Part XXI Overview: Photography in Prussian Blue

Everything you have printed so far worked because a developer found a cluster of about four atoms of silver in a crystal of hundreds of millions and finished the job. Nothing in this part will do that for you. A cyanotype is made by light and by nothing else: every molecule of pigment in the finished picture is there because a photon arrived and was used, and when the light stops the image stops growing. That single fact sets the exposure times, the light source, the negative, the paper, the way you judge when to stop, and most of what goes wrong.

This part is where the course leaves the gelatin emulsion behind. Parts XXI to XXV are five parts of processes you coat yourself from named chemicals — cyanotype here, then salted paper, albumen, the iron-silver pair of Van Dyke and kallitype, and platinum-palladium — and the cyanotype is first among them because it teaches the whole pattern for a few pounds of chemicals and at the lowest hazard of the five. What you learn here about negatives, paper, ultraviolet and coating is not repeated in the four parts that follow. It is assumed.

What changes when the emulsion goes

Four things you have relied on since Part IV are simply absent, and each absence has a visible consequence.

There is no emulsion. No gelatin, no supercoat, no baryta, no crystals suspended in a binder. The sensitiser is a solution that soaks into cellulose, and the image substance is precipitated among and inside the paper fibres themselves. That is why an alternative-process print has no gloss to speak of, why its surface is the paper’s surface, and why — as the next section shows — the choice of paper is a chemical decision rather than an aesthetic one.

A cyanotype, in section

Drawn to show the arrangement, not the depth. Ware's equipment list specifies 100 per cent cotton or linen, sized but unbuffered; the 300 g/m² is the basis weight printed on a UK listing for one of the papers he names. Layer depths are drawn to be readable, not to scale: on real film the base is roughly a hundred times the emulsion, and drawn honestly the emulsion would vanish. Any thickness given in the labels is the real one.

There is no latent image, and nothing amplifies. In a silver gelatin material the photons make an invisible speck of a few atoms and the developer turns that speck into a whole reduced crystal. Here the photons make the product directly, molecule for molecule, and the process stops there. Ware puts the law behind it plainly: the Stark–Einstein law requires one absorbed photon for each molecule transformed, and no iron sensitiser in this cluster beats it by more than a factor of two.

🧮How much you gave up, in orders of magnitude

What a developer was doing for you. Part IV’s amplification arithmetic — the course’s own calculation, from Ware’s coating weights and Kodak’s grain sizes, with its assumptions written out — puts development at of order 10⁷ to 10⁸ atoms of image silver per atom of latent image.

What you get instead. Ware collates the measured quantum yields for the iron systems: about 1.2 iron(II) per absorbed photon for trisoxalatoferrate(III) between 250 and 420 nm, against an ideal maximum of 2 for its two-electron mechanism; a maximum of 0.45 at 365 nm for the citrate system of the classic cyanotype, falling to 0.28 at 436 nm; and about 0.01 for hexacyanoferrate(III) photolysing on its own, which is why Herschel’s first attempt took hours.

And what that costs in speed. Working from the Stark–Einstein law alone, Ware calculates the maximum possible sensitivity of any proto-photographic material, whatever its chemistry: about 34 J/m² for a just-perceptible image, in the near ultraviolet and blue, which on a photographic scale is a speed of about 10⁻⁵ ISO — one ten-millionth of an ordinary 100 ISO film.

One observation, offered as an observation. Ware’s speed ratio of about 10⁷ and Part IV’s amplification factor of 10⁷ to 10⁸ are the same order of magnitude, which is what you would expect if amplification were most of what separates the two families. The course notes the correspondence and does not lean on it: one figure is a ratio of threshold exposures between different materials measured in different ways, the other a ratio of atom counts, and Part IV’s page says in terms why the two are not the same quantity. Two numbers agreeing to an order of magnitude is a reason to look, not a proof.

There is no fixer, and no developer worth the name. Everything light did not use is soluble; the image substance is not. Water alone separates them, which is why Herschel — who had solved silver fixing in 1839 and knew exactly what residual thiosulfate does over decades — cared about this process at all. Where a bath in this cluster is called a developer, read the word carefully. The mineral acid bath of the New Cyanotype strips oxalate ligands off the iron(II) that light already made, so that it can build pigment; the hot oxalate bath of the platinotype dissolves an insoluble photoproduct so it can reach the metal salt sitting beside it. Neither reduces anything the light had not already reduced. Nothing in Parts XXI to XXV amplifies.

And there is no enlarger. At one ten-millionth of a film’s speed, projecting an image onto this material is not slow, it is impossible. Every print in this cluster is a contact print the size of its negative, made through Part XVI’s frame, and exposures run in minutes rather than in the handful of seconds Part XIX worked in. Ware’s arithmetic for the sun is the one to keep in your head: about 900 W/m² arrives, of which some 7 per cent is actinic, so an average ultraviolet irradiance of 30 to 40 W/m² gives a just-perceptible image in about a second and full density in 2 to 4 minutes. Under a lamp, Ware’s own figure for the classic sensitiser is 20 to 30 minutes.

Why cyanotype comes first

Not because it is the oldest — Herschel’s 1842 process is, but that is a reason to teach it in Part I, not to build a part around it. It comes first for four reasons, and the fourth is the one that matters most.

It is the cheapest chemistry in the cluster. Two solids, both cheap, and the whole part’s consumable chemistry costs less than a single box of the paper it is coated on. The figures are under Cost below.

It is the least hazardous of the five. The other four all bring in either silver nitrate, which is corrosive and stains everything it touches, or the soluble oxalates that the platinotype is built on and which are systemic toxins. Cyanotype brings in neither. Ware, a chartered chemist writing about his own field, ends his environmental chapter with the sentence “from all points of view, there is no safer photographic process”. The course does not adopt that as its own classification — see Safety below, where the reasons are set out — but it is a Tier 2 specialist’s considered judgement, it is about the right comparison, and it is not eccentric.

It is handled in room light. Bostick & Sullivan state it for their own kit: the solutions are sensitive to ultraviolet only, a safelight is not necessary, and ordinary incandescent lighting can be used throughout the process. This is the one part of the course where the safelight discipline of Part XVI simply does not apply. It is not daylight handling, and the distinction is worth being exact about: Ware’s instruction is to work under 40 to 60 W tungsten, not under fluorescent tubes and not near a window, and Bostick & Sullivan say to work in a windowless room or shade the windows. What you have escaped is the dark, not the sun.

And its mechanism is the pattern the rest of the cluster repeats. This is the real reason. Ware’s umbrella term for the family is siderotype, from the Greek for iron, and the shape of every one of them is the same two steps with a different second reagent.

The siderotype pattern: one photochemical step, then one ordinary chemical step

1. 1Near-ultraviolet light is absorbed by an iron(III) carboxylateNot by the iron alone. The photoactive species is iron(III) held by an organic acid — citrate in the classic cyanotype, oxalate in the modern ones — and the ligand is what supplies the electron. Iron(III) chloride on paper is not usefully light sensitive
2. 2An electron moves from the ligand to the iron; iron(II) is left and the ligand fragmentsWare writes the oxalate case as one equation and promises the reader no others: two iron(III) plus oxalate become two iron(II) plus two molecules of carbon dioxide. In the citrate case the organic photoproduct is identified as acetone dicarboxylic acid and the iron(II) species itself is still unknown
3. 3The iron(II) reduces or precipitates a second reagent that was never light sensitiveThis is the step that decides which process you are running, and every part from here to XXV changes only this line
4. 4The image substance is left exactly where the light fell; everything else washes outNo amplification, no fixer, and — in cyanotype alone — no clearing bath either, because the wash that develops the print also removes what would go on printing

2 Fe³⁺ + C₂O₄²⁻ + light → 2 Fe²⁺ + 2 CO₂

Ware's one equation for the whole family: light reduces the iron and the carboxylate pays for it with carbon dioxide

The second reagent	What it leaves	The process	Taught in
Hexacyanoferrate(III)	Prussian blue, a pigment	cyanotype	this part
A silver salt	Metallic silver	Van Dyke brown, kallitype	Part XXIV
A platinum or palladium chloro-complex	Platinum or palladium metal	platinotype, palladiotype	Part XXV
A gold(III) complex	Nanoparticle gold	chrysotype	studied in the atlas; no practical page

Cyanotype’s second step is the odd one out in a useful way: it is the only member of the family whose product is a coloured compound rather than a metal, and it is a plain electron transfer between two iron centres that needs no light at all.

Fe²⁺ + [Fe(CN)₆]³⁻ → Fe³⁺ + [Fe(CN)₆]⁴⁻

The oxidation states swap, and the iron(III) and hexacyanoferrate(II) then precipitate together as Prussian blue

Why the resulting pigment is so intensely blue — an intervalence charge transfer absorbing in the red, near 700 nm — and why the historical distinction between Prussian blue and Turnbull’s blue turns out to name one substance, are on the Prussian blue page and are developed properly in the chemistry lesson.

Three things every process here needs

Taught once, in this part, and assumed by the four that follow.

A negative whose density range matches the process. Ware’s rule is blunt and it inverts the habit of the last three parts: the contrast of the negative should always be adapted to the process, not the other way round. There is no contrast filter here and no grade to change: the exposure scale is settled when you choose the formulation, and the only lever left after coating is the development bath of the modern oxalate sensitisers, which the variants lesson deals with. Most of this cluster wants a long-scale negative: the New Cyanotype’s scale is about 2.2, the Simple cyanotype’s about 2.7, and Ware’s point is that one such negative serves New Cyanotype, salted paper, platino-palladiotype and argyrotype alike. The classic cyanotype is the exception, and it is the exception you meet first. Its scale is about 0.9 — three stops, comparable with a grade 3 enlarging paper — which means the negatives already in your file will print on it, and a negative made for platinum will not print on it at all. That inversion is deliberate: it lets you make a print in this part before you have solved the negative problem, and then shows you the wall.

An unbuffered cotton paper. Every conservation-grade paper you have been taught to prefer is the wrong paper here, and the reason is arithmetic rather than taste. Ware reads ISO 9706 and ISO 11108, the permanent-paper and archival-paper standards, as requiring an alkaline reserve equivalent to 0.4 moles of alkali per kilogram of paper, and calculates that a 250 g/m² sheet carrying a typical siderotype coating therefore holds an eightfold molar excess of alkali over the iron(III) you have just applied. Alkali hydrolyses the iron(III) sensitiser before it can work, and alkali destroys the finished Prussian blue too: a buffer at pH 9.4 decolourises it irreversibly in one to ten minutes, and pH 9.4 is what a saturated solution of calcium carbonate sits at. So the paper is unbuffered, 100 per cent cotton, sized; and the print does not go on a buffered mount at the end either, which is the single compatibility rule Part XX hands forward to this one. Where a buffered paper is unavoidable Ware’s remedy is a soak in 10 per cent sulfamic acid to destroy the chalk, and he calls it tedious, undesirable, and damaging to the sheet.

A controlled ultraviolet exposure. Part XVI owns this, and this part cites rather than repeats it. The ultraviolet sources and dose lesson establishes the band the sensitiser can actually use and the reason a lamp peaking at 410 nm is nearly wasted on a cyanotype: hexacyanoferrate absorbs there and acts as an internal filter, so the fraction of incident light reaching the photoactive iron falls from 0.65 at 365 nm to 0.03 at 420 nm. The UVA exposure unit is the enclosed, interlocked source with a measured uniformity map and a dose series, and the contact printing frame is what holds negative and paper in real contact — glazed in glass rather than acrylic, because acrylic cuts off at about 370 nm and would take the exposure with it. Its split back exists for this cluster: you open half the frame during the exposure and look at the print, which is how a printing-out process is judged.

The standard negative

One negative, chosen in this part, printed in every process you can reach, measured every time.

It is introduced in the negatives and papers lesson and it is the spine of the whole cluster. Brief section 31 asks for one negative printed across seven processes in one darkroom, and the printing processes matrix currently says plainly that nobody has done it: what is published there is assembled from different photographers, papers and decades, which tells you what the processes are like and not what they do to the same picture. Your standard negative is how that gets fixed. Each practical page in Parts XXI to XXV adds a row — measured exposure scale, maximum density, image colour, surface, cost, difficulty — and Part XXV’s comparison atlas assignment closes it.

Two consequences follow immediately. The subject has to survive being printed six or seven ways, which means a full tonal range and something in both ends of the scale rather than a striking effect. And because the classic cyanotype’s exposure scale is less than half of what platinum wants — Ware says it is quite unable to render a density range of 2 or 2.4 — the standard negative is made for the long scale and the classic cyanotype prints from something else. That is not a compromise, it is the argument of the part in miniature.

The map

Eight pages, and the order is the reasoning. Chemistry before materials, materials before coating, coating before measurement, and measurement before you are told there is a better formula — because the case for the modern sensitisers is only intelligible to somebody who has watched the classic one wash down the sink.

Page	What it settles	Level
Iron, light and Prussian blue	The photoreduction, the pigment, why it is blue, why it prints out and reverses, and the cyanide question answered properly	—
What these processes need	Exposure scale against density range, ultraviolet density, the paper, the sizing, the additives that ruin prints, and the standard negative	—
Lab: mixing, coating and printing a classic cyanotype	Sensitiser from named chemicals with a recorded batch, rod against brush, a measured dose, and a print	B
Experiment: exposure scale, maximum density and contrast	Your own figures for what the process can render, and which variable actually moves contrast	B
The New Cyanotype and the other ways	What is wrong with the classic sensitiser, what the oxalate formulations fix, and what each costs	—
Lab: toning the Prussian blue image	Destroying the image substance deliberately and rebuilding it in another colour, and what that costs in density	A
Assignment: at scale, on fabric, and the botanical sequence	Four projects that break the single-sheet habit	B
Break/fix: the cyanotype that went wrong	Six failures as symptoms, with the evidence to collect before the diagnosis	B

Behind all eight sits a formulary that is already further ahead than the lessons are. Seven cyanotype sensitisers are written in full, with provenance, mechanism and the function of every ingredient: Herschel’s 1842 formula, the classic sensitiser, the improved classic, the New Cyanotype, the Simple cyanotype, Mike’s cyanotype and the positive-working Pellet process — alongside the tannic acid toner, the two bought kits (Photographers’ Formulary, Bostick & Sullivan) and an encyclopaedia entry for every substance in any of them. No quantity is repeated in the lesson pages. A lesson explains why a formulation is shaped the way it is and sends you to the formulary for what to weigh, which is the division of labour the whole course is built on.

What you have already done, and what changes

Part I’s assignment had you lay pressed specimens on commercially pre-coated cyanotype paper, expose them in sunlight as a timed series, wash them in plain water and keep an exposure log — the course’s first permanent picture, made the way Anna Atkins made about five thousand of them from 1843. Nothing about that was pretend. But the coating was somebody else’s, at a strength you were not told, and the only variable you controlled was time.

Three things change here. You mix the sensitiser from named chemicals, with a strength you chose, a batch number recorded through the batch SOP and a formula version record so that a difference between two prints can be traced to something. You measure what it does — an exposure scale and a maximum density from a step tablet, using the densitometer of Part XV, rather than a judgement that the blue looks about right. And you control the light, with a dose in minutes at a known distance from a characterised source instead of an afternoon and a sky.

Safety framing

The part’s level, 3, describes difficulty. The safety letters are assigned page by page from the classification rubric, and they are not the same scale.

⚠The four rules that govern every practical page here

Hexacyanoferrate is a complex, not a cyanide salt — and it must still never meet acid or heat. The cyanide in potassium ferricyanide is bound to iron in a stable low-spin complex and is not free to act as a poison; the Environment Agency’s WM3 says the same in regulatory language by excluding ferro- and ferricyanides from the hazard statement for liberating a very toxic gas on contact with acid. Ware’s qualification of the frequently quoted warning is worth having exactly: the heating has to be very strong and the acid very concentrated to get much hydrogen cyanide, and neither is done in cyanotype, so the traditional process “will not emit any significant, perceptible or harmful amount” during normal exposure and processing. The rule the course keeps anyway is that no acid stop bath, no acid fixer, no vinegar and no bleach comes near this workflow or its waste bottle, because the cost of being wrong is asymmetric. The one exception is the deliberate, dilute, weak development bath of 1 per cent citric acid or less, which is the process as published.

Ultraviolet is the hazard people forget, because nothing tells you it is happening. A twenty-minute exposure in direct sun is a skin and eye exposure, and a lamp is worse because it is closer and does not feel warm. The controls, and the reason they run enclosure and interlock before eyewear, are Part XVI’s, and the outdoor exposure session SOP governs printing in the sun.

Never re-oxidise a print in a dichromate bath. It was once common practice, especially for fabric. Ware advises strongly against it, records that the consequences to one artist who used a dichromate bath for quilting fabrics are described in the medical literature, and notes that dichromate victims can become particularly sensitised to it. Ammonium dichromate is a chromium(VI) compound under the course’s chromium ruling, and it is the only substance in Ware’s own account of this work that he calls recognised as hazardous. Use air, or hydrogen peroxide.

The used baths are collected, not poured away on the assumption that dilute means safe. They carry unreacted iron salt, unreacted hexacyanoferrate and a colloidal dispersion of the pigment itself — the blue you can see leaving the print is image substance. Bottle it, label it through the labelling SOP, keep acid away from it, and follow the general chemical waste procedure and the course’s disposal ruling. Ware argues at length that treating this particular effluent as chemical waste is environmentally unnecessary, and his grounds are specific and well founded; the classic sensitiser page sets that argument out in full and explains why the course still does not publish a jurisdiction-specific instruction. Check your local regulations; they govern.

Two levels appear in the part and the difference between them is instructive. The toning lab is Level A because it works on a finished, washed print with no hexacyanoferrate present. The coating lab and the experiment are Level B, and the letter comes from potassium ferricyanide alone — ammonium iron(III) citrate is Level A, and so is Prussian blue, whose ECHA aggregation shows 86 per cent of notifiers declining to classify it as hazardous at all. Where a source and this course differ, the difference is stated rather than smoothed: Ware writes that he has no misgivings placing the traditional process in the hands of five-year-olds; the course takes each chemical’s own page as governing and the formula’s level from the highest of them, which lands at B.

☠What this part does not put in your hands

Two cyanotype formulations in the formulary carry no procedure at all, and the reason is Rule 5 rather than difficulty. Ware’s improved Classic sensitiser adds oxalic acid and a dichromate, and Ware’s own comment on it is that if you are prepared to go to that trouble you would do better with the New Cyanotype, and that it “should not be placed in the hands of children”. The New Cyanotype is not in the same position and the difference is worth stating: Ware marks its dichromate optional, so the course publishes that formula without one and classifies it B on its oxalate and its acid instead. Herschel’s hydrargyro-cyanotype applies a mercury(I) salt to a finished blue print; it exists to explain the chemistry and gives no method. Lead, thallium and nickel toners for cyanotype are real and historical, and they belong to Part XXVI at Level D under the Level D policy, which is the level at which this course gives no procedure to anybody.

What you need first

Part I’s Herschel lesson and its photogram assignment, for the history and for having held a finished cyanotype before you try to explain one.

Part XVI’s ultraviolet chain, in the order it was built: the sources and dose lesson for what dose is and why the band is narrow, the UVA unit or an honest substitute for it, and the contact frame glazed in glass. A reader who has none of these has the sunlight route below and should read the Part XVI lesson anyway, because it is what makes sunlight usable.

Part II’s laboratory discipline: concentration and dilution, measurement and its uncertainty, incompatibilities — which is what makes the acid rule above intelligible rather than superstitious — and the waste containers, to which this part adds one label.

Part XIII’s sensitometry, because the experiment measures an exposure scale and a maximum density and will not teach either from scratch, and Part XV’s densitometer, with the honest caveat that ultraviolet density is not visual density and an instrument built for one may not measure the other. The negatives lesson deals with that.

What you do not need is a darkroom. Mixing, coating, exposing, washing and drying all happen in a lit room, and the whole of the light discipline is to work under tungsten or warm LED, keep coated paper away from daylight and fluorescent tubes until it goes in the frame, and shade the window.

What you produce

Six things. Two are pictures, one is a material, and three are records — and it is the records that make the four parts after this one possible.

A mixed sensitiser with a recorded batch number, so that two prints made a month apart can be compared with something. A calibrated exposure — a dose in minutes at a stated distance from a characterised source, or a written sunlight record — replacing the guess you made in Part I. A measured characteristic curve for the process as you run it, giving your own figures for exposure scale and maximum density rather than a number quoted from a book about somebody else’s paper. A toned print, made by destroying the Prussian blue on purpose and building something else where it was, with the density cost measured. A project set: a large print, a botanical sequence, a paper negative and a piece of coated cloth. And one row of the process comparison atlas, which is the thing this part exists to start.

Cost

The band is ££, and the split is worth naming because it is the opposite of what beginners expect. The chemistry is almost free; the paper and the negatives are not.

Against one British retailer on 6 September 2026, potassium ferricyanide was £5.95 for 20 g and £15.95 for 100 g, and ferric ammonium citrate was £9.95 for 50 g and £24.95 for 250 g, both sold for this process and both inclusive of VAT. A 100 per cent cotton unbuffered rag paper from the list Ware names — Hahnemühle Platinum Rag, 300 gsm — was £23.95 for twenty-five 8 × 10 in sheets on the same day, and was shown out of stock, which is itself a fact worth planning around. Those are ranges to plan against, not quotations.

🧮What the two smallest jars actually buy — the course's own arithmetic

From the classic sensitiser’s published strengths and Ware’s stated coating volume, both on its formulary page, and the two prices above:

50 g ÷ 20 g per 100 mL = 250 mL

Solution A, from the 50 g jar

20 g ÷ 10 g per 100 mL = 200 mL

Solution B, from the 20 g jar

Equal volumes are mixed, so 200 mL of each gives 400 mL of sensitiser and leaves 50 mL of A unpaired. At Ware’s guide of about 1.5 mL for a 10 × 8 in print coated by rod, that is

400 mL ÷ 1.5 mL = about 265 prints

Sheets coated

for £15.90 of chemicals — roughly 6p of sensitiser per sheet, against about 96p for the sheet of paper it goes on. Brush coating can consume twice the solution and halves the count. The manufacturers’ own yields are the same order: Bostick & Sullivan claim 200 to 250 8 × 10 in prints from a 250 mL kit, and Ware gives 100 cc of Simple cyanotype sensitiser as sufficient for about 65 prints of that size.

What this arithmetic is and is not. It is the course’s own, from two sourced strengths, one sourced coating volume and two dated listings. It assumes rod coating, no waste, and that you use the paired volumes rather than the whole of both jars. Treat it as an order of magnitude for planning, which is all a per-print figure ever is.

!What the price file cannot yet tell you

The figures above are one retailer’s, read on 6 September 2026, and none of them is in src/data/prices.json. What the file does hold, from the survey of the day after, is a dated record for each of the two salts at a different pack size — ferric ammonium citrate and potassium ferricyanide, £17.99 for 230 g apiece — beside the cyanotype-paper entry for pre-coated sheets, which is Part I’s material rather than this part’s. So the planner can cost the chemistry of this part and not the rest of it, and each practical page still reports its consumables as a floor rather than a total. What is missing is a dated UK listing for unbuffered cotton rag paper in the sizes this cluster prints at, for transparency film for digital negatives, and for the ammonium iron(III) oxalate and iron(III) nitrate that the modern sensitisers need. What would settle it is a retailer’s listing stating pack size beside the sheet count, and one for a transparency film that names its ultraviolet transmission. The paper figure quoted above is the near miss: the file records it as a citation on this page rather than as a price of its own, and says it becomes a record the moment the address it was read from is supplied with it.

Alternative route

Sunlight is the honest alternative, and for cyanotype it is a good one. This is the part of the course where the reader without a facility is least disadvantaged, because the source that every supplier’s own instructions name first is the sky. Bostick & Sullivan list sunlight beside a 1000 W metal halide bulb and ultraviolet tubes without apology; Ware’s arithmetic gives average sunlight 30 to 40 W/m² of ultraviolet and full density in 2 to 4 minutes, and calls the prediction in agreement with all experience.

What you must supply in place of the unit is a written record — date, time, sky, orientation, elapsed time and the print’s appearance at each inspection — and the print-out judgement through a split-back frame, continuing until the shadows have reversed and the image looks blocked up. Part XVI’s alternative route argues this in full and this part does not re-argue it. Use the daylight exposure record and the exposure log; they exist for exactly this.

What you lose is repeatability, and you should not pretend otherwise. The experiment page is the one that suffers, because a measured characteristic curve needs a source that delivers the same irradiance twice. The honest substitute is to expose the whole step-tablet series in one sunlight session under an unchanging sky, so that the comparison between strips is valid even though the absolute dose is unknown, and to record the conditions rather than a dose. Say so in the analysis. A measurement whose limits are stated is a measurement; one that quietly claims a dose it did not have is not.

No route around the ultraviolet hazard. A reader printing outdoors has an unenclosed source of 30 to 40 W/m² that cannot be interlocked; what they have instead is distance, a hat, sleeves and the sense not to look at the sky. Those are administrative controls, the weakest kind, and knowing that is part of what the Part XVI lesson is for.

Where there is no wet space at all, four of the eight pages need no facility of any kind: the chemistry lesson, the negatives and papers lesson and the variants lesson are reading, arithmetic and paper-buying from end to end, and the break/fix page is diagnosis from recorded evidence. The formulary entries carry the mechanism at greater depth than the lessons do, and a reader who works through the classic sensitiser and the New Cyanotype pages side by side has had most of the argument of this part without opening a bottle. What has no substitute is the wash. Ware’s processing is twenty minutes in gently running water, or several changes of standing water, and it is both the development and the fixing; there is no version of this process that leaves it out, and no version that is not wet. If that is the obstacle, this part becomes a reading part, and the honest thing to say is that no page here can be run instead — which is why the section above puts as much work as possible into what can be done with paper, a pencil and the formulary.

The pages of this part

Part21 of 28Level3 — AdvancedPages8Estimated time15.2 hoursHighest safety levelLevel B

✱7 of 8 pages in this part need a darkroom, a UV source or mains-powered equipment, marked below. Each says what can be improvised and, where one exists, gives an alternative route.

0 / 8 lessons in this part completed

• LessonIron, Light and Prussian Blue✱ UV60 min
• LessonWhat These Processes Need: The Negative, the Paper and the Light✱ UV50 min
• LabLab: Mixing, Coating and Printing a Classic CyanotypeB✱ UV180 min
• ExperimentExperiment: The Exposure Scale, Maximum Density and Contrast of a CyanotypeB✱ UV150 min
• LessonThe New Cyanotype and the Other Ways to Make a Blue Print✱ UV50 min
• LabLab: Toning the Prussian Blue ImageA120 min
• AssignmentAssignment: Cyanotype at Scale, on Fabric, and the Botanical SequenceB✱ UV240 min
• Break/fixBreak/Fix: The Cyanotype That Went WrongB✱ UV60 min