The New Cyanotype and the Other Ways to Make a Blue Print

One substance is responsible for most of what is wrong with the print you made in the coating lab. It is not the ferricyanide, it is not the paper and it is not your coating. It is ammonium iron(III) citrate, and the case against it is not folklore: Mike Ware sets out seven numbered drawbacks of the classic process, six of them are properties of the sensitiser rather than of the process, and most of those six belong to that one jar. Every modern cyanotype formulation is an attempt to stop buying it.

This page is about those attempts. What Ware changed in 1995 and why the change had to be that particular one; what he changed again in 2019 when a better answer became possible; what each change costs you in preparation, money and care; why the resulting sensitisers must never be mixed or share a tray; and how to choose. No quantities appear here. The five sensitisers, with their weights, their provenance and the function of every ingredient, are in the formulary; this page is the argument those pages assume.

Seven things wrong with the classic process

Ware prefaces his list with John Tennant, editor of The Photo-Miniature, writing in 1906 that “the blue print process reached finality long ago as far as any important improvement is concerned, and there is little that is new to tell about it.” He was wrong, and the reason is worth stating as chemistry rather than as complaint.

Ware’s drawback	What it actually is	Where you meet it
1. The iron salt is ill-characterised	Iron content ranges from 14 to 28 per cent by weight across samples; no molecular formula can be written; remote practitioners cannot be sure they hold the same chemical	Two printers, one recipe, two different prints
2. It will not keep as a mixed solution	The two stocks must be stored apart, and the citrate stock is “an excellent nutrient medium for the growth of moulds”, covered within a week or two unless dosed with a fungicide such as thymol	A furred bottle in a fortnight
3. It is poorly absorbed by the paper	Some of it stays on the surface or in the coarse pores rather than inside the fibres, and being deliquescent it leaves a tacky coating that can damage a negative in humid weather	Image washing off; a negative that sticks
4. It is slow	20 to 30 minutes under a typical ultraviolet source, against about 2 to 4 minutes for palladiotype or the New cyanotype	Half an hour a sheet, in a teaching session or under a weak sky
5. Its pigment peptizes	The “soluble” potassium form disperses and washes out, causing “a serious loss of gradation in the high values, truncating the tonal scale, and conferring an artificially high contrast”	Blue wash water; empty highlights
6. It bleeds	Heavily exposed areas make surplus iron(II), which diffuses sideways and blues the highlight next to it	A soft blue halo round a dense shadow
7. Its exposure scale is about 0.9	Three stops, comparable with a grade 3 silver-gelatin paper, and quite unable to render the 2.0 to 2.4 density range the rest of the cluster’s negatives carry	One negative for cyanotype, another for everything else

Read the list twice and the structure appears. Drawbacks 1 to 6 are properties of the sensitiser, not of the process. Four of them — the variability, the mould, the poor absorption and the slowness — belong to the iron salt alone. The fifth belongs jointly to the potassium form of the pigment and to the poor absorption that fails to hold it, which is how Ware states it. The sixth is the ratio between the two solids, which cannot be got right while the iron content of one of them is unknown. Only the seventh looks like a property of the process, and it is not — the chemistry lesson establishes that the short scale is made in the wash, out of drawback 5. Change the iron salt and, in principle, all seven go.

⚗Why an 'ill-characterised' substance is a chemical problem and not just a purity problem

An impure chemical is a known compound with something else in it, and you can buy a purer grade. Ammonium iron(III) citrate is not that. Its manufacture — precipitate iron(III) hydroxide with ammonia, dissolve it in citric acid, neutralise, evaporate — starts from a highly polymeric, often colloidal solid, and the product inherits that indefiniteness. The iron content that results runs from 14 to 28 per cent by weight and the colour runs with it, light green to dark brown. There is no purer grade of an undefined thing.

The consequence for the bench is arithmetic. The classic sensitiser page has to reckon the iron-to-ferricyanide ratio as a range, about 1.7 to 2.1 moles of iron per mole of ferricyanide, because the numerator depends on a batch specification most suppliers do not print. You cannot calibrate against a number you are not given, which is why Ware records that the variability is worst for the people working hardest: those calibrating digital negatives to a densitometer.

The New Cyanotype: change the ligand, then throw the potassium away

The 1995 answer replaces the citrate with a crystalline oxalate: ammonium iron(III) oxalate, a compound of known formula, known molecular structure, reproducible composition and worldwide availability. Ware’s reasoning is compact — use “a light-sensitive ferric compound that is pure and reproducible, and essentially monomeric with molecules small enough to penetrate the cellulose structure” — and it answers the drawbacks in the order they were listed.

The drawback	The oxalate’s answer, in Ware’s terms
1. Ill-characterised	Well characterised, chemically consistent, keeps well in the dark, highly soluble
2. Mould and two bottles	No mould grows on it; the mixed sensitiser is a single bottle keeping for years
3. Poor absorption	Penetrates cellulose far more readily, mordants better onto fabric, is non-deliquescent, dries well, retains the image substance
4. Slow	More light-sensitive by a factor of 4 to 8 — two to three stops
5. Peptization	Yields an “ammonium Prussian blue” more resistant to peptization and to alkalies, with a maximum density verging on black
6. Bleeding	Can be mixed in chemically equivalent amounts with ferricyanide, so there is no surplus iron(II) to migrate
7. Short scale	Exposure scale about 2.2, which matches negatives made for salted paper, platino-palladiotype, argyrotype and chrysotype

Six of those follow from using a different molecule. The seventh follows from the fifth, which is the whole argument of this part in one line: the scale is long because the pigment stays in the paper.

And what results is a stronger solution as well as a better-behaved one, which is where the rest of the maximum density comes from. The formulary page’s comparison — the course’s own arithmetic, worked from Ware’s figures — puts the classic mixture at about 0.35 molar in iron and 0.15 molar in ferricyanide, and the finished New Cyanotype at about 0.40 and 0.30. The iron is much the same; the hexacyanoferrate is doubled. So the ratio of iron to ferricyanide falls from about 2.3 : 1 to about 1.3 : 1, which is essentially the 4 : 3 of Prussian blue itself — and that single number is remedy 6 stated arithmetically, because a sensitiser with no surplus iron has nothing left over to diffuse into the highlight beside it.

The difficulty the change created, and the trick that solved it

You cannot simply substitute one salt for the other. Ware is explicit that the early experimentalists found the obstacle: mix ammonium iron(III) oxalate with potassium ferricyanide at the concentrations a pictorial sensitiser needs — around 10 per cent w/v ferricyanide, which is 0.3 molar — and potassium iron(III) oxalate crystallises out, because it is sparingly soluble at 6.6 g per 100 cc of water. A gritty sensitiser prints small white spots, and crystals forming in the coating make what Ware describes as pretty and totally unwanted fern-like patterns.

The nineteenth-century blueprint industry lived with this by keeping the ferricyanide at 5 per cent or less, which avoided the crystals and gave a maximum density too low for any pictorial purpose. That is why commercial blueprint paper never looked like a photograph.

The obvious fix is ammonium ferricyanide, and you cannot buy it. Ware calls it a rare chemical and rather troublesome to make in pure form. So the 1994 procedure makes it in the beaker instead, by exploiting the very insolubility that caused the problem: mix the two salts hot and very concentrated, let the potassium iron(III) oxalate crystallise deliberately, filter it off and throw it away, and what is left in the flask is ammonium iron(III) oxalate plus ammonium ferricyanide in the proportions Prussian blue needs.

!This is why the New Cyanotype's ingredient list is not its composition

The formulary entry carries the weights, and it warns on its first line that they are a list of what goes into the beakers, not into the bottle: about 15 g of green solid leaves between the mixing and the make-up. If you have read a supplier’s kit list and wondered why the numbers do not add up to the strengths Ware quotes, that is why. Do not try to reproduce the sensitiser by dissolving those weights together in one vessel. You would get the crystals the procedure exists to remove, and you would get them in the coating.

The date. Ware developed the procedure in 1994 and published it as the New Cyanotype process of 1995. His own endnote gives the first appearance as “A New Blueprint for Cyanotypes” in Ag+ Photographic, volume 7, 1995, pages 74 to 80, with an independent account by John Barnier, “The New Cyanotype”, in Photo Techniques, volume 18 number 1, 1997, pages 12 to 15. The course cites those two references as Ware reports them and has read neither; what it has read is Cyanomicon, his own website article and his 2009 workshop notes, all three of which print the same formula. Under the course’s evidence rule that a formula published by the person who devised it is primary evidence for that formula, that is sufficient for the chemistry; it is not sufficient for a claim about what the 1995 article says, so no such claim is made.

What changes in the tray when the chemistry changes

A reader who has only printed the classic process tends to assume the wet work is the same for all of them. It is not, and one of the differences is a real hazard boundary rather than a preference.

The first bath is an acid, and it is doing chemistry

The classic process develops in plain water, or at most in 1 per cent w/v citric acid. The New Cyanotype wants a dilute mineral acid: nitric at 0.1 to 1 per cent v/v, hydrochloric at the same strengths, sulfamic acid at 1 to 5 per cent w/v, or citric acid at 1 to 5 per cent as a last resort. Acetic acid is specifically rejected, because it tends to form insoluble, strongly coloured basic iron(III) acetate.

The reason the oxalate sensitiser can take a stronger acid and the citrate ones cannot is the ligand. The chemistry lesson sets out the step: the iron(II) light makes is still wearing its carboxylate ligands and has to shed them before it can enter the Prussian blue lattice. For oxalate, hydrogen ions compete for the ligand to make undissociated oxalic acid and the equilibrium runs strongly to the right, so more acid means more free iron(II) means a deeper black. For the citrate sensitisers the same acid produces blue chemical fog in the highlights — free iron(II) appearing where you did not want it — which is why Ware’s limit for them is “a very weak acid”.

⚗One bath, two jobs, and the number that comes out of it

The acid bath is also the New Cyanotype’s contrast control, and it is the only one in the family with a usable range. Ware’s step-tablet figures: water alone gives an exposure range of about 2.6 density units, with the most open shadow gradation, a reduced maximum density and a curve shape he calls rather unusual; 1 per cent nitric acid gives about 2.1 with the deepest black; anything between is available by dilution.

(2.6 − 2.1) ÷ 0.3 = about 1.7 stops

The range, in stops — the course's own arithmetic from Ware's two figures

That is nearly two stops of exposure scale, chosen after the exposure, on a print you have already made. The classic process cannot do it at all — its contrast is decided by the negative and by nothing else — and the Simple Cyanotype’s tray-side version, described below, only goes one way and costs maximum density when it does.

The wash behaves differently, and so does the drying

Washing. For the classic process the twenty-minute wash is a compromise: it must clear the yellow sensitiser out of the highlights, and while it does so it carries a significant fraction of the image down the sink. For the New Cyanotype the same twenty minutes is nearly free — Ware records very little loss of image substance — so the end point is purely the disappearance of the yellow stain, checked by looking through the sheet against a bluish light. The practical consequence is that the print-out image you see in the frame is a usable preview of the finished print, which is not true of the classic process at all.

Humidity. The classic coating is deliquescent: in damp weather it goes tacky and can pull at a negative. The New Cyanotype’s coating is not, and Ware lists that among its advantages precisely because it reduces the likelihood of damaging a negative during contact printing. This is a genuine difference in convenience, and it is the one that will decide whether you can print at all in an unheated room in a wet autumn.

Heat-drying, and the second thing everyone tries. Ware measured both of the shortcuts people reach for, with Stouffer step tablets in duplicate, and both of them work — in the sense of moving the curve — and both cost something.

What you do	When you do it	What it does to the exposure scale	What it costs
Heat-dry the coated sheet, about 42 °C for five minutes	Between coating and exposure	Shortens it by about a stop, 2.3 to 2.0	May slightly coarsen the texture
“Darktime”: leave the exposed print in the dark before processing, 19 hours at 58 per cent RH and about 20 °C	Between exposure and processing	Lengthens it by about a stop, 2.3 to 2.6	Weakens the maximum density, flattens the shadows, causes edge fog with a rim at the tideline
Both	—	Returns to about 2.3, where you started	The lost maximum density, kept

Ware’s conclusion is a single sentence and the course does not soften it: “so it is best to do neither”. He allows one exception — darktime is worth resorting to if essential highlight detail would otherwise be lost. Note also that heat-drying the paper before coating was found to give no benefit at all, which disposes of a common piece of workshop advice.

⁇Whether heat-drying changes contrast, from one author's three printings

Ware’s 2009 workshop notes say the dryness of a cyanotype paper “does not appear to influence image colour or contrast”. Cyanomicon 7.3.6 says heat-drying “does not appear to influence image colour but it may increase contrast slightly”, and 7.3.12 then measures it at about one stop.

What this course uses and why. The later statement, because it is the one with a published step-tablet comparison behind it rather than an impression, and because it is the author revising his own earlier account after making the measurement. A reader working from the handout should know its sentence has been superseded by its own writer. The formulary entry records the same disagreement; it is repeated here because a reader deciding between processes will meet the handout before they meet the book.

Two later answers to the same question

The New Cyanotype fixes the chemistry and hands you a preparation that Ware himself calls somewhat challenging: dissolving by heating, a controlled fractional crystallisation, filtration, an overnight wait, and a yield variable enough that the sensitiser’s concentration differs from batch to batch. For a printer calibrating digital negatives, that variability is the whole problem coming back in a new form. In 2019 Ware answered it a second time.

The Simple Cyanotype, and the first cyanotype with a contrast control

The Simple Cyanotype does not buy an iron complex at all. It buys a simple salt of iron — iron(III) nitrate nonahydrate — dissolves it into citric acid that is already present in excess so that hydrolysis cannot start, and then neutralises with ammonia until the monomeric dicitratoferrate(III) anion is the main species in solution. That anion was isolated and its structure determined by X-ray diffraction in 1998, three years after the New Cyanotype; the insight of 2019 is that a photographer does not need the solid, only the solution.

Two things follow, and the second one is new in the whole history of the process.

The nitrate left behind is not a spectator. Ware credits it with four benefits, and marks the weaker two as expectations rather than findings. It should prevent iron(II) forming in the bottle during dark storage, so it acts as the preservative that dichromate is in the New Cyanotype. On development in mild citric acid it helps re-oxidise the solarised Prussian white in the shadows, which is why the reversed tones come back without peroxide. It is possible that the hygroscopic ammonium nitrate in the dried layer assists print-out by holding humidity in the paper. And it is likely that it is what inhibits mould. Read possible and likely as the words he chose.

The ammonia dose sets the contrast. This is the part with no precedent. The extent of the neutralisation, and therefore the pH of the finished sensitiser, sets the exposure scale, and Ware publishes three grades that differ in the volume of ammonia and in nothing else at all. His own step-tablet results, on Buxton 160 gsm with 0.3 per cent surfactant, an eight-and-a-half-minute exposure and half a minute in 1 per cent citric acid:

Grade	Sensitiser pH	Exposure scale	In stops
A, low contrast	about 4	about 2.7	nine
B, medium	about 6	about 2.4	eight
C, high contrast	about 8	about 1.8	six
Any of them developed in water alone	—	about 1.3	four

Intermediate values are obtained by mixing grades A and C proportionally, which is exact and repeatable in a way that a development-time adjustment never is. Ware states the claim plainly and the course records it as his: he believes this is the first cyanotype sensitiser with fully controllable contrast.

What it costs. Speed. The Simple Cyanotype is less than half the New Cyanotype’s, though still about four times a typical classic — 5 to 10 minutes under an average 365 nm source. Its maximum density is about 1.4 in Cyanomicon and about 1.5 in the 2022 instruction sheet, against 1.7 for the New Cyanotype. And the single-bottle version keeps about a month at room temperature, much longer refrigerated, with a two-bottle version published for anyone who needs a stock that lasts.

!One word of Ware's that this course does not adopt

Ware’s one-line summary of this formulation is “easy to make; cheap; contrast control; non-toxic”, and he writes elsewhere that the sensitiser contains neither dichromate nor oxalate. The adjective is not a hazard classification and this course does not adopt it on his word. The formula contains iron(III) nitrate nonahydrate, classified Danger as an oxidiser with skin, eye and respiratory statements, and concentrated ammonia, and its formulary entry carries the classifications and the controls. What is true and useful in his sentence is the comparison he is drawing: this sensitiser omits the two most hazardous substances in the other two formulations.

Mike’s Cyanotype: the same solution, without the ammonia bottle

Mike’s Cyanotype exists for one reason, and Ware states it: some workers cannot obtain, or cannot tolerate handling, strong ammonia solution. His answer is not a weaker formula but a different route to the same one. Instead of adding citric acid and then neutralising it with ammonia, you buy the citric acid with its ammonia already attached, as triammonium citrate, a bland white powder. Three ammonias per citric acid is exactly the composition of the tribasic salt, so the arithmetic closes; Ware’s stated intention is that the preparation produces, odourlessly, a sensitiser identical with Simple Cyanotype’s grade A.

What it costs is the contrast control. The neutralisation has already been done by the manufacturer, so there is no ammonia dose to vary and only one grade is published — the low-contrast one, exposure scale about 2.7. If matching a sensitiser to a measured negative is what you are doing, this is not the formula for it. And there is one shopping trap, set out at length on its formulary page: tribasic ammonium citrate is a different substance from the dibasic salt that a label reading only “ammonium citrate” may mean, and only the CAS number on the safety data sheet settles it.

They do not mix, and this is not a matter of taste

It is tempting, and it circulates, to mix New Cyanotype sensitiser with Classic or Simple. Ware calls it “a chemical nonsense”, and the reason is exactly the reason the New Cyanotype preparation exists.

What mixing a New Cyanotype sensitiser with a citrate one actually does

1. 1The preparation removes potassiumThe whole point of the fractional crystallisation is to take potassium ions out and leave ammonium ones in their place.
2. 2The classic sensitiser is full of potassiumIts ferricyanide is the potassium salt and nothing has been done to remove it.
3. 3Mixing puts back what was filtered outPotassium ions meet iron(III) oxalate again, at concentrations chosen to make a pictorial sensitiser.
4. 4Potassium iron(III) oxalate crystallisesSolubility 6.6 g per 100 cc. It forms in the bottle, in the syringe, and worst of all in the coating.
5. 5The print is wreckedSmall white spots typical of surface crystals; fern-like crystal patterns; brush marks where friction has promoted microcrystals.

Ware's account in Cyanomicon 6.8 and 7.2.4, with the solubility figure from his note 618. He adds that some workers get away with it depending on how far they dilute their solutions, which is why the failure is intermittent and therefore hard to diagnose.

There is a second incompatibility, and it is the one that catches people who never mix a bottle. The two chemistries want different development baths, and the baths are mutually hostile. The New Cyanotype needs a mineral acid for its best maximum density; the Simple and Classic sensitisers tolerate only a very weak acid, above which they fog blue in the highlights. A tray that has held a 1 per cent nitric bath and been rinsed rather than washed is a trap for the next citrate print through it.

⚠The equipment rule, and how much of it is sourced

What the sources establish. Ware’s two incompatibilities are between the sensitisers and between the development baths. Both are documented and both are his.

What the course adds, as an inference and labelled as one. A coating rod, a brush, a measuring syringe and a tray all carry residue, and the residue is the substance the incompatibility is about. The course’s rule is therefore one set of coating tools and one set of trays per chemistry, labelled and never crossed, and it is a precaution taken because the cost of being wrong — an intermittent, undiagnosable spotting or fogging fault — is much larger than the cost of a second brush. Ware’s observation that brush marks may themselves be microcrystals promoted by friction is the nearest thing to direct evidence for it, and it is not proof.

The one part of this that is a hazard rule rather than a quality rule is the acid. Concentrated acid never goes near a hexacyanoferrate, in any of these processes; the dilute development bath is a controlled use at one part in a hundred or less, and the chemistry lesson sets out exactly where that boundary is and why it is drawn stricter than the evidence strictly requires.

Ware also has something to say about the practice of mixing two established formulae and naming the result after the mixer, which is worth reading in full in Cyanomicon and can be summarised here without losing its point: artists may do as they please, but it has yet to be demonstrated with step tablets, not pictures that New cyanotype is improved by mixing it with Classic. That is the standard this course applies to any claim of that shape, including its own.

Contrast: four places you can set it, and one the course will not use

Gather what is now on the table. A cyanotype’s contrast can be set in four places, and they are not equivalent.

1. In the negative. Always available, always the first answer, and the subject of the negatives and papers lesson. For the classic process it is the only answer.
2. In the sensitiser, at mixing. Only the Simple Cyanotype offers this as a variable: 2.7 down to 1.8 by the ammonia dose, with any intermediate value obtained by mixing two grades proportionally. Mike’s Cyanotype reaches the same solution by another route but publishes one grade only.
3. In the development bath. Only the New Cyanotype offers a usable range, 2.6 in water down to about 2.1 in 1 per cent nitric acid. The Simple Cyanotype has a one-way version of this — water instead of citric acid drops it to about 1.3 — which Ware treats as a last resort because it costs maximum density.
4. In the drying and the darktime, about a stop each way, at the costs tabulated above and against the author’s own advice.

☠The fifth lever: dichromate, and why this course will not print a procedure for it

What it does. Ammonium dichromate at 0.1 g per 100 cc is part of Ware’s published New Cyanotype and of the commercial kits. Its stated purpose is preservation — it holds the bottle at an oxidising potential so that traces of reducing impurity cannot start making Prussian blue in the stock, extending the shelf life from months to several years — and Ware records a second effect: it “will also tend to increase the contrast of the sensitizer significantly”. A potassium dichromate bath was also once routine in the wet processing of fabric cyanotypes, for reoxidation and reputedly for light-fastness.

Why the course gives no procedure, at any level. The chromium policy is the single place that rule lives and it is not decided on this page: chromium(VI) is never used anywhere in this course. Ammonium dichromate carries a harmonised CLP classification, signal word Danger, with H340 for genetic defects, H350 for cancer and H372 for organ damage on repeated exposure each reported by 100 per cent of notifiers to the ECHA inventory and not one notifier declining to classify it. Potassium dichromate’s record is the same shape across twelve statements. HSE’s EH40 sets chromium(VI) compounds at 0.01 mg/m³ as chromium with Carc, Sen and a biological monitoring value attached — which is to say the recognised way of knowing whether the control worked is to test the worker’s urine. That is not a control a home darkroom has.

This is not the course editing a formula behind your back. Ware marks the dichromate optional in three publications, writes that it “may be omitted to satisfy prevailing safety regulations” and states that it “can be entirely omitted”. The characteristic curve he publishes for the acid-development contrast control is captioned, in his own words, as being for New Cyanotype without dichromate. And his strongest safety statement in the whole chapter is his own: never use a dichromate bath to re-oxidise a cyanotype, a practice he calls particularly inadvisable on fabrics that touch skin, citing a well-documented case history of an artist made ill by it.

What you give up, and what you get instead. You give up the long shelf life — months rather than years — and some contrast. You get the contrast back in the development bath, where this process happens to have the family’s only development-stage control with a usable range, and you get the shelf life back by mixing less at a time. If you buy a New Cyanotype kit rather than mixing one, expect the dichromate to be in it: the Photographers’ Formulary kit supplies it as 10 mL of a 1.25 per cent solution and its own safety notes single it out.

The wetting agent, and what it is actually doing

Every one of these formulations offers a surfactant, and it is worth understanding rather than adopting, because the same additive is essential in one formulation and usually pointless in another for a reason that is entirely mechanical.

Where the sensitiser ends up, and what the wash can reach

1. 1Sensitiser in the coarse pores and on the surface — where a poorly absorbed, polymeric, large-molecule sensitiser stops; the pigment forms loose and the wash takes it
2. 2Sensitiser inside the interfibrillar space of the fibre — where the pigment is trapped as it forms; Ware states that success depends on reaching here
3. 3The wash — the same twenty minutes in both cases; what differs is how much image it can reach

Drawn to show the argument. The fibre dimensions are not to scale and no source read for this course publishes how deep the sensitiser actually penetrates a given sheet.

Ware states the requirement in one sentence: the success of cyanotype depends on the sensitiser penetrating the interfibrillar space of the surface cellulose fibres, where the pigment will be trapped, and not simply remaining in the coarse pores of the paper, from which it washes out. A surfactant is a molecule with a water-liking end and a water-avoiding end. Put enough of them at the liquid’s surface and the surface tension falls, so the drop that was sitting on top of a sized fibre spreads and enters it instead. That is the whole mechanism, and everything else follows from it.

• The classic citrate sensitiser benefits most from one, because its molecule is large, polymeric and poorly absorbed; the surfactant is the remedy Ware names for drawback 3, and in his own classic formula he calls it the only innovation he has added to Herschel’s, one that can greatly improve the final print on many papers.
• The Simple Cyanotype uses one routinely, at about 0.25 to 0.5 per cent in the sensitiser about to be coated.
• The New Cyanotype usually does not, and this is the interesting case: Ware writes that a wetting agent is “usually quite unnecessary” because the oxalate solution is much more readily absorbed by paper fibres than the citrate. It earns its place only on a few hard-sized papers such as Buxton, at about 0.1 to 0.25 per cent, where it improves both the ease of coating and the retention of the pigment.

Four practical rules come with it, all Ware’s. It never goes into the stock bottle — the right amount depends on the paper and has to be found by trial, and the dilute stock solution does not keep, so it is replaced if it shows a deposit. It may interact unfavourably with gelatin-sized papers. The specific surfactant he uses is the non-ionic polysorbate sold as Tween 20, and its sibling Tween 80 is explicitly rejected because it foams. This course has no chemical encyclopaedia entry for either, which is why the class rather than the brand is what is taught here; the brand is recorded because it is what the source specifies and what a reader will be sold.

Pellet’s process, and what makes a positive possible at all

Everything so far makes the blue where the light fell. One historical process makes it where the light did not, and understanding why is the cleanest lesson in the family about what a developer actually decides.

The classic cyanotype contains potassium ferricyanide — hexacyanoferrate(III), the red prussiate — and that reagent gives its blue with the iron(II) light has made. Pellet’s paper contains no ferricyanide at all. It goes after exposure into a bath of potassium ferrocyanide — hexacyanoferrate(II), the yellow prussiate — and that reagent gives its blue with iron(III), the iron light has not touched. With iron(II) it gives Prussian white, which is colourless.

Fe³⁺ + [Fe(CN)₆]⁴⁻ → Fe[Fe(CN)₆]⁻

Pellet's development: the surviving iron(III) meets hexacyanoferrate(II) and gives the same Prussian blue by the other route

So the polarity is not a property of the sensitiser. It is set by which oxidation state of hexacyanoferrate you develop in. That is the whole of the difference, and a student who has coated both papers has learned something no amount of reading about “positive-working processes” conveys.

What makes it work rather than merely happen is the second ingredient. Henri Pellet, a carbohydrate chemist, patented the process in Britain in 1877, in France the same year and in the United States in 1881, under the names Cyanofer and Papier Gommoferrique. His addition was gum arabic with iron(III) chloride, and it rests on Alphonse Poitevin’s observation of 1863 that iron(III) salts harden gum and make it insoluble in water while iron(II) salts do not. In the exposed regions the iron is reduced, the gum re-dissolves, and the colourless Prussian white that forms there can simply be wiped away — which fixes the image without much disturbing the blue in the unexposed regions. Herschel had invented a positive-working cyanotype in 1842 and never made it reproducible; the gum is the answer to the problem he could not solve.

🕰Why the drawing offices could not make it work, and how a conservator tells one today

Pellet’s prints were exhibited at the 1878 International Exhibition in Paris and, on Ware’s account, excited J. M. Eder’s admiration. His patents, though, are written in what Ware calls “catch-all” language, deliberately: solutions of many types and compositions are cited as examples, and the development bath is specified as “concentrated, or dilute, neutral, slightly acid, or slightly alkaline, hot, or cold.” Ernst Lietze, having transcribed it, records that he tried the process several times “without succeeding in producing good prints. The background was always more or less colored.” The formulary entry records that every source it read which tried the process reports the same trouble with the whites, and by 1943 it was deemed defunct. The quantities on that page are therefore not Pellet’s at all, but Pizzighelli and von Itterheim’s of 1881.

Identifying one. The Getty Conservation Institute’s atlas states that Pellet prints are indistinguishable from cyanotypes by microscopy and by X-ray fluorescence, since the image substance is the same Prussian blue. The criterion that works by eye is Ware’s: a Pellet print carries a deposit of insolubilised gum holding the pigment, so it shows a binder layer in relief in the shadows and looks more like a blue Woodburytype than like a cyanotype, whose pigment sits in the paper fibres with no layer at all.

Cyanotype off paper

The sensitiser is a solution of two salts and it will go wherever a solution goes. Three destinations have real literature behind them and one has a definite answer that beginners keep asking for.

Fabric and textiles. This is the substrate the iron processes are best suited to, and the reason is the absence of a binder: no gelatin, no gum, nothing to stiffen the cloth or spoil its drape. The chemistry was in the dye trade before it was in photography — nineteenth-century practice mordanted cloth in a bath of an iron(III) salt, rinsed, then formed the pigment in place in an acidic ferrocyanide bath. The one finding a printer can act on is mordanting efficiency: Ware records the textile artist Wendy Wilson’s observation that mordanting to cotton and silk fibres appears to be better with the oxalate sensitiser than with the citrate, which is the same molecular-size argument as the paper case, in another material. That is a named practitioner’s reported experience and not a measurement the course has verified. The dichromate bath that used to be part of fabric processing is dealt with above, and this is the application in which Ware is most emphatic against it: a cloth that will be handled and worn is the worst place in the whole process to put a sensitising, allergenic carcinogen.

Ceramics. Two facts settle this one. Prussian blue decomposes thermally at around 200 °C with the evolution of cyanogen and hydrogen cyanide, so it cannot be fired under a glaze — you would destroy the image, leaving only a weak brown iron-oxide picture, and generate two toxic gases doing it. But the sensitiser is readily imbibed by an unglazed, bisque-fired tile body, whose porosity has to be restricted first with a sizing agent such as gelatin, and Studio Glithero in London have made botanical photograms on tiles and vases this way, sealing the surface afterwards. The prohibition that catches people is the one that catches them on paper too: calcium carbonate in gesso is an alkali, and alkali destroys Prussian blue. It is also, Ware notes, why the pigment was never used for fresco.

Diapositives on glass and film. The nineteenth-century route was to coat a plate with gelatin, or to fix the silver out of an old fogged lantern plate and harden what remained, then soak the gelatin in sensitiser. The modern route is easier and it is one Ware has used himself: the transparency films sold for making digital negatives take aqueous iron sensitisers very well, the coating holds the image nanoparticles without bleeding, and — because the film contains no alkaline buffer — it is used out of the box with no pre-treatment at all. The one difficulty is laying the coat down evenly: a glass rod suffers intolerable friction on that surface, and a squeegee, a foam brush or a PTFE rod is what works.

🧮Cyanotype negatives made in camera, and the pinhole question answered once

The arithmetic. Ware rearranges the Jones-Condit camera-image equation and reduces the result to one usable ratio: an in-camera exposure is about 5 A² times the contact exposure, where A is the f-number, and four times that again at 1:1 reproduction. He calls it order-of-magnitude, good to a factor of two or three either way.

f-number	2.8	4	5.6	8	11	16
Times the contact exposure	40	80	160	320	640	1280

With a lens it is possible and tedious. Ware reports that a camera image can just be obtained on New Cyanotype at f/4 with a minimum of an hour of sun, where the traditional formula takes a day, and publishes a 6 × 9 cm negative made on transparency film at f/3.5 with a five-hour exposure.

With a pinhole it is not. A pinhole camera’s f-number is the focal length over the hole diameter, which for most is between 100 and 1000. Take 200: the exposure runs to (200/4)² = 2500 hours of sun, about a year of eight-hour sunny days — and that assumes reciprocity holds, which over such a period it will not, and that the coating does not chemically fog first, which it will. Ware’s conclusion is one sentence: “it won’t work!” The only image you might get is a streak from the sun’s own disc.

And the negatives are poor even when you get them. A cyanotype negative is blue, so it has low optical density in the ultraviolet that the alternative processes actually print by, and the positives made from it are dingy for that reason. Its tonal rendering is also unbalanced, because the sensitiser responds only to ultraviolet and blue: green foliage does not register at all. The practical answer found in 2002 by John Beaver is to scan the cyanotype negative and invert it digitally rather than print it.

And no enlarger. Projection printing fails for the ordinary reason — the lamp has no ultraviolet and the condensers and lens attenuate what little there is — and the arithmetic in the chemistry lesson shows why no amount of lamp power fixes it. Ware allows one exception in principle: the kind of solar enlarger once used commercially for enlarged platinotypes.

The variants this course studies and does not reproduce

Three additions to the cyanotype sensitiser have real literature and no place in a home darkroom. They are named here so that a reader meeting them can identify them, and the chemistry, the history and the toxicology belong to Part XXV’s collected page, which is where the additives inside processes this course does print are gathered.

Uranium(VI) in the sensitiser. Uranyl salts were claimed to enhance sensitivity by catalysing the photoreduction of iron(III). Ware tested the claim himself and reports the result with unusual precision: uranyl nitrate at 1 to 2 per cent added to either an oxalate or a citrate sensitiser does increase printing speed, by about one stop — a factor of two — and further increases in concentration bring little additional advantage. His conclusion is the one this course adopts: its use is not worth the toxic risk. One stop is what a slightly stronger lamp gives you for nothing.

Vanadium pentoxide, for which a fourfold speed increase was claimed in a US patent of 1924. The course has not read that patent and states no more about it than that Ware records it.

The uranotype, sometimes miscalled “uranium cyanotype”. Here the sensitiser is a uranium(VI) salt entirely — no iron at all — reduced by light to uranium(IV), and developed in ferricyanide to give a red image in uranyl ferrocyanide. It was originated by Burnett. It is not a cyanotype and the name is a serious misnomer, as Ware says; there is no Prussian blue in it.

And the mercury one. Herschel’s hydrargyro-cyanotype is not a sensitiser variant but a post-treatment: a finished cyanotype washed with a mercury(I) salt, bleached blank, then brought back in brown by heat. It is Level D for the mercury and for the heating, and its formulary entry is worth reading for a different reason entirely — it is the clearest demonstration anywhere in early photography that an image is a substance and not a picture.

⚗Two contemporary systems that are cyanotypes in the pigment only

Ware records that until 2017 every cyanotype chemistry rested on the light-sensitivity of an iron(III) polycarboxylate. Two do not, and both are early work rather than established practice.

Blythtype, from Jeff Blyth and Martin Richardson at De Montfort University, uses the dye thionine — Lauth’s violet — as its ferricyanide salt. Light bleaches the dye to a colourless reducing form, which reduces the ferricyanide; the print is then developed in an acidic iron(II) sulfate bath, so the blue forms where less light fell. Its three claimed advantages are worth knowing because they are exactly the three things a cyanotype cannot do: it is positive-working so needs no negative, it is sensitive to visible light out to the red end, and it can therefore be printed by projection from a digital projector or a slide projector.

A “mixed sandwich” iron complex. A group of Russian workers prepared the ferricyanide salt of the benzene(cyclopentadienyl)iron(II) cation and found it sensitive to visible light, yielding Prussian blue after ten minutes under a 60 W incandescent bulb. Ware notes that only a simple image of writing was published, with no step-tablet test of the tonal gradation, and that the starting complex is exotic and not commercially available.

Neither is offered here as a procedure and neither has a formulary entry, because neither is a formula this course can source at the standard the rest of the part is written to. They are recorded because “cyanotype means an iron sensitiser” is a sentence that stopped being universally true in 2017.

Choosing

Here is the whole decision, and the first question is not about chemistry.

Which sensitiser, in the order the questions actually arise

1. 1What is the ultraviolet density range of the negatives you have?About 0.9 to 1.2 — the range of a negative made for grade 2 or 3 enlarging paper — points at the classic process, which was built for it. 1.8 and above points at everything else. Measure before you shop.
2. 2Can you get the chemicals where you live?The two classic salts are sold by photographic retailers as named products. The oxalate and the nitrate are laboratory or specialist purchases, and in one dated UK check neither was listed at all.
3. 3What can you actually do in your workspace?Heating to 70 °C, filtration and an overnight crystallisation for the New Cyanotype; concentrated ammonia for the Simple; neither for the classic or for Mike’s.
4. 4How deep does the black have to be?Maximum density about 1.7 for the New Cyanotype, about 1.4 to 1.5 for the Simple, and for the classic whatever survives the wash.
5. 5How long can you stand in front of the frame?20 to 30 minutes a sheet for the classic, 5 to 10 for the Simple, 2 to 5 for the New Cyanotype under a typical source.
6. 6Do you need to choose the contrast, or only to live with it?Simple Cyanotype sets it at mixing; New Cyanotype sets it in the development bath; the classic sets it in the negative and nowhere else.

The figures are Ware's and are collected with their sources on the five formulary pages. The availability observation is a single dated retailer check and is described as such below.

Ware’s own one-line summaries are the most honest short comparison in the literature, because he wrote three of the five formulas himself and still names their faults first:

Formulation	His summary
Classic cyanotype	easy to prepare; uncertain; limited tonal range
Classic improved	more toxic; better range and density
New cyanotype	faster; finer; higher Dmax; toxic; chemically tricky
Simple cyanotype	easy to make; cheap; contrast control
Mike’s cyanotype	Simple cyanotype without using ammonia

The fourth row omits one word of his, and the omission is deliberate: the reason is in the callout above.

!Cost and availability, with the part of it the course could not establish

What is sourced. At one British photographic retailer, checked on 6 September 2026, potassium ferricyanide was £5.95 for 20 g and £15.95 for 100 g, and ferric ammonium citrate £9.95 for 50 g and £24.95 for 250 g — the two classic chemicals, listed as their own products under their process name. A search of the same shop for “ferric” on the same day returned those four products and nothing else: no standalone ammonium ferric oxalate and no ferric nitrate, the oxalate appearing only inside the description of a bought cyanotype kit. Ware’s own comparison is that the New Cyanotype is more expensive in chemicals than the classic, and the Simple Cyanotype cheap.

What the course could not establish. The manifest asked this page to compare the current UK price of ammonium iron(III) oxalate against ammonium iron(III) citrate. It could not, because the one dated retailer check available did not list the oxalate at all, and the course’s price file prices neither salt — a gap its own notes already record. So the comparison below is availability, not price, and a reader in another country may find the position reversed.

Availability, which for most readers is the real constraint. The classic salts are photographic-shop chemicals. The oxalate and the nitrate are laboratory or specialist-supplier purchases, and the New Cyanotype is also sold ready-made: as a kit whose contents include the dichromate, and as a premixed single solution whose composition, exposure time and maximum density its seller does not publish at all. Buying the premix means buying a sensitiser you cannot characterise, which is a real trade and not merely a lazy one.

And the answer for most readers is still the classic process, which is worth saying plainly on a page that has spent two thousand words on its faults. It is two solids and two bottles. It needs no heat, no filtration, no concentrated acid, no ammonia and no oxidiser. Ware — who devised three of the four alternatives — writes that its chemicals are not dangerous and can be handled by children under supervision. He says something comparable about only one other formulation in the chapter, the Simple Cyanotype, and that adjective of his is not one this course adopts either, for the reason given above. The classic process’s faults are real, and you have measured two of them — the exposure scale and the maximum density — in the exposure-scale experiment. Move on when you have a reason: a negative that will not fit in three stops, a black that is not black enough, a class that cannot stand round a frame for half an hour, or a calibration that has to repeat. Any of those is a reason. “Because it is newer” is not one.

Alternative route

This page declares that it needs an ultraviolet source, and everything it recommends you do with one is a comparison rather than a procedure: print the same negative on two sensitisers and look at what the highlights do.

Sunlight is the honest alternative, and for this comparison it is a good one. Ware’s arithmetic gives average sunshine 30 to 40 W/m² of ultraviolet, and his stated exposure times — 20 to 30 minutes for the classic, 5 to 10 for the Simple, 2 to 5 for the New Cyanotype — are given for “a typical ultraviolet source” of either kind. What you lose is repeatability, and the substitute is the one this part has used throughout: expose both sheets in the same session under an unchanging sky, so that the comparison between them is valid even though the absolute dose is unknown, and record the conditions rather than a dose. Part XVI’s ultraviolet lesson makes that argument in full and this page does not re-make it.

Everything else here needs no facility at all. The seven faults and their seven answers, the incompatibility rule, the contrast inventory, the surfactant mechanism, the Pellet argument, the substrate survey and the whole of the choosing section are reading and arithmetic. A reader with no lamp, no sink and no sun can finish this page and arrive at a decision about which sensitiser to buy, which is what the page is for.

Two things have no alternative route and the page says so rather than implying one. Ultraviolet is an eye and skin hazard whether it comes from a lamp or from the sky, and a reader printing outdoors has an unenclosed source that cannot be interlocked; distance, sleeves, a hat and not looking at the source are administrative controls, the weakest kind. And there is no route at all to the dichromate contrast increase described above — not a safer substitute, not a smaller quantity, not a better-ventilated version. The course’s answer is the acid development bath of the New Cyanotype and the ammonia dose of the Simple Cyanotype, both of which are contrast controls the classic process never had, and neither of which is a dichromate at any dilution.

Recap

The classic cyanotype has seven documented faults and six of them are properties of the sensitiser rather than of the process. Ammonium iron(III) citrate is not a compound but a family of polymeric solids running from 14 to 28 per cent iron, so two printers cannot be sure they hold the same substance; it grows mould, it does not penetrate the fibre, it is slow, its potassium Prussian blue washes out of the highlights, and it carries twice the iron it can use, which is where bleeding comes from. The short exposure scale of about 0.9 is a consequence of the fifth fault, not an independent one.

The New Cyanotype of 1995 answers all seven by changing the ligand from citrate to oxalate — a pure, monomeric, small-molecule salt that penetrates the fibre — and then removing the potassium, which it does by provoking potassium iron(III) oxalate to crystallise where it can be filtered off. What you buy with that is four to eight times the speed, a maximum density verging on black and an exposure scale of about 2.2; what you pay is a preparation involving heat, filtration and an overnight wait, a variable yield, an oxalate hazard a level above the classic, and a mineral-acid development bath. That bath is also the family’s only development-stage contrast control with a usable range.

The Simple Cyanotype of 2019 builds a well-defined citrate complex in the beaker from iron(III) nitrate, citric acid and ammonia, and turns the amount of ammonia into a contrast control that sets the exposure scale anywhere from 2.7 to 1.8 before you coat anything. Mike’s Cyanotype reaches the same solution using triammonium citrate, for anyone who cannot have ammonia in the building, at the cost of that control.

The three chemistries do not mix and do not share trays: the citrate sensitisers put back the potassium the New Cyanotype exists to remove, and the mineral acid one process needs fogs the highlights of the other two. Polarity, as Pellet’s process shows, is decided by the developer and not the sensitiser. And the dichromate that appears in Ware’s own formula and in every commercial kit is ruled out here at every level, under a course-wide policy resting on a harmonised carcinogen classification and an exposure limit enforced by testing the worker.

Three things this page could not establish, and says so rather than filling in: the UK price of ammonium iron(III) oxalate against the citrate, because the one dated retailer check did not list the oxalate at all; what the 1995 Ag+ Photographic article actually says, because the course has read Ware’s later printings and not that one; and whether coating tools and trays must really be segregated by chemistry, which is the course’s own precaution drawn from Ware’s two documented incompatibilities rather than a rule he states.

Check your understanding

Question 1. Six of Ware’s seven shortcomings of the classic cyanotype have a single common cause. What is it?

• Potassium ferricyanide absorbs ultraviolet that the iron needs
• The properties of ammonium iron(III) citrate as a substance
• The twenty-minute wash
• The alkaline buffer in modern papers

Show the answer and why

Answer: The properties of ammonium iron(III) citrate as a substance

Variability, mould, poor absorption, slowness, peptization of the potassium form and bleeding are all properties of the iron salt or of what it is mixed with. Only the short exposure scale looks independent, and it is a consequence of the peptization. That is why every modern formulation begins by not buying the citrate.

Question 2. Why can the New Cyanotype’s sensitiser not simply be made by dissolving ammonium iron(III) oxalate and potassium ferricyanide together in one beaker?

• The two salts react in the bottle to make Prussian blue immediately
• Potassium iron(III) oxalate is sparingly soluble and crystallises out, in the bottle and in the coating
• The oxalate is destroyed by the potassium ion
• The mixture would be too concentrated to coat

Show the answer and why

Answer: Potassium iron(III) oxalate is sparingly soluble and crystallises out, in the bottle and in the coating

At the concentrations a pictorial sensitiser needs, potassium iron(III) oxalate — soluble to only about 6.6 g per 100 cc — crystallises out and produces white spots and fern-like patterns in the print. The commercial blueprint papers avoided it by keeping the ferricyanide at 5 per cent or less, which is why their maximum density was too low for pictorial work. Ware’s preparation provokes the crystallisation deliberately, hot and concentrated, where it can be filtered off and discarded.

Question 3. A printer develops a Simple Cyanotype in the 1 per cent nitric acid bath left over from a New Cyanotype session. What should they expect, and why?

• A deeper maximum density, because mineral acid releases iron(II) more completely
• No change, because both are cyanotypes and Prussian blue is Prussian blue
• Blue chemical fog in the highlights, because the citrate sensitisers tolerate only a very weak acid
• The image will dissolve, because Prussian blue is decomposed by acid

Show the answer and why

Answer: Blue chemical fog in the highlights, because the citrate sensitisers tolerate only a very weak acid

This is the second of Ware’s two incompatibilities. The New process needs a mineral acid for its best maximum density; the Simple and Classic sensitisers tolerate only a very weak acid, above which stronger acids cause blue chemical fog in the highlights. Prussian blue itself is stable in acid down to about pH 1 or 2, so the fourth answer is wrong for a different reason.

Question 4. Which statements about setting the contrast of a cyanotype are supported by the sources on this page? Select all that apply.

• The Simple Cyanotype’s exposure scale is set by the volume of ammonia used in its preparation, from about 2.7 down to about 1.8
• The New Cyanotype’s exposure scale is set by the strength of the acid development bath, from about 2.6 in water to about 2.1 in 1 per cent nitric acid
• The classic cyanotype’s contrast can be adjusted in the development bath in the same way
• Heat-drying the coated sheet shortens the exposure scale by about a stop, and leaving the exposed print in the dark lengthens it by about a stop

Show the answer and why

Answer: The Simple Cyanotype’s exposure scale is set by the volume of ammonia used in its preparation, from about 2.7 down to about 1.8, The New Cyanotype’s exposure scale is set by the strength of the acid development bath, from about 2.6 in water to about 2.1 in 1 per cent nitric acid, Heat-drying the coated sheet shortens the exposure scale by about a stop, and leaving the exposed print in the dark lengthens it by about a stop

The third is the trap. The classic process develops in water or in a very weak citric acid and has no development-stage contrast control at all; its contrast is decided by the negative. Ware measured the drying and darktime effects with step tablets in duplicate and advises against using either, because the combination returns the scale to where it started while losing maximum density.

Question 5. Pellet’s process makes a positive print — blue where the light did not fall. What produces that reversal?

• The gum arabic, which blocks the pigment where light struck the sheet
• Developing in hexacyanoferrate(II), which gives blue with the iron(III) the light did not reduce, and a colourless salt with the iron(II) it did
• A second exposure through the back of the sheet
• An iron(II) sensitiser rather than an iron(III) one

Show the answer and why

Answer: Developing in hexacyanoferrate(II), which gives blue with the iron(III) the light did not reduce, and a colourless salt with the iron(II) it did

The reversal is entirely in the developer. Ferricyanide, hexacyanoferrate(III), gives blue with iron(II); ferrocyanide, hexacyanoferrate(II), gives blue with iron(III) and Prussian white with iron(II). The gum is not what reverses the image — it is what fixes it, by hardening where iron(III) survives and re-dissolving where light made iron(II), so the colourless salt can be wiped away.

Question 6. Why does the New Cyanotype usually need no wetting agent while the classic sensitiser benefits from one?

• The oxalate solution is much more readily absorbed by paper fibres than the citrate, so it reaches the interfibrillar space without help
• The oxalate is coated at a higher temperature, which lowers the surface tension
• The New Cyanotype is always coated by brush, and brushes need no surfactant
• Surfactants react with oxalate to form insoluble soaps

Show the answer and why

Answer: The oxalate solution is much more readily absorbed by paper fibres than the citrate, so it reaches the interfibrillar space without help

The image survives the wash only if the pigment forms inside the surface cellulose fibres rather than in the coarse pores between them. The classic sensitiser is a large, polymeric, poorly absorbed molecule and a surfactant is what gets it in; the oxalate is small and monomeric and largely gets in on its own, which is why Ware calls a wetting agent "usually quite unnecessary" for it and reserves it for a few hard-sized papers.

Question 7. A reader asks whether they can make cyanotype negatives in a pinhole camera. What does the published arithmetic say?

• Yes, at about an hour in sunlight, the same as a lens at f/4
• Yes, but only on the New Cyanotype and only in summer
• No: at a typical pinhole f-number of 200 the exposure works out at about 2500 hours of sun, and reciprocity failure and chemical fog would prevent an image forming at all
• No, because pinhole cameras transmit no ultraviolet

Show the answer and why

Answer: No: at a typical pinhole f-number of 200 the exposure works out at about 2500 hours of sun, and reciprocity failure and chemical fog would prevent an image forming at all

Ware works it through: an in-camera exposure is about five times the square of the f-number times the contact exposure, a New Cyanotype image can just be got at f/4 with an hour of sun, and scaling that to f/200 gives (200/4)² = 2500 hours, about a year of eight-hour sunny days. His conclusion is "it won’t work!" — and a pinhole has no glass to absorb the ultraviolet, which is why the last answer is wrong for the right reason.

Check my answersClear