Three-solution drop system for platinum and palladium

Three bottles and a shot glass, and the number of drops taken from each bottle is the contrast grade. No filter, no second paper, no change of developer: you decide how hard the print will be before you have coated anything, by counting. It is one of the oldest surviving pieces of working photographic craft in this formulary — published in Vienna in 1882, carried into American practice in 1937, and printed on the instruction sheet of a kit you can buy this week — and understanding why it has three bottles instead of two explains most of what a siderotype is.

The third bottle contains an oxidising agent, and its job is to destroy part of what the light has just made.

☠Level D: this page is not a procedure, and two separate course rules put it here

Platinum. The course ruled on 4 September 2026 that students perform palladium and never handle platinum, and the Level D policy carries the argument: the statement that governs it is H334, may cause allergy or asthma symptoms or breathing difficulties if inhaled, and HSE’s EH40 sets halogeno-platinum compounds at 0.002 mg/m³ as platinum with the Sen notation, the tightest limit anywhere in this course.

Potassium chlorate. Its encyclopaedia entry is at Level D too, on the rubric’s second criterion rather than on acute hazard: a safer modern route to the same result exists and this course teaches it. That page gives no quantity, no drop count and no sensitiser formula, and says so.

Either ruling alone would land this entry here. So this page gives no procedure: no coating routine, no drying regime, no exposure time or processing sequence appears below as an instruction to a reader, and no protective equipment list is offered, because equipment is not what the classification turns on.

The composition and the drop tables are published because a formula that cannot be read cannot be identified — and identification, of a bottle on a shelf, of a print in a collection, of a set of numbers in a manual, is the use a reader has for this entry. The relatives this course can publish are the palladiotype as a process at Level B, and, in the formulary, the kallitype sensitiser, which is this same iron chemistry with silver in place of the noble metal and no oxidant anywhere in it.

Solution 1 — ferric oxalate (Sensitizer A) — the light-sensitive iron, with no contrast agent in it

Ingredient	Quantity	Form the source specifies
Ferric oxalate	20 g	trihydrogen ferric oxalate, H3Fe(C2O4)3, prepared by the supplier by the iron alum and oxalic acid route and containing a slight excess of oxalic acid. The kit supplies it ready-made in 30 mL bottles and the reader never weighs it
Water	to make 100 mL	The sheet does not have the reader make this bottle; it states its strength as 20 per cent, and 20 g of solute in 100 mL of made-up solution is what that unit means. Bostick & Sullivan's equivalent bottle is 27 per cent and Ware's standard is 25 per cent with 2 g of oxalic acid per 100 cc.

Solution 2 for platinum — ferric oxalate with potassium chlorate (Sensitizer B) — the same iron, carrying the oxidant that shortens the exposure scale

Ingredient	Quantity	Form the source specifies
Solution 1 — ferric oxalate (Sensitizer A)	30 mL	the kit's second bottle of Solution 1, which is identical to the first until the sachet goes in
Potassium chlorate	0.18 g	supplied as a pre-weighed sachet, which the sheet instructs be transferred whole because the amount is small enough that any left behind matters
Water	to make 30 mL	No water is added. The sachet is tipped into the 30 mL bottle of Solution 1 and shaken until it dissolves, and the bottle is still 30 mL. The make-up volume is recorded because the schema requires a solution to state one, and it is the bottle's own.

Solution 2 for palladium — ferric oxalate with potassium chlorate (Sensitizer B) — the same solution again at twice the chlorate, because palladium answers it less

Ingredient	Quantity	Form the source specifies
Solution 1 — ferric oxalate (Sensitizer A)	30 mL	the palladium kit's second bottle of ferric oxalate, at the same 20 per cent
Potassium chlorate	0.26 g	the palladium kit's sachet — about 1.4 times the platinum dose in this pair of sheets, which Bostick & Sullivan describe more roundly as twice as much
Water	to make 30 mL	Again no water is added, and again the bottle is its own make-up volume.

Solution 3 — platinum — the image metal, and not light-sensitive at all

Ingredient	Quantity	Form the source specifies
Potassium tetrachloroplatinate(II)	20 g	potassium chloroplatinite, K2PtCl4. The sheet warns that the salt carries varying amounts of the much less soluble potassium hexachloroplatinate(IV), that a 20 per cent solution is near saturation, and that a fine red precipitate is common and should not be carried over into the sensitiser
Water	to make 100 mL	The kit supplies 5, 15 or 30 mL of a stated 20 per cent solution and the reader never makes it. Ware's own standard platinum solution is about 18.5 per cent, or 0.45 M, and he notes that the salt's true solubility is not clearly documented.

Solution 3 — palladium — the image metal for the cheaper and more reactive half of the pair

Ingredient	Quantity	Form the source specifies
Sodium tetrachloropalladate(II)	15 g	Na2PdCl4. The sheet gives the strength and not the hydrate; the trihydrate and the anhydrous salt differ by about a sixth in palladium content
Water	to make 100 mL	The kit supplies 15 or 30 mL of a stated 15 per cent solution ready-made. Ware's own standard is a 0.5 M solution, made either from the solid double salt or from palladium(II) chloride dissolved in hot sodium chloride solution.

Mixed in the ratio — platinum, for very soft negatives

0.478 parts Solution 2 for platinum — ferric oxalate with potassium chlorate (Sensitizer B) + 0.521 parts Solution 3 — platinum

FOR VERY SOFT NEGATIVES — Solution A 0, Solution B 0.478, Platinum Solution 0.521.

The chlorate end of the range: every drop of iron carries the oxidant. Solution 1 is absent altogether, which is why it does not appear here.

Mixed in the ratio — platinum, for soft negatives with highlight densities about 1.1

0.174 parts Solution 1 — ferric oxalate (Sensitizer A) + 0.304 parts Solution 2 for platinum — ferric oxalate with potassium chlorate (Sensitizer B) + 0.522 parts Solution 3 — platinum

FOR SOFT NEGATIVES (HIGHLIGHT DENSITIES ABOUT 1.1) — Solution A 0.174, Solution B 0.304, Platinum Solution 0.522.

Mixed in the ratio — platinum, for average negatives with highlight densities about 1.3

0.304 parts Solution 1 — ferric oxalate (Sensitizer A) + 0.174 parts Solution 2 for platinum — ferric oxalate with potassium chlorate (Sensitizer B) + 0.521 parts Solution 3 — platinum

FOR AVERAGE NEGATIVES (HIGHLIGHT DENSITIES ABOUT 1.3) — Solution A 0.304, Solution B 0.174, Platinum Solution 0.521.

The sheet's own reference point: a negative that gives a good print on grade 1 paper will print correctly with the average print mixture.

Mixed in the ratio — platinum, for moderately contrasty negatives

0.391 parts Solution 1 — ferric oxalate (Sensitizer A) + 0.086 parts Solution 2 for platinum — ferric oxalate with potassium chlorate (Sensitizer B) + 0.521 parts Solution 3 — platinum

FOR MODERATELY CONTRASTY NEGATIVES — Solution A 0.391, Solution B 0.086, Platinum Solution 0.521.

Mixed in the ratio — platinum, for very contrasty negatives with highlight densities about 1.5

0.478 parts Solution 1 — ferric oxalate (Sensitizer A) + 0.521 parts Solution 3 — platinum

FOR VERY CONTRASTY NEGATIVES (WITH NORMAL SHADOW DENSITY AND HIGHLIGHT DENSITIES ABOUT 1.5) — Solution A 0.478, Solution B 0, Platinum Solution 0.521.

No chlorate at all, and the sensitiser reverts to the two-bottle mixture Willis used and Ware still uses. This row is the whole argument of the page: the third bottle exists for negatives that are not contrasty enough, and a negative that is contrasty enough does not need it.

Mixed in the ratio — palladium, for very soft negatives

0.478 parts Solution 2 for palladium — ferric oxalate with potassium chlorate (Sensitizer B) + 0.521 parts Solution 3 — palladium

FOR VERY SOFT NEGATIVES — Solution A 0, Solution B 0.478, Palladium Solution 0.521.

Mixed in the ratio — palladium, for soft negatives with highlight densities about 1.1

0.174 parts Solution 1 — ferric oxalate (Sensitizer A) + 0.304 parts Solution 2 for palladium — ferric oxalate with potassium chlorate (Sensitizer B) + 0.522 parts Solution 3 — palladium

FOR SOFT NEGATIVES (HIGHLIGHT DENSITIES ABOUT 1.1) — Solution A 0.174, Solution B 0.304, Palladium Solution 0.522.

Mixed in the ratio — palladium, for average negatives with highlight densities about 1.3

0.304 parts Solution 1 — ferric oxalate (Sensitizer A) + 0.174 parts Solution 2 for palladium — ferric oxalate with potassium chlorate (Sensitizer B) + 0.521 parts Solution 3 — palladium

FOR AVERAGE NEGATIVES (HIGHLIGHT DENSITIES ABOUT 1.3) — Solution A 0.304, Solution B 0.174, Palladium Solution 0.521.

Mixed in the ratio — palladium, for moderately contrasty negatives

0.391 parts Solution 1 — ferric oxalate (Sensitizer A) + 0.086 parts Solution 2 for palladium — ferric oxalate with potassium chlorate (Sensitizer B) + 0.521 parts Solution 3 — palladium

FOR MODERATELY CONTRASTY NEGATIVES — Solution A 0.391, Solution B 0.086, Palladium Solution 0.521.

Mixed in the ratio — palladium, for very contrasty negatives with normal shadow detail and highlight densities about 1.5

0.478 parts Solution 1 — ferric oxalate (Sensitizer A) + 0.521 parts Solution 3 — palladium

FOR VERY CONTRASTY NEGATIVES (WITH NORMAL SHADOW DETAIL AND HIGHLIGHT DENSITIES ABOUT 1.5) — Solution A 0.478, Solution B 0.0, Palladium Solution 0.521.

The palladium sheet's five parts figures are identical to the platinum sheet's, to the third decimal place, including the 0.522 that appears once in each and looks like a rounding slip carried across both documents.

Purpose

To fix the contrast of a platinum or palladium print before the paper is coated, by choosing how much of an oxidising agent goes into the sensitiser.

That is a strange place to put a contrast control and it is worth saying why it ended up there. A silver gelatin printer changes contrast at the moment of exposure, with a filter or a graded paper, and can change their mind after seeing a test strip. A platinum printer has no such lever: the paper is made one sheet at a time in a shot glass, the metal is expensive, and the exposure takes a quarter of an hour under ultraviolet. So the decision was moved to the only place it could go — into the mixture — and the mixture was made adjustable by keeping the iron in two bottles that differ in exactly one substance.

Three bottles, but only two chemistries. Solution 1 and Solution 2 are the same ferric oxalate at the same strength, and the only difference between them is a little potassium chlorate in the second. Drawing more drops from bottle 2 and fewer from bottle 1 raises the chlorate concentration in the sensitiser without changing the amount of iron, the amount of metal or the volume that has to be spread over the sheet. That last point is the elegant part, and it is why the published tables hold the metal at a constant 0.521 or 0.522 of the mixture in every one of the five grades: only the ratio of the two iron bottles moves.

Recommended uses

What the sources record, as history and as current commercial practice. The course recommends none of them, and the section is here because reading a period manual or a kit sheet requires knowing what the numbers were for.

A negative whose density range is shorter than the process’s exposure scale. Ware puts the unmodified platinotype’s exposure scale at about 2.0 and observes that the drop-counting system “has become standard practice today with many users, whose negatives presumably have insufficient density range for successful printing by the unmodified platinotype sensitizer”. The suppliers agree from the other direction: Bostick & Sullivan ask for a negative of 1.35 to 1.50 density range, the Formulary’s palladium sheet for 1.3 to 1.5, and both Formulary sheets give the average mixture as the one for a negative that would print well on grade 1 paper.

Flat negatives, in the period literature. Abney’s fault list for the platinotype runs in both directions — prints too flat, use more chlorate; prints too hard, too much chlorate — which tells you the drop count was adjusted print by print rather than set once.

Palladium more than platinum. The palladium sheet’s own explanation for doubling the chlorate is that palladium is less responsive to it, and the palladiotype’s native contrast is lower. Ware records that very few contemporary practitioners attempt pure platinum at all, most using palladium or a mixture in which palladium predominates.

Not for a negative made on purpose for the process. This is the recommendation the sources themselves converge on and it is the one the course follows. Ware: 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”.

When another formula is preferable

• The palladiotype, with the contrast put in the negative. This is the course’s route to a noble-metal print: the same iron photochemistry, the same coating craft, a metal that is neither a respiratory sensitiser nor five times the price, and no oxidant in the bottle. The Willis potassium oxalate developer and the sodium citrate developer are its two published developers, and the clearing sequence is the step that decides whether the print lasts.
• Ware’s print-out platino-palladiotype. His own answer to the six objections he raises against the traditional method, of which the chlorate is one. It replaces ill-characterised ferric oxalate with ammonium iron(III) oxalate, coats by rod rather than by brush, prints out so that test strips become unnecessary, and takes its contrast from controlled humidity and from the platinum-to-palladium ratio instead of from an oxidant.
• A calibrated negative. The course’s honest answer. A correction curve applied to a digital negative gives an exposure scale matched to the process for nothing, without graininess and without a strong oxidiser on a domestic bench. The chlorate bottle exists because nineteenth-century printers could adjust the sensitiser and could not adjust the negative.
• The kallitype sensitiser, to learn on. Two bottles of the same iron chemistry with silver instead of platinum, at a fraction of the cost, and the coating, drying, exposing and clearing are the same skills. A reader who cannot yet coat an even sheet should not be practising on palladium and must not be practising on platinum.
• The classic cyanotype sensitiser, to learn on more cheaply still, if the point is to understand ultraviolet contact printing rather than noble-metal chemistry.
• Sodium hexachloroplatinate(IV), if a contrast agent is used at all. Ware records it as the nineteenth-century alternative oxidant, still sold at a high price, and said to cause less image deterioration than chlorate. It carries its own problem: it cannot be used with any ammonium ion in sensitiser or developer, because ammonium hexachloroplatinate(IV) is barely soluble and crystallises out. It is also platinum, so it is not a route this course opens.

Mixing

This section gives no mixing procedure. What can be described without giving instructions is the shape of the thing, because in this formula the shape carries almost all of the chemistry.

Three bottles, two of which start out identical. The kit ships two bottles of ferric oxalate solution and one pre-weighed sachet of potassium chlorate. One bottle is Solution 1; the sachet goes into the other and makes it Solution 2. No water is added at any point, which is why the make-up volume recorded for Solution 2 above is the bottle’s own 30 mL. The metal solution is bought ready-made and never mixed at all.

The reason the sachet is pre-weighed is worth pausing on, because it is also the reason the chlorate page identifies weighing the dry salt as the operation that decides the classification. 0.18 g is a quantity a domestic balance reads badly, and the sheet warns that all of it must be transferred because there is so little. A supplier weighing it once, in a place set up to weigh oxidisers, removes that operation from the darkroom — and does not remove the reason the course teaches the alternative instead.

🧮Reading the parts column, and what 20 drops per millilitre means

Photographers’ Formulary publishes each mixture twice: as parts of the total volume, and as drops for two print sizes. The parts are the formula; the drops are the formula realised at a sheet size.

Take the average mixture. The parts are Solution 1 at 0.304, Solution 2 at 0.174 and the metal at 0.521, which sum to 0.999 — they are fractions of the whole, not a ratio to be scaled. Multiply by the volume you need in drops and round:

0.304 × 55 = 16.7 drops

Solution 1, for a 55-drop mixture

0.174 × 55 = 9.6 drops

Solution 2, for a 55-drop mixture

0.521 × 55 = 28.6 drops

Metal solution, for a 55-drop mixture

The sheet’s own worked example rounds these up and notes that you cannot measure a fraction of a drop, so the mixture becomes 56 drops rather than 55. That rounding is not fussiness: at these volumes one drop is about two per cent of the whole.

The drop is the unit and the drop is not standard. The sheet’s own assumption is 20 drops per millilitre, so a drop is 0.05 mL — which is also the figure Ware uses when he works out that Irving Penn’s worksheet volumes cannot have been drops. A different dropper gives a different drop, and every figure on this page moves with it. That is the standing weakness of drop counting as a method, and it is why the conservation literature states coating volumes per square metre instead.

The two suppliers do not agree about the iron, and neither agrees with the conservation standard. Photographers’ Formulary supplies 20 per cent; Bostick & Sullivan supply 27 per cent for both their #1 and #2 bottles; Ware’s standard siderotype solution is 25 per cent with 2 g of oxalic acid per 100 cc, which he says is the strength Pizzighelli and Hübl’s analysis originally specified. A drop count written for one of those bottles is not a drop count for another, and this is the commonest way a formula from a book fails in a darkroom stocked from a different supplier.

Ferric oxalate is the reason the strengths cannot simply be converted. Ware calls it “a chemists’ nightmare”: polymorphic, apparently uncrystallisable, and sold in states of hydration that sources give variously as the tetrahydrate, the pentahydrate and the hexahydrate, with formula weights from 375.76 to 483.84. Twenty grams of it is a weight of an ill-defined solid rather than a known quantity of iron, and a page that converted 20 per cent to a molarity would be inventing the certainty the substance does not have.

One thing the sheets are clear about and it is a matter of chemical identity, not of strength. Two substances are sold as “ferric oxalate”: trihydrogen ferric oxalate, H₃Fe(C₂O₄)₃, and tripotassium ferric oxalate, K₃Fe(C₂O₄)₃. Both are light-sensitive; only the acidic form is sensitive enough to be useful, and Photographers’ Formulary states plainly that it does not recommend the green solid potassium salt for platinum, palladium or kallitype printing. The green salt is converted to the active form by acid, which is one of the things the excess oxalic acid in the bottle is doing.

The metal solutions are near saturation and behave like it. The platinum sheet warns that a 20 per cent solution of potassium chloroplatinite is close to its solubility limit, that the commercial salt carries some of the far less soluble hexachloroplatinate(IV), and that a fine red precipitate at the bottom of the bottle is normal and must not be transferred. Ware adds a second reason the solubility figures in the literature disagree: the dissolved anion slowly aquates, with a measured half-time of 2.4 hours, which is why a freshly made platinum solution is left to mature for a day before first use.

Behaviour

Recorded from the sources so that the system can be understood, not so that it can be run.

Almost nothing appears during the exposure, and what does appear is the wrong colour. The Getty atlas describes step 4 of the process as producing “a faint brownish image made from the photochemically generated ferrous oxalate”. The picture is not there yet; a distribution of iron(II) is.

Then the image arrives all at once. Both suppliers describe development as complete within seconds. Bostick & Sullivan instruct that the developer be poured over the print fast enough to break air bubbles, because an area the developer reaches a second late is an area that develops differently. Photographers’ Formulary makes the same point about a heated bath and streaking.

The developer is a variable, not a constant. Bostick & Sullivan state that the bath works from room temperature to boiling and that print colour and contrast vary with the temperature; the Getty atlas explains the direction — hotter development gives smaller metal particles and browner prints. The conservation summary from the AIC’s Photographic Materials Group puts the balance well: image contrast is achieved mostly by exposure rather than by the developer, although contrasting agents can be applied to the process.

Humidity is a hidden variable and a large one. Ware’s tests attribute the tonal reversal seen in some palladium prints principally to low relative humidity in the dried coating, below about 30 per cent, and reports the effect disappearing at 67 per cent on otherwise identical prints. A sensitiser mixed identically twice will not print identically twice if the room is different.

The chlorate bottle ages faster than anything else in the kit. Photographers’ Formulary states that the chlorate slowly decomposes over a period of weeks and the solution loses potency, and advises a test strip weekly. So the contrast grade of Solution 2 is not a fixed property of the bottle: a mixture counted out in the same numbers in March and in May are two different mixtures.

Clearing is not an afterthought and the two metals need different baths. The platinum kit’s citric acid is 30 g made up to 1500 mL and the palladium kit’s is the same 30 g made up to 2000 mL, with the reason stated: palladium metal is etched off the print by an acid strong enough to be routine for platinum. Ware reports the same asymmetry from the historical record at 1 : 60 against 1 : 200 hydrochloric acid, and notes that American workers who used the platinum strength on palladiotypes were partly rescued by a second error, because the stronger acid removed the palladium fog their oxalate developer had produced.

Image characteristics

Platinum is a matte, deep, rich yet warm black; palladium is warmer, yellower and browner. The AIC’s identification notes add that mixtures modulate between the two and that the mixture was not commonly employed by historic practitioners — the mixed Pt/Pd print of a modern portfolio is a modern object.

Both metals give a wider scale of grey tones and less contrast than silver, which is the property the process is admired for and also the property that created a market for a contrast agent.

What the chlorate does to the picture is not what “more contrast” usually means. This is the most important sentence on the page and it is Ware’s: the effect of these oxidising agents “is not to uniformly contract the tonal scale, but simply to truncate the high values”. A grade-3 paper redistributes every tone; a chlorate dose removes the weakest exposures altogether and leaves the rest where they were. What you get is separation in the mid-tones bought with empty paper in the highlights, which Ware calls a false sparkle of lost gradation and describes as aesthetically unacceptable for the platinotype ethos, where subtly nuanced high values are the point.

And it costs texture. Ware records an increase in image graininess and a deterioration in image quality. The Getty atlas records the same defect from the analytical side, and its observation is the one a conservator uses: the chlorate itself cannot be detected, because it is too soluble to leave a residue, but it is responsible for a visually detectable patchiness in the platinum image. A patchy platinum print with no chromium in its X-ray fluorescence spectrum is a chlorate print.

Maximum density is a known weakness of the traditional route. Ware attributes the practice of double-coating to it, and traces the underlying cause to the platinum salt’s solubility: the sensitiser cannot carry as much metal as the image would like.

The mechanism

The chemistry is in three acts, and the chlorate acts between the first and the second.

One: light makes an insoluble iron(II) salt

Ferric oxalate absorbs in the ultraviolet and an electron transfers from a coordinated oxalate to the iron, reducing iron(III) to iron(II) and destroying the ligand as carbon dioxide. In the dry coating the product is ferrous oxalate, which is almost insoluble — Ware gives 0.022 g per 100 cc of water.

UV + Fe₂(C₂O₄)₃ → 2 FeC₂O₄ + 2 CO₂

Ware's equation for the photolysis of solid iron(III) oxalate

The metal salt sits in the paper alongside it and nothing happens, because the reducing agent cannot move.

Two: the developer dissolves rather than reduces

This is the fact that makes the platinotype family unlike every other developing process in this course. The developer contains no developing agent. Saturated potassium oxalate supplies free oxalate ions, which take the immobile ferrous oxalate into solution as a mobile complex.

FeC₂O₄ + C₂O₄²⁻ → [Fe(C₂O₄)₂]²⁻

Oxalate ions dissolve the immobile photoproduct into a mobile complex

Now the iron(II) can reach the metal salt, and the reduction light had already prepared finishes itself. Two iron(II) per metal atom, because each metal(II) takes two electrons and each iron supplies one.

2 [Fe(C₂O₄)₂]²⁻ + [PtCl₄]²⁻ → 2 [Fe(C₂O₄)₂]⁻ + Pt + 4 Cl⁻

Ware's reduction of platinum(II) by the iron(II) oxalato complex

2 [Fe(C₂O₄)₂]²⁻ + [PdCl₄]²⁻ → 2 [Fe(C₂O₄)₂]⁻ + Pd + 4 Cl⁻

The same reaction with palladium, which is the one the course teaches

Ware’s redox potentials say why it goes: the iron(III)/iron(II) oxalato couple sits at +0.02 V, the tetrachloroplatinate couple at +0.73 V and the tetrachloropalladate at +0.62 V. Both metals are comfortably reducible; the citrato-iron couple of a cyanotype sensitiser, at +0.372 V, is not strong enough for either, which is why an iron-citrate process makes Prussian blue and gold and silver but never platinum.

Three: the chlorate removes part of act one before act two can use it

Potassium chlorate is a strong oxidising agent dissolved in the coating from the start. It does not touch the light-sensitive iron(III) and it does not touch the metal salt. What it does is re-oxidise some of the iron(II) that exposure has just produced, so that less of it survives to reduce the metal. Ware’s phrasing is exact: the effect is “to reoxidise some of the iron(II) photoproduct, thereby making it unavailable for reducing platinum(II) and so truncating the exposure scale of the process”.

⚗Why this is a subtractive control, and why the course will not write it as an equation

Two properties follow from the mechanism and neither is obvious from the phrase “contrast agent”.

It is a threshold, not a gradient. The chlorate is spread evenly through the coating and consumes iron(II) wherever it finds it. In a shadow that received a large exposure, plenty of iron(II) survives and the tone is barely touched; in a highlight that received little, the whole of it can be consumed and the tone disappears. That is what “truncating the exposure scale” means physically, and it is why Ware describes the result as removing the high values rather than redistributing the scale.

It competes with the developer for the same species. The iron(II) has two possible fates and the chlorate wins the ones the metal salt is slow to claim — which is part of why platinum, whose complex reacts more slowly than palladium’s, is the metal more damaged by the practice.

What this page does not do is write a balanced equation for it, and it does not because the chlorate page has already established that no source read gives the stoichiometry or says what the chlorate is reduced to in a ferric oxalate sensitiser at the pH it is used at. Abney’s 1905 account adds a second consequence specific to platinum — that some of the platinite is converted to a platinic salt — and the two accounts are consistent, but neither is a stoichiometry. An equation assembled from first principles would be an invention with a citation attached.

Why palladium needs twice as much

The palladium sheet gives 0.26 g against the platinum sheet’s 0.18 g and states the fact without explaining it; Bostick & Sullivan give the same comparison as “twice as much”. Ware’s chemistry supplies the explanation, and it is the same one that explains everything else about the pair.

The electron transfer proceeds not through the [MCl₄]²⁻ ion but through its aquated form, in which water has displaced one chloride. At 20 °C the equilibrium constant for that substitution is 0.17 for palladium against 0.015 for platinum, so at sensitiser strength more than half the palladium is already in the reactive form and less than a fifth of the platinum is. Palladium(II) complexes are more labile in any case, being those of a second-row transition metal rather than a third-row one.

So the palladium reduction is fast, and a fast reaction gives the chlorate less time to intercept the iron(II). More oxidant is needed to remove the same fraction. The same speed accounts for palladium’s lower native contrast, its greater tendency to fog in an oxalate developer, and — by the von Weimarn coefficient, which says faster precipitation gives smaller particles — its warmer colour.

Function of every ingredient

Ferric oxalate, 20 g to make 100 mL, in Solution 1 — the only light-sensitive substance in the system. Iron(III) held by oxalate ligands. Ultraviolet transfers an electron from a ligand to the metal, reducing it to iron(II) and destroying the ligand as carbon dioxide; the iron(II) is the reducing agent that will make the picture. More of it and the coating carries more reducing power than the metal can absorb, which wastes iron and increases the residual iron the clearing baths have to remove; less and the metal is under-reduced and the maximum density falls. It is also the least defined substance on the page: polymorphic, of variable hydration and variable purity, decomposing in solution over months, and — in Ware’s judgement — the root of most of what goes wrong with the traditional process. The kits supply it with a slight excess of oxalic acid, which keeps it soluble, converts any of the potassium double salt present to the active acidic form, and, as Ware found at around 2 per cent, slightly increases the amount of print-out during exposure by producing a little of the soluble trisoxalatoferrate(III) anion.

Solution 1 itself, 30 mL, as the ingredient of both versions of Solution 2. The formula’s own stock, drawn on to make another of its own solutions, which is how both sheets print it: you do not mix a second ferric oxalate solution, you take a bottle of the first. Recording it that way keeps one fact visible that a flat ingredient list would lose — that Solution 1 and Solution 2 are the same iron at the same strength, and that any difference in behaviour between them is the chlorate and nothing else.

Potassium chlorate, 0.18 g in 30 mL of Solution 1 for platinum and 0.26 g for palladium — the contrast agent, and the reason this page is at Level D. A strong oxidising agent that re-oxidises part of the iron(II) the exposure produced, so that less survives to reduce the metal. Because it is spread evenly and consumes iron(II) wherever it meets it, it removes the weakest exposures entirely while barely touching the strongest: the tonal scale is truncated rather than compressed. More of it and the highlights empty, the grain coarsens and the Getty atlas’s “patchiness” appears; less and the print flattens toward the sensitiser’s native exposure scale of about 2.0. None of it at all is the two-bottle sensitiser Willis used, which the platinum sheet itself prints as the mixture for a very contrasty negative. It decomposes in solution over weeks, so the same drop count gives a softer print as the bottle ages. Away from photography it is the classic laboratory oxygen source and a substance whose hazards belong to its mixtures rather than to itself; see its own page, and the Level D policy, for why the course teaches a calibrated negative instead.

Potassium tetrachloroplatinate(II), 20 g to make 100 mL, in Solution 3 — the image, in the metal this course will not let anyone handle. Potassium chloroplatinite, K₂PtCl₄, 415.1 g/mol. It is not light-sensitive; it waits in the paper until the developer sets the iron(II) free, then takes two electrons and precipitates as metallic platinum among the cellulose fibres, where the electron microscopy reported in the Platinomicon finds it as single ellipsoidal nanoparticles 15 to 25 nm across, distributed through the body of the surface fibres. More of it is barely possible: a 20 per cent solution is near saturation, which is precisely why printers resorted to double-coating to raise maximum density. Less and the image is weak, fibrous and grainy — Ware’s account of what happens when the wet processing washes the platinum salt out before it has finished reacting, which is also why period formulas reached for lead and mercury salts to hurry the reduction along. Its hazard is not acute toxicity but sensitisation, and that is what decides its classification.

Sodium tetrachloropalladate(II), 15 g to make 100 mL, in the palladium version of Solution 3 — the same job in the metal the course does teach. Na₂PdCl₄, 294.2 g/mol anhydrous, usually sold as roughly the trihydrate; the sheet gives the strength and not the hydrate, and the difference between the two forms is about a sixth of the palladium. Not light-sensitive either. It reduces faster than platinum for the reasons set out under The mechanism, giving a warmer, browner, lower-contrast image, a greater tendency to fog in an oxalate developer and a susceptibility to being etched off the print by acid that platinum does not share. More of it and the print deepens; less and it thins, on the same argument as platinum, but the economics are different enough that palladium is where most contemporary printing actually happens. The potassium salt is the same anion with a different cation and about a tenth less palladium per gram, and Ware’s own preparation makes the sodium salt from palladium(II) chloride dissolved in hot sodium chloride solution, which is the cheaper route to the identical solution.

Water, as the make-up. Distilled, and it is not a passive ingredient here: the sheets specify distilled water for rinsing brushes and glassware between coatings precisely because calcium is a problem in this family. Calcium strips oxalate from the iron complex and starts the hydrolysis that leaves the yellow iron stain; once it has gone as far as goethite the ordinary clearing baths will not take it out, and Ware records that only a more energetic reductant such as sodium dithionite may.

Interactions

Solution 1 against Solution 2 is the contrast control, and nothing else in the mixture moves. In all five published mixtures the metal holds at 0.521 or 0.522 of the volume. What changes is the split of the remaining 0.478 between the two iron bottles: all of it in bottle 2 for the softest negative, all of it in bottle 1 for the hardest. The iron total, the metal total and the coating volume are constant, so the only variable reaching the paper is the chlorate concentration. That is a well-designed control, and it is worth admiring even while declining to use it.

Chlorate against the metal’s reaction rate. The oxidant and the metal salt are competing for the same iron(II), so the dose that gives a particular contrast depends on how fast the metal reacts. This is not a separate fact from the doubled palladium dose; it is the same fact stated as an interaction. It also means that a mixed platinum-palladium sensitiser is not a simple average: Bostick & Sullivan’s instruction for their traditional kit is to use the palladium version of the chlorate bottle and then “use less #2 for the proportion of platinum solution”.

Excess oxalic acid against print-out and against reversal. Ware found that the customary 2 per cent of free oxalic acid converts about a tenth of the ferric oxalate to the soluble trisoxalatoferrate(III) anion. That anion’s photoproduct is soluble, so a little of the image forms during the exposure instead of waiting for the developer — and the partial print-out is one of the conditions for the tonal reversal that palladium prints sometimes show.

Oxalate developer against palladium highlights. Ware’s comparative tests show the oxalate developer reducing some palladium(II) on its own, fogging the high values brownish-grey, distinct from the yellow of an iron stain, and worse in humid conditions. Willis specified a citrate developer for the palladiotype and his distributor said why — the citrate baths do not injure the paper surface the way prolonged oxalate and acid do. A chlorate dose partly hides that fog, which is a real interaction and a poor reason to add an oxidant.

Hydrochloric acid against palladium. The clearing bath that suits platinum dissolves palladium. 1 : 60 for platinum and 1 : 200 for palladium in the historical record; a citric acid bath a third weaker for palladium than for platinum in the modern kits. Anderson recorded American printers using the platinum strength on palladiotypes and Ware’s reading is that the two errors — oxalate fog and over-strong acid — partly cancelled, which is why nobody noticed.

Gelatin against platinum. Ware records that gelatin sizing binds platinum(II) strongly and irreversibly and de-activates it. A paper choice is therefore a chemical variable, not only a surface one, and it is one of the reasons Willis avoided gelatin size in his own papers.

Metal against metal, in the container. Both suppliers insist on glass for the metal solutions, never metal vessels, because platinum and palladium plate out onto a more reactive metal surface. The same reasoning closes off stainless tanks and metal trays for the developer.

Variants

The ancestor: Pizzighelli and Hübl, 1882. Ware’s Table 5.1 records the shape the system began in — four stock solutions rather than three, with a platinum solution at about 15.6 per cent, an iron solution based on sodium trisoxalatoferrate(III), a Solution C that is potassium chlorate at 0.4 parts in 100 parts of the iron solution, and a Solution D containing mercury(II) chloride for brown tones. Ware prints the mixing ratios with the letters attached — “For black tones: A : B : C = 5 : 6 : 2. For brown tones: A : C : D = 5 : 4 : 4” — and the second of those omits B, which is either Pizzighelli’s intention or a slip in transmission; the course reproduces it as printed rather than repairing it. Ware’s own comment on Solution C is the position this course takes: it “serves only to increase the sensitizer contrast for negatives with an inadequate density range, and causes a deterioration in quality”, and the simplified black-tone mix without it is 5 : 8. The course’s separate entry for the 1886 formulations owns that material.

Irving Penn’s four bottles. Ware’s Table 4.1, compiled from the Irving Penn Foundation’s own formulary, records the practice of the twentieth century’s most famous platinum printer: a platinum solution at 16.7 per cent, a concentrated palladium at 20 per cent, an iron solution coded “2” at 21 per cent with 1 per cent oxalic acid, and an iron solution coded “2H” identical to it but with 1 per cent potassium chlorate. It is the same two-bottle switch, and the “H” stands for what you would expect. Penn’s 1 per cent is a stronger dose than either modern kit: the course’s own arithmetic on the Formulary’s figures puts its platinum bottle at 0.6 per cent w/v (0.18 g in 30 mL) and its palladium bottle at about 0.87 per cent, so Penn printed platinum at nearly twice the chlorate the platinum kit supplies. Those two percentages are the course’s conversion of a stated weight and a stated volume, not figures any sheet prints.

Bostick & Sullivan’s stronger iron. The same three-bottle system at 27 per cent instead of 20, with drop charts organised by negative size rather than by contrast grade, and with the chlorate weight unpublished. Their traditional kit adds a fourth bottle by letting platinum and palladium be mixed in one sensitiser. Their sheets are recorded here as corroboration of the system and their numbers are not transferable to the Formulary’s bottles.

The Na2 route, which replaces the chlorate bottle rather than filling it. Bostick & Sullivan’s current kits for digital negatives substitute a “sodium platinum” solution they call Na2, used at 5 per cent or 20 per cent. The sheets print no formula for it; Ware names the nineteenth-century alternative oxidant as sodium hexachloroplatinate(IV), Na₂PtCl₆, “still sold (at a very high price, of course) for this purpose today”, and the trade name follows that formula, so the course reads the two as the same substance and marks the identification as an inference. The instructions are explicit that it is used in place of both Ferric Oxalate #2 and the traditional platinum solution, and that it must not be combined with either. Their published drop counts for an 8 × 10 print are 20 drops of ferric oxalate #1, 15 of palladium #3 and 5 of the 5 per cent Na2, calibrated to the supplier’s own correction curves. It is a genuinely different design: the contrast agent has become a platinum(IV) salt rather than a chlorate. Ware records that the thermodynamics would allow hexachloroplatinate(IV) to be reduced — its couple sits at +0.68 V — but that the complex is far too inert kinetically to yield an image in the few minutes available, which is exactly why it can act as an oxidant without becoming image metal, and why Willis’s own platinum(IV) experiments of 1872 failed. Ware adds that it is said to cause less image deterioration than chlorate. It is still platinum, so this course does not open it either.

Dichromate in the developer instead of chlorate in the sensitiser. Willis & Clements recommended a very small amount of potassium dichromate in the oxalate developer for “brilliant prints” by palladiotype, and Ware records that Paul Strand and Ned Scott used it. It leaves this course by a route that has nothing to do with this page: the chromium policy rules chromium(VI) out at any level anywhere. Ware also notes it is not even stable in the bath, since dichromate is reduced by oxalic acid to oxalato-chromium(III) complexes.

Lead and mercury, which were contrast and colour controls before chlorate was either. Willis’s patents specify lead(II) nitrate and lead chloride to hasten the platinum reduction and smooth the tones, and mercury(II) chloride to shift the colour to sepia; the Getty atlas finds lead in early platinotypes and mercury in the sepia papers. Both are Level D and both have entries of their own — the lead contrast additions and the mercuric sepia platinotype.

And the variant that is not a variant of this at all. Ware’s print-out platino-palladiotype changes the iron salt rather than adding a bottle, and every objection to the drop system falls out with it. The course’s other siderotype sensitisers — the kallitype, Ware’s argyrotype — are the same first act with a different metal in the second.

Safety

Level D, on two independent grounds, and the page states both because a reader who knows only one of them will draw the wrong conclusion about the palladium half.

Platinum, for irreversibility. Potassium tetrachloroplatinate(II) carries H334 — may cause allergy or asthma symptoms or breathing difficulties if inhaled — in 99.5 per cent of 212 notified reports, with H301, H315, H317 and H318 alongside it. HSE’s EH40 sets halogeno-platinum compounds, defined in its paragraphs 25 and 26 as coordination compounds in which the platinum atom is directly bonded to halide ions, at 0.002 mg/m³ as platinum with the Sen notation, against 5 mg/m³ for platinum metal itself: a limit 2,500 times tighter for the compound than for the element. Ware’s own health warning is blunt and is the practical form of the same fact — never touch the surface of platinum-sensitised paper or put ungloved fingers in the processing solutions, dust inhalation is the worst hazard, and it is better not to store large amounts of dried sensitised paper. He also records where the hazard was first identified: platinum allergy was described in 1911 as an occupational disease of photographic factory workers handling platinotype paper. The Level D policy carries the argument for why a hazard whose consequence is permanent lands in the same classification as an acute one.

Potassium chlorate, for the availability of a better route. Its own page sets out the classification: signal word Danger, H271 as a strong oxidiser under the harmonised CLP entry, with the acute-toxicity statement genuinely disputed between notifiers. Its own conclusion is that the deciding operation is weighing the dry salt onto a bench carrying paper and dust, and that the contrast the reagent buys is available from a calibrated negative without the graininess and without an oxidiser in the darkroom.

What is not the reason, and it matters. Neither ruling is about ferric oxalate or about sodium tetrachloropalladate(II), which are Level B substances, or about the oxalate developer, which is Level B and is genuinely poisonous — an anticoagulant that both kit sheets single out for tongs and gloves, with EH40 listing oxalic acid at 1 mg/m³ long-term and 2 mg/m³ short-term, and the potassium oxalate page recording CAMEO’s account of systemic poisoning from inhalation of the dust. Those are the hazards of the palladiotype, they are controllable at Level B, and the palladiotype process page teaches them. Remove the chlorate and the platinum and what is left is a Level B process this course does teach.

EH40 lists no palladium compound at all, and its own introduction says that absence from the list does not indicate that a substance is without risk. Palladium being permitted is a judgement about controls, not an absence of hazard: the salt still carries H302, H317 and H318, and both aquatic statements.

Nothing above is a control and none is offered. The classification is not a statement about how difficult the controls would be; it is a statement that a page cannot assume them.

Storage

What the sources record.

The iron is the perishable one. Bostick & Sullivan give their ferric oxalate about a year, longer refrigerated, and say they mix it from powder just before shipping so that it ships fresh; they also insist it be brought back to room temperature before use, because a cold concentrated solution is not the solution it was when it was made. Ware records six to nine months in the dark as the figure “said by some”, and notes that at least one well-respected printer makes it fresh the night before every session. Both sheets keep it out of sunlight and intense ultraviolet, since it is the only light-sensitive thing in the kit.

The chlorate bottle is perishable in a way that is easy to miss, because nothing about it looks different. Photographers’ Formulary state that the chlorate slowly decomposes over weeks and the solution loses potency, so a bottle that has been on the shelf a month is a lower contrast grade than the one on the sheet.

The metal solutions keep almost indefinitely and their enemy is contamination. Bostick & Sullivan: they do not oxidise or age, refrigeration is unnecessary, and the two things to avoid are cross-contamination with ferric oxalate and any contact with metal, on which the noble metal plates out. Glass only, and a separate pipette for each bottle.

Coated paper is not stock. Ware records that sensitised platinotype paper does not keep unless it is thoroughly dried to inhibit the decomposition reactions, which is why the commercial papers were sold in sealed tins containing anhydrous calcium chloride as a desiccant. The kit sheets simply have the paper coated and printed the same day.

None of this is stocked in a home darkroom on this course’s account. The chlorate page says so for the oxidiser and the Level D policy says so for the platinum; this entry creates no exception.

Incompatibilities

Potassium chlorate against combustible material, which is the whole of its hazard profile. CAMEO’s entry describes a very flammable mixture with combustibles, explosive if the combustible is finely divided, ignitable by friction, fires or explosions with strong sulfuric acid, and spontaneous decomposition with ammonium salts. A darkroom bench carries paper, dust and cloth.

Ferric oxalate against the metal solutions, in the bottle. The suppliers’ repeated warning. The two are meant to meet in the coating vessel and nowhere else; a pipette used in both, or a brush washed carelessly, contaminates a bottle that costs more per millilitre than anything else in the darkroom.

Any oxalate against silver. The course’s oxalic acid page carries the incompatibility with silver compounds, and the reason is silver oxalate, which is explosive when dry. Every bench in this family carries oxalate, and a course that teaches both salted paper and the noble-metal processes has to keep the two sets of glassware apart absolutely.

Calcium against the iron complex. Hard water, a calcium-carbonate-buffered paper or a calcium carbonate alkaline reserve will strip oxalate from the iron and start that hydrolysis. Ware writes the first step of it out:

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

Ware: calcium precipitates as the oxalate and pushes the iron complex towards hydrolysis

This is why the sheets specify distilled water, and why paper choice in this family is a chemical decision as much as a surface one.

Gelatin against platinum(II), which it binds and de-activates.

Ammonium ions against hexachloroplatinate(IV), if the Na2 route is met in a manual: ammonium hexachloroplatinate(IV) is barely soluble and crystallises out of the sensitiser.

Alkali against the clearing baths. Ware’s chapter on clearing siderotypes is exact about the sequence: above about pH 4 the iron(III) hydrolyses to a polymeric colloidal hydroxide that lodges in the fibres, and while that hydroxide redissolves in dilute acid at first, if it is not removed at the wet processing stage it transforms irreversibly into iron(III) oxyhydroxide — the mineral goethite — which dilute acid will not touch. The yellow highlight stain of an old platinum print is that mineral.

See incompatibilities.

Waste

Nothing here is generated on this course, because the system is not carried out. What follows is what the chemistry implies, for a reader who has such solutions and needs to think about them.

Three separate problems, not one. A spent bath from this family carries iron and oxalate, a noble metal and, in the chlorate case, an oxidiser. They do not have the same answer.

The noble metal is worth recovering and is the reason a commercial platinum printer keeps spent developer and clearing baths rather than discarding them. The metal accumulates in the developer — both sheets note the colour darkening as it does — and the sludge that settles out and is filtered off with a coffee filter is not nothing.

An oxidiser is not neutralised by dilution, and the chlorate page makes the point that soaking it into absorbent paper to bin it creates precisely the mixture every hazard entry for the substance warns about.

Both metals carry the aquatic statements, though not unanimously: H400 and H410 — very toxic to aquatic life, with long lasting effects — appear in 76.7 per cent of the 176 notified reports for sodium tetrachloropalladate(II) and in 49.1 per cent of the 212 for potassium tetrachloroplatinate(II). Potassium chlorate carries H411 in 93.9 per cent of 262 reports.

Local regulation decides, and this course cannot tell you what it says where you are. See the disposal ruling.

Troubleshooting

There is no procedure to troubleshoot. What follows is what the sources record going wrong, kept because each entry explains a piece of the chemistry and because a reader working from a period manual or a kit sheet will meet all of it.

“The highlights are empty and the print looks brittle.” Too much chlorate, which is Abney’s own fault entry in 1905 and Ware’s aesthetic objection a century later. The tonal scale has been truncated rather than compressed: the weakest exposures were cancelled outright.

“The print is grainy and patchy, and there is no chromium in the analysis.” The Getty atlas’s diagnostic. Chlorate leaves no analysable residue because it is too soluble, so patchiness with no detectable contrast agent is the chlorate’s signature, where dichromate would show up under X-ray fluorescence a century later.

“The same drop count gives a flatter print than it did last month.” The chlorate in Solution 2 has decomposed. Photographers’ Formulary’s own remedy is a weekly test strip, which is an admission that the bottle’s contrast grade drifts.

“The highlights are yellow.” Residual iron: the print has not been cleared. Bostick & Sullivan state it directly — yellowing in the highlights means the clearing is incomplete — and Photographers’ Formulary warn that a print that is not properly etched will darken with age and can be destroyed. This is the failure that decides whether a print survives, and it is not a cosmetic one.

“The palladium print lost density in the clearing bath.” The acid was too strong. Palladium is etched by a citric acid bath a third stronger, or by hydrochloric acid at the 1 : 60 platinum dilution instead of 1 : 200.

“The palladium highlights are a brownish grey rather than paper white.” Chemical fog from an oxalate developer reducing palladium(II) on its own, worse in humid conditions, and distinct from the yellow of iron. Willis’s citrate developer leaves clear highlights.

“The shadows went backwards.” Tonal reversal, which Ware’s tests associate with a dried coating below about 30 per cent relative humidity, an oxalate developer, a contrasty negative and a little print-out. It happens far more with palladium than with platinum, and it does not happen at all in the print-out process.

“The platinum image is weak, fibrous and grainy.” Ware’s account: the platinum salt was washed out of the paper before it had finished reacting with the iron(II), because platinum reduction is slow. It is the problem lead and mercury salts were introduced to solve and the problem double-coating was introduced to work around.

“There is a red precipitate in the platinum bottle.” Normal, per the sheet: the commercial salt carries some potassium hexachloroplatinate(IV), which is much less soluble. It should not be carried into the sensitiser, where its complex is too inert kinetically to be reduced in the few minutes available.

“The coating is streaky or the image is splotchy.” Both sheets attribute splotchiness to paper that was not completely dry before exposure. Streaking in development is attributed to a print entering a hot developer unevenly.

“The bottle of ferric oxalate has gone dark, or the test says it is bad.” Both kits publish a spot test rather than a keeping date, and it is a nice piece of chemistry: a drop of the ferric oxalate into a little potassium ferricyanide solution should darken only slightly, because Prussian blue forms wherever iron(II) meets ferricyanide. A deep blue in the dark means the solution has already reduced itself; a deep blue forming on the side nearest a lamp within a minute means it is still photoactive. The same reaction that makes a cyanotype is being used as a quality control on a platinum sensitiser.

Experiments

No experiment below involves making or using this formula.

Rebuild the drop table from the parts column, and see the constraint. Take the five published mixtures and check, for each, that the metal holds at 0.521 or 0.522 while the two iron bottles trade places. Then work out the chlorate concentration in the final sensitiser at each grade, using the sheet’s own 0.18 g per 30 mL and 20 drops per millilitre. The five numbers you get are the actual contrast scale of the system, and seeing how small they are explains both why the control is delicate and why a month-old bottle behaves differently.

Test the substitution the course actually makes. The claim replacing the chlorate is that a correction curve on a digital negative delivers the same contrast without the cost. That claim is testable on any ultraviolet process — cyanotype is the cheapest — by printing one step tablet, measuring the exposure scale, building a curve to match it, and comparing the result against a print made from an uncorrected negative. If the corrected print holds its high values and the uncorrected one does not, you have reproduced the argument this page rests on.

Separate “truncating” from “compressing” on paper you already have. Print one negative on grade 2 and grade 4 silver gelatin paper and plot the two curves. Both are contrast changes and neither is what a chlorate dose does. Draw what a truncation would look like on the same axes — the toe cut off, the rest unmoved — and you will have the shape Ware objects to, without mixing anything.

Use the ferricyanide spot test as a redox lesson. The kit test for ferric oxalate quality is a Prussian blue reaction being used as an iron(II) detector. Run it on a cyanotype sensitiser instead: before exposure, after exposure, and after a day in the dark. It teaches the photochemistry of act one directly, and it is the same measurement a platinum printer makes on a bottle worth several hundred pounds.

Trace a formula’s paperwork. Three suppliers and one chemist give three strengths for the same bottle — 20, 25 and 27 per cent — and only one of them says what the hydrate is. Take any alternative process formula you can reach in two published versions and find the disagreement. Learning that a formula is a document with a provenance, rather than a fact, is the skill this formulary exists to teach, and this entry is one of its better examples.

Assess the classification yourself, which is Part XXVI’s own assignment. This page is unusual in having two independent Level D grounds with completely different logic: one about an irreversible health outcome that controls cannot recover from, one about a superseded technique for which a better route exists. Write the assessment for each, decide whether either alone would be enough, and then read Safety above to see whether you got there by the same route.