than at Level C; and, where the safer route the course teaches exists, the page that teaches it.

What Level D forbids at any length or in any tone: quantities to weigh out, a sequence of steps, a temperature and a time, a “safer” substitution, or a sentence of the form “if you must do this, then…”. A partial procedure is more dangerous than none, because it reads as permission and omits the controls.

!Where a controlled account exists, this page points to it

The course looked for institutions and qualified chemists who work with these additions under conditions a home darkroom cannot provide, and publish what they find. Four are worth naming, because they are the right answer to “but somebody must be doing this”.

Mike Ware himself, a chartered chemist and former university chemistry lecturer, prepared a mercury-containing platinum sensitiser in the laboratory at a stated molar concentration, coated and exposed it across a range of times, and published the quantitative X-ray fluorescence analysis of the resulting images — the proportion of mercury deposited against exposure — rather than a recipe to follow. That is the shape of a controlled demonstration: the measurement is published, the procedure is not offered to a general reader, and the author states his own hazard position separately.

Saori Kawasumi Lewis and Koseki, in 2015, analysed finished mercury-containing platinum prints by X-ray fluorescence and separated two historical practices by their signatures — mercury in the sensitiser against mercury in the developer.

The National Gallery of Art’s laboratories attempted to reproduce the discoloration of palladiotypes by accelerated ageing of simulacra, and Ware notes the word choice deliberately: simulacrum rather than facsimile, because it is curatorially undesirable to make facsimiles of images by Stieglitz.

The Getty Conservation Institute’s atlas documents each of these additions as an analytical signature in real objects — where the mercury is, where the lead is, what uranium looks like in an infrared spectrum — which is the use a conservator has for this chemistry and the use this page is written for.

None of the four publishes a procedure for a general reader, and this page does not become the place where one appears.

Mercury, inside two processes this cluster prints

Mercury turns up twice in this cluster and the two appearances have nothing chemically in common beyond the element.

Herschel’s hydrargyro-cyanotype, 1842

The oddest experiment in the whole cyanotype literature, and Ware — who has repeated it — cannot fully explain it. A finished, processed cyanotype is coated with a solution of a mercury(I) salt. Within a few hours the blue image disappears and the sheet is washed and dried, apparently blank. Heated strongly with a flat-iron, but not enough to scorch, a brown image appears. It fades slowly in the dark over days and can be raised again by further heating.

What is conjectured to be happening, and Ware marks it as conjecture: the mercury(I) may be trapped in the Prussian blue lattice in the same way thallium(I) is known to be — the mechanism the toners further down this page depend on — and then reduce the pigment to Prussian white, which is colourless. Heating may decompose the mercury ions to finely divided mercury metal, which appears brown. The observations, he says, deserve further scientific investigation.

Why it is here rather than in Part XXI or Part XXVI. It is an addition applied to a print made by a process this course teaches, so it is this page’s; the toxicology of mercury salts in depth is Part XXVI’s; and Ware’s own verdict is the one that ends the discussion — he has repeated and confirmed the process, and “cannot recommend it generally owing to the high toxic hazard presented by the mercury salts, especially when heated.” A procedure that ends by heating a mercury compound with a domestic iron is a procedure that ends by making a vapour, and mercury vapour is what NIOSH’s skin-notated limits and EH40’s 0.02 mg/m³ with a biological monitoring value are about. The formulary entry records the transformation, records that no source publishes a strength for the mercury bath, and gives no method.

The sepia platinotype, 1878 to 1911

This one was a commercial product for a quarter of a century, and the received account of it is wrong.

It was not a hot developer. Ware states it directly: it is sometimes mistakenly said that sepia platinotypes were made simply by developing at about 71 °C, and the degree of warming that achieves is found to be “only very slight”. The brown came from mercury.

The history is a chemist changing his mind about where to put it. Willis’s 1878 patent puts a small quantity of mercuric chloride in the developer, used hot. The problem that emerged was one he called “double tones” — a colour shift between the brown high values and the neutral dark shadows — which he found aesthetically unacceptable, and which he explained himself in a way that is a small lesson in concentration: where the ferrous image is strong, there is not enough mercuric chloride relative to it to have the same effect as on the weaker areas. Moving the mercury into the coating, in its proper proportion throughout, removed the problem, and that is how the commercial sepia papers were made from about 1885.

What the mercury was doing chemically is the most interesting part, and it is the same mechanism that explains lead and silver in the same sensitisers.

⚗Chloride scavenging: why a metal that is not the image improves the image

The chemistry lesson established that the electron transfer does not proceed through the [PtCl₄]²⁻ ion but through its aquated form, and that only about a fifth of the platinum is aquated at sensitiser strength.

H₂O + [PtCl₄]²⁻ ⇌ [Pt(H₂O)Cl₃]⁻ + Cl⁻

The aquation equilibrium that limits how fast platinum can be reduced

That equilibrium releases a chloride ion, so anything that removes free chloride pushes it to the right and increases the concentration of the reducible species. Mercury(II) binds chloride exceptionally strongly — Ware gives log K₁ = 6.74 and log K₂ = 6.48 for the first two steps and an overall formation constant of 1.2 × 10¹⁵ for [HgCl₄]²⁻ — so:

H₂O + [PtCl₄]²⁻ + Hg²⁺ → [Pt(H₂O)Cl₃]⁻ + HgCl⁺

Ware's chloride-scavenging hypothesis, written for mercury(II)

The prediction the hypothesis makes is testable and was tested from the other end: added chloride should inhibit the image, and excess chloride added to a platinotype sensitiser as an alkali metal chloride or as hydrochloric acid is indeed found to inhibit platinum image formation. Ware also notes a consequence that follows for the process itself — chloride is released by the reduction, so the development of a platinum print may be chemically self-inhibiting, which is part of why platinum exposures are longer than palladium’s from the same iron sensitiser.

But mercury is not only a scavenger. Its own redox potential is +0.854 V, well above the iron(II) oxalato couple, so it is itself reduced to metallic mercury and co-precipitates with the platinum. Ware’s own controlled preparation — a platinum sensitiser containing mercury(II) nitrate at a 1:1 molar ratio, coated, exposed across a range and analysed by X-ray fluorescence — found mercury precipitated far more readily than platinum, with Hg:Pt of about 4 at low exposures and about 2 at high. So a “sepia platinotype” made that way is substantially a mercury print.

Ware also lists chemistry the model does not cover: mercury(II) chloride is known to add to some platinum(II) complexes to give species with a platinum–mercury bond, so the sensitiser chemistry may be more complex than the equilibrium above.

What the image effect was, and what it cost. An image resembling palladium’s: lower in contrast, brown and very smooth. And the permanence record is bad in a specific, documented way. The Platinotype Company’s own manager, W. H. Smith, warned in 1911 that excessive mercuric chloride in the developer gave a disagreeable colour and an impermanent image, and stated that the company “had never advocated the use of mercury in their developers simply because it was not stable, and they would not include a formula in their instructions which would injure the reputation of platinum printing for permanency.” In 1915 he showed severely faded specimens to make the point. F. C. Lambert added a diagnostic in 1911 that a conservator can still use: mercury-developed sepia images are soluble in Farmer’s reducer. The AIC records the same thing from the objects: mercury-developed photographs can reduce in colour over time and appear lighter brown.

And the two practices leave different signatures. Lewis and Koseki’s 2015 X-ray fluorescence work found mercury-sensitised prints at Hg:Pt of 0.2 to 0.3 and mercury-developed prints at 0.8 to 1.1 — so a conservator can tell which method made a given print a century later, from the ratio.

What the course does instead: prints in palladium. This is the substitution the reader is already making without noticing. Ware’s statement is exact: mercury and lead “need not be introduced into a palladium sensitiser”, because palladium images are already naturally brown, and mercury(II) makes them more neutral rather than warmer, so there is little reason to use it. The colour that the sepia platinotype bought with a mercury compound is the colour palladium gives for nothing.

Lead, and a correction this course has already made

Lead salts in platinum sensitisers are commonly described as contrast additions, including in the curriculum entry that asked for this page. Part XXVI has corrected that against Willis’s own patents and the correction is not restated here. What matters for this cluster is the shorter version: lead was never a contrast control, it was a chloride scavenger like mercury, and it is not in the image.

Two facts make that concrete and both are analytical rather than argumentative. Lead is far too electropositive to be reduced by the iron(II) photoproduct — Ware gives Eº(Pb²⁺/Pb) as −0.126 V, against the iron(II) oxalato couple’s +0.02 V — so it cannot become image substance. And Matthew Clarke’s extensive X-ray fluorescence measurements find that the lead signal does not correlate with image density but sits as a more or less uniform background in the paper, suggesting an insoluble colourless salt such as the oxalate or the chloride. It is in the sheet and not in the picture. The formulary entry works the chemistry through, including the solubility products that make lead chloride and lead oxalate the plausible species.

Lead’s second appearance in this cluster is a real toner and it is a different mechanism. A cyanotype treated in a lead(II) acetate bath shifts towards a more violet blue — a procedure attributed to Oscar Bolle around 1900. Ware’s account of the mechanism is that the lead(II) cation is incorporated into the voids of the microporous Prussian blue structure, and the evidence is that in X-ray spectrometry the lead signal correlates with the iron signal of the pigment. He reports two consequences worth knowing: the colour change is permanent, and it stabilises the pigment against light fading by an exposure factor of about four — while the vulnerability to alkali is undiminished, so the treatment fixes one of cyanotype’s weaknesses and not the other.

The classification. Lead(II) nitrate is Danger with a reproductive-toxicity statement, a suspected-carcinogen statement and organ damage on repeated exposure — and with unusually weak agreement between notifiers, which is a reason to read the occupational regime rather than the label. EH40 does not carry lead in its table at all, because lead in Great Britain is regulated separately, under its own Control of Lead at Work Regulations. Ware’s own note on the lead cyanotype toner is the practical one: lead(II) acetate is seriously toxic by ingestion, it is said to taste sweet — its common name is sugar of lead — and experimental confirmation of that claim is “definitely not recommended for this cumulative poison”.

What the course does instead. For platinum, nothing: the metal that needed the help is not handled. For palladium, nothing is needed — palladium reduces fast enough without a scavenger, which is the whole of Ware’s argument for why it is the easier metal. For cyanotype, the substitution is tannic acid toning, which Part XXI teaches at Level A; what it gives up is the violet, and it does not deliver lead’s fourfold light-fastness gain either.

Uranium, as a sensitiser and as a toner

Uranium appears inside these processes in two quite different roles, and both are worth understanding because both were genuinely attractive.

As a sensitiser additive, it made a cyanotype faster. Uranium(VI) salts added to a cyanotype sensitiser were claimed to enhance sensitivity by catalysing the photoreduction of iron(III) to iron(II). Ware tested the claim himself and reports the result with its limit: adding uranyl nitrate at the 1 to 2 per cent level to both ferrioxalate and ferricitrate sensitisers does increase printing speed by about one stop, a factor of two, but further increases in uranium concentration bring little additional advantage — “so its use is not worth the toxic risk”. A second additive with a speed claim is recorded in the same passage and is not a heavy metal: vanadium pentoxide, from a 1924 US patent claiming a fourfold increase.

As a toner on a platinotype, it worked catalytically and produced something nobody wanted. Ware’s account of uranyl ferrocyanide as a toner of platinotypes is short and it is the best warning on this page. Treating an uncleared platinum print — one that still contains iron(III) — with this toner forms two pigments at once: uranyl ferrocyanide, which is red, and Prussian blue, which is blue, “and the mixtures of them can therefore produce a whole gamut of unattractive false colours superimposed upon the long-suffering platinum image.”

The Getty atlas records the procedure historically, attributed to Horsley-Hinton, as toning a fully processed black platinotype in a bath of uranium nitrate and acetic acid with potassium ferricyanide and ammonium sulfocyanide, to a deep-brown or red-brown tonality — and records that it can be reversed by washing the print in dilute ammonia, which is the property that condemns it. The atlas also gives the analytical signature: uranium’s L-alpha and L-beta X-ray fluorescence peaks at 13.61 and 17.22 keV, and the deposit identified as the uranium complex of the hexacyanoferric anion by its carbon–nitrogen stretch at 2062 cm⁻¹ in the infrared.

☠Why uranium is excluded, in the one sentence this page owes it

Two hazard regimes describe uranyl salts and this course can only read one of them. The chemical hazard is documented: NIOSH’s symptom list for soluble uranium compounds is largely a renal picture, the recommended limit is 0.05 mg/m³, and uranyl nitrate’s notified classification is Danger with fatal if swallowed and fatal if inhaled, plus an oxidiser statement. The radiological hazard is described nowhere in that record — no GHS classification covers it, EH40 does not list uranium at all, and NIOSH attributes the cancer potential to alpha emission and decay products such as radon rather than to the chemistry. The chemical toxicity generally governs at photographic quantities, and that statement is a reading rather than a source’s conclusion. A control cannot be specified against a hazard that has not been characterised, which is why this exclusion has a different logic from every other on this page.

Part XXVI carries the argument in full, including the radiological measurement on a toned print, the UK waste position and what an owner of such an object should do. This page states the one sentence and stops.

What the course does instead. For speed in a cyanotype, the substitution is a different formulation rather than an additive: the New Cyanotype and the Simple cyanotype are faster than the classic sensitiser by more than uranium’s one stop, and Part XXI’s variants lesson compares five formulations. For a red-brown platinum print, there is no substitution and the course says so: the colour uranium gave is not available by any route this course teaches, and what is available is a warm palladium image at low humidity, which is a different colour.

Thallium and nickel, in the Prussian blue lattice

These two are cyanotype toners and they share one mechanism with the lead bath above, which is why they belong together.

The mechanism is incorporation rather than conversion. Treating a Prussian blue image with a solution of a heavy metal salt may cause the metal cation to be taken into the pigment’s own microporous lattice and so modify its colour — the same structural property that makes Prussian blue an ion-exchange material and a clinical antidote to thallium and caesium poisoning. Ware conjectures why only some metals work: the metal may need two accessible oxidation states to intervene in the intervalence charge-transfer transition that gives the pigment its colour. Lead (+2, +4), thallium (+1, +3), nickel (+2, +3) and copper (+1, +2) all have a perceptible effect; alkali and alkaline-earth cations of similar size and charge do not.

Thallium gives a cornflower blue and is the most dangerous substance on this page by classification. Ware describes the shift as “slight but agreeable” and then declines to recommend it to the general public owing to the severe toxic hazard. Thallium(I) sulfate’s classification is harmonised rather than merely notified — fatal if swallowed, with organ damage on repeated exposure — and EH40 gives soluble thallium compounds a long-term limit of 0.1 mg/m³ with the Sk notation for skin absorption. A harmonised classification is a legal determination rather than a company’s filing, and it is the strongest form of hazard statement in the European system.

Nickel gives a slight greenish-blue shift and one substantial benefit, which is why it is not simply a curiosity: it markedly improves the pigment’s resistance to alkaline hydrolysis, which is cyanotype’s characteristic failure. Nickel(II) sulfate is a listed carcinogen, and that is what removes it.

The formulary entry records all three baths as chemistry with no procedure.

What the course does instead. For colour, tannic acid at Level A, taught in Part XXI’s toning lab; what it gives up is the specific hues these three cations produce, and the cost in density is measured on that page. For alkali resistance there is no substitution at all, and the honest statement is that the course has none: the answer it gives instead is the mounting rule — an unbuffered mount, no alkaline reserve anywhere near the print — which prevents the attack rather than resisting it.

Chromium(VI), and a distinction this page has to correct

The dichromates are the one family on this page that a reader might reasonably expect to be treated differently, and the reason is worth setting out, because this page’s own curriculum entry anticipated a different answer from the one the course settled on.

What the additives actually were. Two of them, in this cluster. Pizzighelli and Hübl’s 1882 formulations and the Getty atlas both record a dichromate as an alternative to chlorate for adjusting platinotype contrast — and the Getty adds the analytical detail that makes them distinguishable a century later, since chlorate is too soluble to leave a residue while chromium shows up under X-ray fluorescence. And Willis and Clements’ own instructions recommended a very small amount of potassium dichromate in the oxalate developer for “brilliant prints” by palladiotype, a practice Ware records Paul Strand and Ned Scott using. Ware notes it does not even keep: dichromate reacts readily with oxalic acid, being reduced to oxalato-chromium(III) complexes, so the additive loses effectiveness in an oxalate bath.

⁇Level C or never? The manifest said one thing and the written policy says another

What this page’s curriculum entry expected. It asked for chromium(VI) to be “distinguished carefully from the rest because it is Level C rather than Level D”, on the reasoning that dichromate contrast agents are real, effective and still sold, and that they need controls a home darkroom does not have rather than being unusable in principle.

What the course actually ruled. The chromium policy, settled on 4 September 2026 and written before this page, states the rule as chromium(VI) never used, at any level, anywhere, with chromium(III) as chrome alum permitted at Level B. It is unambiguous and it is a standing policy.

This page follows the written policy and records the discrepancy rather than quietly writing to the manifest, which is the same thing Part XXVI did when its own manifest entry described lead as a contrast agent and the patents said otherwise.

And the distinction the manifest was reaching for is real, so here it is properly. The exclusions on this page are not all the same kind of exclusion, and knowing which kind you are looking at is the transferable skill:

• Mercury, lead, uranium and thallium are excluded by Level D, which is a statement that no page can assume the controls their hazards require of a domestic reader.
• Chromium(VI) is excluded by a substance policy that applies course-wide, in every part, whatever the procedure. It reaches the same practical result by a different route, which is why the drop-system entry can say that two independent rulings remove the same bottle.
• Potassium chlorate is excluded on the rubric’s second criterion, which is neither of those: not acute hazard, not a policy, but the availability of a better route to the same result. Its own page puts the deciding operation at weighing the dry oxidiser on a bench carrying paper and dust, and the better route is a calibrated negative.

Three exclusions, three arguments, one result. A reader who can tell them apart can assess a substance they have never met, which is what Part XXVI’s assignment asks for.

Why chromium(VI) earns it. Potassium dichromate’s notified classification carries twelve hazard statements including H340 for genetic defects and H350 for cancer, alongside acute toxicity by three routes, severe burns, skin sensitisation and respiratory sensitisation. EH40 sets chromium(VI) compounds at 0.01 mg/m³ with the Carc, Sen and biological-monitoring annotations — and the presence of a biological monitoring guidance value is the telling detail: the recognised way of confirming that the control worked is to test the worker’s urine after the shift. That is what chromium(VI) control looks like where it is done properly, and it is not a thing a domestic darkroom does.

What the course does instead, and Ware’s own conclusion is the course’s: with the better controls available for modern negative-making, especially by digital means, contrast-enhancing agents such as chlorate, hexachloroplatinate(IV) or dichromate “become unnecessary when a correctly calibrated negative is made”. The one contrast agent this cluster does run is a 3 per cent hydrogen peroxide dose in the developer tray, published by Jean-Claude Mougin and discussed with its limits in the tonal-scale experiment.

The substitutions, addition by addition

Every row below states what the substitution costs, because a substitution offered without its cost is not an honest one.

What was added, and what this course does instead

1. Mercury(II) in a platinum sensitiser, for a sepia image → print in palladiumCosts: nothing, and this is the strongest substitution on the page. Palladium is naturally brown, Ware records mercury(II) making a palladium image more neutral rather than warmer, and the mercury route was documented as impermanent by the manufacturer's own manager in 1911 and demonstrated as faded in 1915costs nothing
2. Mercury(I) on a finished cyanotype, for a brown image → no substitution offeredCosts: the effect itself. Herschel's transformation has no safe analogue the course can source, and its chemistry is still conjecture. What the course offers instead is the honest statement that this one is not availableno route
3. Lead(II) in a platinum sensitiser, to make the platinum reduce → palladium needs no scavengerCosts: nothing, for the same reason. Ware: mercury and lead need not be introduced into a palladium sensitiser. The lead was solving a problem specific to platinum's slow aquation
4. Lead(II) on a cyanotype, for violet and light-fastness → tannic acid toningCosts: the violet hue, and the roughly fourfold light-fading protection Ware measured for the lead bath. Tannic acid gives a different colour and the course has no equivalent light-fastness figure for it
5. Uranyl in a cyanotype sensitiser, for one stop of speed → the New or Simple cyanotype formulationsCosts: nothing, and you gain. The modern oxalate sensitisers are faster than a uranyl-doped classic one, and Ware's own test found the uranium worth about a factor of two with little further gain
6. Uranyl ferrocyanide as a toner on a platinum print → no substitution offeredCosts: the red-brown. Nothing this course teaches makes that colour on a noble-metal print. Note that the toner was reversible by dilute ammonia even when it worked, which is a permanence objection independent of the toxicityno route
7. Thallium(I) or nickel(II) on a cyanotype → tannic acid for colour; nothing for alkali resistanceCosts: the cornflower blue, and — more seriously — the order-of-magnitude gain in resistance to alkaline hydrolysis that the nickel bath gives. The course's answer to that is a mounting rule rather than a treatment, and the rule prevents the attack instead of resisting itpartial
8. A dichromate or a chlorate in the sensitiser, for contrast → a calibrated negativeCosts: the ability to change contrast after the negative is made. Ware's own view is that this costs nothing at all with modern negative-making, and that both oxidants damaged image quality; the drop-system entry sets out the "false sparkle" objection in full. The one oxidiser this cluster runs is a peroxide dose in the tray

Three of the eight have no substitution, and saying so is the point of the column. A course that offered a safe version of everything would be inventing chemistry.

How to read a historical formula without being tempted by it

You will meet these formulas. They are in period manuals, in reprints, on websites and in the appendices of books the course itself cites. Six questions, in this order, will keep you out of trouble — and they are the questions this page has been applying throughout.

1. What does this addition actually do, mechanically? Not what the manual says it does. Lead in a platinum sensitiser was called a contrast addition for a century, and the patents say a reducing action and the X-ray fluorescence says the lead is not in the image. A stated function that no mechanism supports is a warning that the formula has been copied rather than understood.

2. Is the substance the image, or is it a helper? Mercury in a sepia platinotype is in the image, at two to four times the platinum. Lead is not in the image at all. Those are different objects and different conservation problems, and no amount of reading the recipe tells you which you have.

3. What is its classification, from a named source, with a date? Not what the manual believed, which is very often wrong: Wall’s 1924 handbook calls skin absorption of mercuric chloride “practically harmless” — a claim the current notified classification flatly contradicts with H310, fatal in contact with skin. A period source is primary evidence for what was done and worthless as evidence for whether it was safe.

4. Is there an exposure limit, and does it come with a notation? The notations are where the real information is. Sen means a sensitiser and a control failure may be permanent. Carc means a carcinogen. Sk means skin absorption, so a glove is a primary control rather than a courtesy. BMGV means the recognised way of confirming the control worked is to test the worker. And absence from the list is not a clearance — EH40 says so in its own introduction, and uranium and palladium are both absent for quite different reasons.

5. What did the substance’s own contemporaries say about it? This is the question people skip, and it is often the most damning. The Platinotype Company’s manager said in 1911 that his own company never advocated mercury in developers because it was not stable, and showed faded prints in 1915. Ware tested uranyl in a cyanotype sensitiser, measured the gain, and called it not worth the toxic risk. Wall said the uranium intensifier was not to be recommended except for extremely thin negatives — in 1924, on photographic grounds alone.

6. What does the course teach instead, and what does that cost? If the answer is “nothing”, that is information too. Three of the eight substitutions on this page do not exist, and a page that invented them would be doing the thing this whole part is written against.

☠And the question that is not on the list

“Could I do it just once, carefully, outdoors, in a small quantity?”

No, and the reason is in the classifications rather than in the quantity. A sensitiser is a sensitiser at any dose — HSE’s own paragraphs are explicit that once the airways are hyper-responsive, further exposure to tiny quantities may provoke symptoms, and that it is impossible to identify in advance who will become hyper-responsive. A carcinogen has no threshold that a domestic reader can establish. A cumulative poison is cumulative. And every one of these substances leaves a waste stream that has no domestic route: what may lawfully be done with it differs between jurisdictions and changes, and this course publishes no jurisdiction-specific instruction. See the disposal ruling and check your local regulations; they govern.

The honest form of the impulse behind the question is curiosity about the chemistry, and that is what this page is for.

Recap

Five families of addition were made inside the processes this cluster prints. Mercury(II) in a platinum sensitiser scavenged chloride to make the platinum reduce, co-precipitated with it at two to four times its own molar share, gave the sepia platinotype its colour, and was documented as impermanent by the manufacturer’s own manager in 1911. Lead did the same scavenging job and stayed in the paper rather than the picture, and in a quite separate role it walks into the Prussian blue lattice and turns a cyanotype violet. Uranium made a cyanotype about a stop faster and made an uncleared platinum print two colours at once. Thallium and nickel enter the same lattice as lead, one for a cornflower blue and one for alkali resistance. And a dichromate raised contrast in a platinum sensitiser or a palladium developer.

None of it is given here as a procedure. Three of the five are excluded at Level D, because no page can assume the controls a domestic reader would need. Chromium(VI) is excluded by a course-wide substance policy that this page follows in preference to its own curriculum entry. Potassium chlorate is excluded on a third criterion again — not hazard, not policy, but a better route to the same result.

And the substitution that matters most in this part costs nothing at all. The brown image that the sepia platinotype bought with a mercury compound is the image palladium gives by itself.

Check your understanding

Question 1. For three of the additions on this page, state what each was used for inside which process, and what the course teaches instead. Which set is correct?

• Mercury(II) for contrast in a cyanotype → tannic acid; lead for speed in a kallitype → a faster developer; uranium for colour in a salt print → gold toning
• Mercury(II) for a sepia image in a platinum sensitiser → print in palladium, which is naturally brown; lead(II) for a violet shift on a cyanotype → tannic acid toning, giving up the violet and the light-fastness gain; uranyl for about a stop of speed in a cyanotype sensitiser → the New or Simple cyanotype formulations, which are faster still
• Mercury(I) for contrast in a palladium developer → hydrogen peroxide; thallium for speed in a platinotype → a calibrated negative; nickel for colour in a Van Dyke → gold toning
• A dichromate for permanence in a platinum sensitiser → longer clearing; uranium for contrast in a kallitype → a different developer; lead for colour in a salt print → a warmer paper

Show the answer and why

Answer: Mercury(II) for a sepia image in a platinum sensitiser → print in palladium, which is naturally brown; lead(II) for a violet shift on a cyanotype → tannic acid toning, giving up the violet and the light-fastness gain; uranyl for about a stop of speed in a cyanotype sensitiser → the New or Simple cyanotype formulations, which are faster still

Each pairing has to name the right substance, the right process, the right function and the right substitution, and the wrong options each break at least two of those. The mercury one is the most instructive: it is the only substitution on the page that costs nothing, because palladium gives the colour mercury was added to produce — and Ware records that mercury(II) actually makes a palladium image more neutral, so adding it to palladium would work against the effect it was used for in platinum.

Question 2. You find a nineteenth-century platinotype manual giving a contrast addition containing a mercuric salt. What do you do with it, and why?

• Try it once outdoors in a very small quantity, with gloves
• Read it as evidence of what was done, check the current classification of the substance rather than the manual's opinion of it, ask what the addition did mechanically and whether it is in the image, and note that this particular claim is probably wrong — mercury in a platinum sensitiser was a chloride scavenger and a co-precipitating image metal, not a contrast control in the sense a modern reader means
• Discard it as unreliable, since period manuals are not sources
• Substitute a modern oxidiser at the same molar concentration and follow the rest of the procedure

Show the answer and why

Answer: Read it as evidence of what was done, check the current classification of the substance rather than the manual's opinion of it, ask what the addition did mechanically and whether it is in the image, and note that this particular claim is probably wrong — mercury in a platinum sensitiser was a chloride scavenger and a co-precipitating image metal, not a contrast control in the sense a modern reader means

Options one and four are the two failure modes this page is written against — the small-quantity exception, which the classifications refuse because a sensitiser is a sensitiser at any dose, and the "safer substitution" that turns a Level D account into a procedure. Option three throws away real evidence: a period manual is primary evidence for what was done and worthless as evidence for whether it was safe, and those two uses have to be kept apart. And the specific correction matters: Ware's chloride-scavenging account plus his X-ray fluorescence measurement of Hg:Pt ratios between 2 and 4 describe something quite different from a contrast agent.

Question 3. A reader asks you where to find three things: the toxicology of uranium salts in depth, the uranium toning of a developed silver gelatin print, and the uranium sensitiser of a cyanotype variant. Name the page that owns each.

• All three are on this page, which collects the heavy-metal chemistry of the whole course
• Toxicology in depth and conservation handling: Part XXVI's heavy-metal treatments lesson. Uranium toning of a developed silver gelatin image: Part XX's "the toners we study and do not use". The uranium sensitiser inside a cyanotype: this page, because it is an addition inside a process this cluster prints
• Toxicology: this page. Silver gelatin toning: Part XXVI. Cyanotype sensitiser: Part XXI
• All three are in the formulary, and the lesson pages only cross-reference it

Show the answer and why

Answer: Toxicology in depth and conservation handling: Part XXVI's heavy-metal treatments lesson. Uranium toning of a developed silver gelatin image: Part XX's "the toners we study and do not use". The uranium sensitiser inside a cyanotype: this page, because it is an addition inside a process this cluster prints

The division is by what the substance is acting on, not by which element it is, and that is what makes it hold: the same salt does genuinely different chemistry as a sensitiser additive, as a toner on silver, and as an object in a collection. The formulary does carry the compositions, which is why option four is half right — but a formulary entry records a formula, and the argument about ownership, mechanism and substitution lives in the lessons.

Question 4. Mercury, lead, uranium and thallium are excluded at Level D; chromium(VI) is excluded by a course-wide substance policy; potassium chlorate is excluded on the rubric's second criterion. What is the difference between the second and the third?

• There is none; both mean the substance is too dangerous for a home darkroom
• The chromium ruling is about the substance and applies in every part whatever the procedure, resting on a carcinogenicity and sensitisation classification and a control regime that includes biological monitoring; the chlorate ruling is about the availability of a better route to the same result, and turns on the fact that a calibrated negative gives the contrast without the oxidiser
• Chromium is excluded because it is expensive and chlorate because it is unstable
• Chlorate is the more hazardous of the two, which is why it is excluded more strictly

Show the answer and why

Answer: The chromium ruling is about the substance and applies in every part whatever the procedure, resting on a carcinogenicity and sensitisation classification and a control regime that includes biological monitoring; the chlorate ruling is about the availability of a better route to the same result, and turns on the fact that a calibrated negative gives the contrast without the oxidiser

Three exclusions, three arguments, one practical result — and being able to tell them apart is the transferable skill. Chlorate is a strong oxidiser with a real reaction hazard on a bench carrying paper and dust, but its own page is explicit that the deciding consideration is that the contrast it buys is available from a calibrated negative without the graininess and the "false sparkle" of truncated high values. Nothing comparable is true of a dichromate, whose classification carries H340 and H350 and whose exposure limit comes with a biological monitoring guidance value.

Question 5. Ware records that adding uranyl nitrate to a cyanotype sensitiser increases printing speed by about one stop. Why does the course not treat that as a reason to consider it?

• Because the measurement has not been replicated
• Because Ware himself measured the gain, found that further increases in uranium brought little additional advantage, and concluded it was not worth the toxic risk — and because the modern oxalate cyanotype formulations the course already teaches are faster than a uranyl-doped classic sensitiser anyway, so the substitution costs nothing
• Because uranium cannot legally be bought
• Because a one-stop gain is too small to be measurable

Show the answer and why

Answer: Because Ware himself measured the gain, found that further increases in uranium brought little additional advantage, and concluded it was not worth the toxic risk — and because the modern oxalate cyanotype formulations the course already teaches are faster than a uranyl-doped classic sensitiser anyway, so the substitution costs nothing

Two independent reasons, and the second is the one that makes the exclusion easy rather than regrettable. The person who published the measurement drew the conclusion, which is the strongest form this kind of evidence takes; and the alternative the course already teaches is better on the same axis, so nothing has to be given up. Compare that with the nickel cyanotype toner, where the benefit — an order-of-magnitude gain in resistance to alkaline hydrolysis — has no substitute at all, and the page has to say so rather than pretending the trade is free.

Check my answersClear