Rochelle salt kallitype developer

One salt, hot water, and a footnote explaining what the salt is called. That is the whole of this bottle, and the interesting thing about it is not what is in it but what is not: no developing agent, no preservative, no restrainer, and — the omission that decides the colour of the print — no borax.

The reducing agent was made by light, in the paper, hours ago. It is iron(II) oxalate, it sits where the light put it, and it is almost insoluble. This bath dissolves it. Everything the kallitype becomes follows from that, and the tone it comes out of the tray follows from the fact that the ligand doing the dissolving is a tartrate and there is no borate beside it.

Ingredient	Quantity	Form the source specifies
Potassium sodium tartrate tetrahydrate	100 g	The sheet writes "Rochelle salts" and footnotes the systematic name; it names no hydrate. The encyclopaedia's entry is the tetrahydrate, KNaC4H4O6·4H2O, formula weight 282.22, which is the commercial article; the anhydrous salt at 210.16 is sold beside it and carries 34 per cent more tartrate per gram
Water	200 mL, added	at 52 °C; Distilled water at 52 °C/125 °F, and an added volume rather than a make-up one: the sheet's table gives the water and the solid and states no final volume for the stock. The warmth buys dissolving time rather than solubility headroom, and the solution cools itself as the salt goes in.

The stock above is not the developer. Photographers’ Formulary’s traditional kallitype kit sells two stock solutions and prints three working developers made from them, and only one of the three is tartrate alone:

Working developer	Rochelle salt stock	Borax stock	Water	Time and temperature
Sepia tone	48 mL	none	to make 500 mL	10 minutes at room temperature
Brown tone	96 mL	64 mL	to make 500 mL	5 minutes, warm
Black tone	72 mL	128 mL	to make 500 mL	5 minutes, around 38 °C or higher

The borax stock is 75 g of borax in 400 mL of water at 52 °C, and it is published separately as the borax kallitype developer. This page is about the left-hand column and the first row.

Purpose

To dissolve the iron the light reduced, so that it can hand its electron to silver; to carry the iron that light never touched out of the paper; and to do both without an alkali in the tray, which is what makes the print sepia rather than black.

The first two duties are common to every siderotype developer and are set out in full on the sodium citrate page. The third is this bath’s own, and it is a choice rather than an accident. Photographers’ Formulary states the rule in its own words: “if you increase the amount or proportion of Stock Developer Solution A (Rochelle salts) in the developer, you will increase the sepia-tone of the print. If you increase the amount or proportion of Stock Developer Solution B (borax), you will increase the blackness of the print.” Take the borax to zero and you are at the warm end of the range the kit can reach.

Mike Ware’s account of the platinotype gives the definition of a developer in this chemistry, and it is worth having in mind before any of the rest: Willis’s process needed a noble-metal salt, a light-sensitive iron salt, and third, “a ‘developer’, potassium oxalate, to dissolve the insoluble ferrous oxalate produced by light, providing a strongly reducing solution” which could then reduce the metal. Not to reduce anything itself. To dissolve.

Recommended uses

A kallitype where you want a warm print out of the tray. This is the only bath in the traditional kit that gives sepia, and it gives it without an additive, a temperature or a ratio to get wrong. If the print is going to stay untoned, or be toned only lightly, this is the developer that puts the colour where the process is most at home.

The warm end of a deliberate colour range. The reason to own both stocks is that the pair spans a range no single bottle covers. A printer working the traditional kit has three published mixtures and can interpolate between them; the experiment at the foot of this page is how to map that range on one negative rather than guessing at it.

Where there is no way to warm a tray. The sheet asks for around 38 °C for both baths that contain borax and asks for nothing at all here. A kallitype developed at room temperature in a cold darkroom is a working proposition in this bath and a compromised one in the other two.

As the demonstration that a developer can choose an image colour. One negative, three sheets, and the three published mixtures give three different pictures from identical exposures. Very little else in this formulary shows the connection between a bath’s composition and the colour of metallic silver so directly, and the chemical page for the salt sets out the same three mixtures from the reagent’s side.

When another formula is preferable

• When you want a neutral black. Use the borax developer. That is not a matter of taste alone: Ware’s argument, set out under The mechanism, is that the kallitype family uses “alkaline-buffered developers of high pH, e.g. Borax” specifically to stop the nitrate anion dissolving the colloidal image silver. The borate bath protects the image better than this one does, and it deposits more iron in the paper for doing so.
• When permanence is the whole point and the print will be toned. Use the sodium citrate developer. Sandy King’s method reduces the developer question to one bath on the grounds that development is where most of the residual iron is removed and that colour should be decided later, in the toner, “when it is fairly obvious if the print is a keeper or not”. A developer that also chooses the colour takes that decision earlier and takes it in the dark.
• When the image metal is platinum. Use Willis’s potassium oxalate developer. The redox table under The mechanism says why: the weaker complexed-iron couples reduce silver(I) and gold(III) and do not reduce platinum(II).
• When there is to be no developer at all. The Van Dyke Brown sensitiser and Ware’s argyrotype print out in the frame and go straight to a wash. You give up the shadow depth a developing-out process buys and you save a tray.
• When you are reconstructing the process as Nicol first published it. In Nicol’s kallitype the silver is in the developer rather than in the paper. That is a different formula and this bath is not a substitute for it.

Mixing

One hundred grams of Rochelle salt into two hundred millilitres of distilled water at 52 °C/125 °F. That is the sheet’s instruction and there is nothing else in the bottle. Put the hot water in the vessel first and add the solid to it, which is the ordinary reason for that order — a solid tipped into a dry vessel and then wetted cakes at the bottom — and, here, a specific one as well.

The solution gets colder while you stir it. The sheet says so plainly: “Rochelle salts dissolve with absorption of heat: therefore, the solution will cool and it may be necessary to warm the bowl slightly. Upon stirring, the mixture turns cloudy but, in time, will clear.” That is an endothermic dissolution, and it is established chemistry rather than a peculiarity of this salt: pulling the ions out of a crystal lattice costs more energy than hydrating them returns, and what drives the process anyway is the entropy gained when an ordered crystal becomes ions dispersed through water. The practical consequence is that a stock started at 52 °C is well below that within a minute, and that the cloudiness is undissolved crystal, not a precipitate. Warm the bowl and keep stirring.

🧮What the hot water is actually for

It is not for solubility headroom. PubChem carries no solubility section for the tetrahydrate at all, so the figures have to come from the anhydrous salt’s own record, and they are generous: about one gram in one millilitre of water, and 47.4 g per 100 mL as low as 6 °C.

The stock is 100 g of the tetrahydrate. Of that, the four waters of crystallisation are 4 × 18.015 = 72.06 of the formula weight of 282.22, so 100 g of the crystals carry about 74.5 g of anhydrous salt and about 25.5 g of water of their own. The bath is therefore roughly 74.5 g of tartrate in about 225 g of water — call it 0.33 g per millilitre — against a cold-water solubility of 0.47 g per millilitre and a warm-water one of about 1.

So the stock is about a third of saturation even in cold water, and there is no risk whatever of it failing to dissolve. The 52 °C buys rate, not capacity. A hundred grams of coarse crystals takes a long time to go into cold water, and the dissolution cools the water further as it goes, which slows it again. Heat breaks that loop. Nothing is lost by mixing the stock cold except patience.

Use distilled water, as the sheet requires in capitals, and the reason is calcium. Ware gives the reaction that hard water sets off in this chemistry, and it happens in the tray rather than in the bottle, because the oxalate comes off the sheet:

Ca²⁺ + Fe(C₂O₄)₃³⁻ + 2 H₂O → CaC₂O₄ + Fe(C₂O₄)₂(H₂O)₂⁻

Calcium displacing iron from its oxalate complex, after Ware

Calcium precipitates as insoluble calcium oxalate and pushes the iron down a chain of aquated species that ends in iron(III) hydroxide in the paper — which is exactly the thing this bath exists to prevent. A developer whose ligand is occupied holding your tap water’s calcium is a developer with less ligand for your print’s iron.

🧮How strong is the working bath? A bound, not a number

This is the one piece of arithmetic the sheet will not let you finish, and it is worth seeing why.

Stock Developer Solution A is 100 g of solid added to 200 mL of water. The sheet states no final volume for it, which under the convention Kodak states in its own 1928 primer — “cold water to make” is reserved for the line that fixes a solution’s strength — means the stock’s volume is not published and is certainly greater than 200 mL, because 100 g of salt occupies real space.

Call the stock’s volume V. The tartrate in 48 mL of it is 48 × 100 ÷ V grams. Since V is greater than 200, that is less than 48 × 100 ÷ 200 = 24 g, and it goes into a bath made to 500 mL. So the sepia-tone working developer is less than 4.8 per cent w/v, and by an amount nobody read publishes enough to compute, because no source read gives a density for the salt or a volume for the stock.

Put that against the two nineteenth- and twentieth-century printings. Wall’s 1912 sepia formula is half an ounce of Rochelle salt in ten fluid ounces of distilled water: 4.99 g per 100 mL if his ounce is the avoirdupois one of 437.5 grains, 5.47 g per 100 mL if it is the apothecaries’ ounce of 480 grains, and the ambiguity does not matter at this precision. Thomson’s developer, in Wall’s 1924 collection, is equal parts of a 5.2 per cent Rochelle salts solution and a 9.4 per cent borax solution, so 2.6 per cent tartrate in the bath as used.

Three sources across a century put a tartrate kallitype developer between roughly two and a half and five and a half per cent, and the modern kit sits inside that band. That is corroboration of a working strength, not of a formula, and it is recorded here as such: the quantities in the object above are Photographers’ Formulary’s and nobody else’s.

Making the working bath. Put 200 mL of distilled water at 20 °C/68 °F in the tray or a bowl, add 48 mL of the stock, then add a further 252 mL of water to bring the bath to 500 mL. Stir it. The sheet’s own reason for making the developer up in the mixing bowl rather than the tray is that you may want to store it, and it says to do so in “any convenient container”.

Behaviour

The image arrives quickly and the ten minutes are for the iron. No source read times the appearance of the image in this particular bath; the two sheets that publish a citrate developer both describe an image that is visually there in well under a minute, and the mechanism is the same. What the clock is for is the second job. Wall states it in 1912 for all four of his kallitype developers, in a sentence that has not been improved on since: the print “must be left in the developer for about ten or fifteen minutes. This is necessary in order to ensure that the iron salts shall be all dissolved. The presence of the iron salts is a most prolific cause of failure.”

Ten minutes here against five in either borax bath, and no heat. That is the trade this developer makes, and the sheet makes it explicit by warming the other two and not this one. It is the slow, cool, undemanding bath of the three, and it is also the one that gets the least help from temperature in shifting iron out of the paper — which is a reason to respect the ten minutes rather than trim them.

Prolonged immersion does not over-develop the print. Wall’s assurance is the one thing a printer coming from silver-gelatin most needs to hear: leaving the sheet in “will not produce over-development, providing the exposure has been correctly timed”. There is no reservoir of developing agent waiting to build density; there is only as much reduced iron as the light made, and once it has done its work the bath has nothing left to add.

The developer has to reach the whole sheet at once. Wall’s technique for a large print is to immerse it face downwards and turn it over immediately, breaking any air bubbles with a fingertip. A print wetted in two stages develops in two densities and the tide-line does not come out afterwards.

A dark residue collects in the bath. Photographers’ Formulary describes it for the citrate developer in its later kit — “a residual black substance forming from and next to dark tones” — and directs the printer to agitate so it does not settle on the highlights, and to filter it out of the stored bath through a coffee filter. No source read describes it for this bath specifically, and no source read identifies what it is. The course’s reading, offered as a reading and not a finding: a tray of tartrate solution holding reduced iron is a reducing solution, and any silver nitrate that has diffused off the sheet meets that solution in bulk, so finely divided silver reduced in the liquid rather than in the paper is the obvious candidate. Tartrate has a second candidate the citrate bath does not: silver tartrate is sparingly soluble, and a bath carrying dissolved silver and a large excess of tartrate is a place for it to appear. Neither is established here.

Exhaustion announces itself in the next tray, not in this one. This is the most useful thing to know about any kallitype developer in use. The tartrate is nowhere near consumed after a few prints — a sheet carries at most a few millilitres of sensitiser and the bath holds tens of grams of ligand — but the iron it has accumulated makes each successive print harder to clear. King states the consequence for his own bath and the chemistry is not different here: an unreplenished developer leaves “not only an unpleasant stain in the masked areas of the print, but may also decrease permanence, because the stain consists in large part of residual ferrous iron.” This sheet publishes no replenishment rate and no discard point at all, which is a real gap in the evidence and not something the course can fill by arithmetic.

The print is not finished when it leaves the developer. It goes through a clearing bath — potassium oxalate in this kit, at 60 g in 500 mL for five minutes — and then a fixer, and both change what you are looking at. Bostick and Sullivan put the warning plainly: “Kallitype prints will bleach in the fix, so either tone before fixing or overprint by a stop or two.” Judge the developer by the print you get at the end of the line, not by the one in the tray.

Image characteristics

Colour: sepia, and the sheet’s own numbers say why. Set the three published mixtures side by side and one column moves monotonically with the tone and the other does not.

Working developer	Rochelle salt stock	Borax stock	Ratio A to B	Colour
Sepia tone	48 mL	0 mL	all A	warmest
Brown tone	96 mL	64 mL	1.5	middle
Black tone	72 mL	128 mL	0.56	coldest

The brown-tone bath carries twice the tartrate of the sepia one and is colder, so the amount of tartrate is not what sets the colour. The borax rises 0, 64, 128 as the print goes sepia, brown, black, and the A-to-B ratio falls steadily. The manufacturer’s rule is stated in terms of proportion, and its own table is why.

⚗Wall's four developers of 1912 order themselves the same way

The same arithmetic on a source seventy years earlier gives the same answer. Wall’s four kallitype developers, in ten fluid ounces of distilled water each, carry:

• Sepia — Rochelle salt half an ounce, no borax.
• Purple — Rochelle salt 450 grains, borax 120 grains.
• Black — Rochelle salt three-quarters of an ounce, borax one ounce.
• Warm maroon — sodium tungstate one ounce, and neither of the two salts above.

Laid out warm to cold, the borax climbs from nothing to 120 grains to a full ounce, and the tartrate goes 219, 450, 328 grains, which is not an order at all. Two independent sources, seventy years apart and in different unit systems, both put borax in charge of the colour and neither puts tartrate in charge of it. The course states that as an observation about two published tables, not as a mechanism; the mechanism it suggests is below.

Wall’s fourth developer is a reminder not to over-generalise: sodium tungstate contains no tartrate and no borate and gives a fourth colour again. The chemical encyclopaedia has no entry for sodium tungstate, so the course can go no further with it than to record that Wall printed it.

Why a kallitype is brown at all is particle size, not chemistry. Ware’s measurements on plain-paper silver prints put the image particles at colloidal dimensions, around twenty nanometres, far smaller than the wavelengths of visible light, and their colour comes from an absorption that depends on the particles’ “shape, size, state of aggregation and chemical environment” rather than on anything about silver as an element. The same metal in the micron-sized filaments of a developed silver-gelatin print reads as neutral black. Ware reproduces Wiegel’s calculated and experimentally confirmed table for silver hydrosols: 10 to 20 nm particles transmit yellow, 25 to 35 nm red, 35 to 45 nm purplish-red, 50 to 60 nm violet. Departure from a spherical shape moves the absorption again, and so does the refractive index of the matrix the particles sit in.

Why one developer should give a different colour from another is the course’s reading, and it is built on a statement Ware makes about a different system. Discussing Talbot’s ammonia-buffered paper, Ware writes that “the more neutral image colour is also consistent with a larger particle size for the silver colloid formed in such a sensitizer” — a buffered alkali, a larger particle, a colder colour. Borax is a buffer at about pH 9.5; a Rochelle salt solution is not buffered in any useful sense and sits far below it. If the alkalinity of the bath is what pushes the reduction towards fewer and larger particles, then the borate bath should give the colder print and the tartrate bath the warmer one, which is what both sources report. That is a hypothesis consistent with the evidence and not a measurement. Nobody read has measured the particle sizes any kallitype developer produces, and this page therefore gives the principle and no number.

The sizing is part of the colour too, and it is easy to forget. The same sheet says that prints on paper sized with arrowroot starch “will have a brown color while those sized with gelatin will have a blue tone”. If you are comparing developers, size all the sheets the same way or you are comparing two variables.

Contrast: none is offered here, and that is a policy rather than an omission. The only contrast control this sheet publishes is 5 to 20 drops of a 10 per cent potassium dichromate solution per 500 mL of mixed developer, and it is emphatic in bold type that the dichromate goes into the developer and not into the sensitiser. The course does not use chromium(VI) at any level; the chromium policy sets out the rule and the sensitiser page collects what the three suppliers and King each publish. What is left is the negative, and the sheet’s own claim for the process is that a negative with a density range up to 1.85 can be printed.

Tonality: long, and mostly decided elsewhere. This developer contributes a colour and no contrast control, so the tonal decisions happen in the negative before it and in the clearing bath, the toner and the fixer after it.

The mechanism

Four steps, of which this bath performs exactly one.

Step one, in the printing frame: light makes an insoluble solid. Ultraviolet light decomposes iron(III) oxalate, reducing the iron at the expense of an oxalate ligand which leaves as carbon dioxide. Döbereiner reported the reaction in 1831, decades before anyone printed with it:

Fe₂(C₂O₄)₃ + UV → 2 FeC₂O₄ + 2 CO₂

Döbereiner's photolysis, after Ware

The product is iron(II) oxalate, and Ware gives its solubility as 0.022 g per 100 cc — for practical purposes it stays exactly where the photon left it. That is what makes the latent image an image rather than a smear, and it is also what stops anything further happening until a developer arrives.

Step two, in this tray: the tartrate dissolves it. Tartrate is a ligand — a molecule with donor atoms that bind a metal ion — and the tartrate dianion offers two carboxylates and two hydroxyls to an iron centre. Complexing the iron(II) takes it out of the insoluble oxalate salt and puts it into solution, mobile, and still reducing. Nothing in the bath is oxidised or reduced at this step. It is a complexation, and it is the entire contribution of the only substance in the bottle.

Step three, still in this tray: the dissolved iron(II) reduces silver(I) to metal. The silver nitrate was coated on the sheet with the ferric oxalate and has been sitting there in the dark. The moment mobile iron(II) reaches it, one electron moves:

Fe²⁺ + Ag⁺ → Fe³⁺ + Ag

The image-forming step, stripped of its ligands

Wall states the same thing in 1912 without the notation: the exposed paper carries “an image made up of ferric and ferrous salts, which, under the action of the developer, will reduce the silver nitrate to metallic silver, thus giving an image of that metal.” His 1924 collection puts the sequence in order — “the ferrous salts dissolve in the developer, and reduce silver nitrate to the metallic state at the points where the ferric salt has been reduced by light” — and that last clause is the whole reason the process makes a picture: the reduction happens where the iron dissolved, which is where the light fell.

Step four, over the following minutes: the tartrate carries the unexposed iron away. Every part of the sheet that was masked still holds iron(III), and iron(III) left in paper is what kills a kallitype. Ware is unambiguous about the failure mode: “the inherent problem of the iron-based silver processes lies in the danger of leaving residual ferric iron in the print — to its ultimate undoing, because iron(III) will oxidise silver with consequent degradation of the image.” Tartrate binds iron(III) as well as iron(II), so the same bath that developed the print begins clearing it.

Deeper: the redox arithmetic, and the number this page has not got

Ware gives standard potentials for the two complexed-iron couples that matter in siderotype:

Couple	Potential
Iron(III)/iron(II) as the oxalato complex	+0.02 V
Iron(III)/iron(II) as the citrato complex	+0.372 V
Iron(III)/iron(II) chelated by EDTA	−0.12 V
Silver(I) to silver metal	+0.80 V
Palladium(II) tetrachloride to palladium	+0.62 V
Platinum(II) tetrachloride to platinum	+0.73 V

Reduction is thermodynamically allowed when the metal’s potential is the more positive of the pair, and every one of the iron couples clears silver’s +0.80 V with room to spare. That is why an iron developer reduces silver at all.

The number this page has not got is the tartrato couple. No source read publishes a potential for iron(III)/iron(II) complexed by tartrate, and the course will not estimate one. What the table does establish is a bracket: the ligand changes the potential over a range of nearly half a volt — EDTA pulls it down to −0.12 V, citrate pushes it up to +0.372 V — and even the least reducing of the three still reduces silver(I) comfortably. Ware says so explicitly of citrate: it “does not reduce platinum(II) or palladium(II), although it will reduce gold(III) and silver(I), under the printing conditions”. A tartrato couple anywhere in that region will make a silver image. Whether it sits above or below the citrato couple, and what that does to the rate, is not established by anything read for this page, and the rate is exactly what the particle-size argument above would need.

Deeper: the bargain this bath strikes, and the side it takes

Ware states the kallitype’s central difficulty in two sentences, and the two halves pull opposite ways. First: “nitrate is an oxidising anion, and tends to dissolve the colloidal image silver during wet processing, especially under acidic conditions. To minimise this loss of image the Kallitype process employs alkaline-buffered developers of high pH, e.g. Borax.” Then, immediately: “Alas, these create a new problem, because they cause hydrolysis of the excess iron(III) in the sensitizer and the deposition of insoluble ferric hydroxide in the image, which ultimately causes it to fade.”

So there are two ways for a kallitype to die and the developer’s pH decides which one you are more exposed to.

• A low-pH bath leaves the iron(III) soluble and complexed, so more of it washes out — but leaves the nitrate free to etch the finest image particles while the print is wet.
• A high-pH bath protects the silver — but Ware records that iron(III) hydrolyses above about pH 4, chemisorbs to the hydroxyl groups of cellulose, and on drying converts irreversibly to goethite, which no clearing bath will lift.

A tartrate-only developer sits on the low-pH side of that bargain and the borax developer on the high side. That is the honest description of the difference between this page and its neighbour, and it is worth stating plainly because the usual account — that one bath is “for sepia” and the other “for black” — makes it sound like a matter of taste.

How far down the pH scale this bath sits is not published. No source read gives a measured pH for a Rochelle salt kallitype developer. What can be said from established chemistry, and is said here as a reading rather than a measurement: Rochelle salt is the fully neutralised salt of tartaric acid with potassium and sodium, two strong bases. IUPAC’s high-confidence dissociation-constant dataset gives tartaric acid pKa1 3.17 and pKa2 4.91 for the meso form, assessed reliable, and approximate values of 3.03 and 4.46 for the D form a photographic supplier actually sells; on either set the second proton is long gone by pH 6. A solution of the fully neutralised salt is therefore mildly alkaline through hydrolysis of the tartrate dianion, and, sitting several pH units above the higher pKa, it is very poorly buffered: it has almost no capacity to resist the acid that development and the paper bring into it. Borax at about pH 9.5 is a buffer in the proper sense and holds its pH. Anyone who wants the real number should measure it, and the experiment section says how.

Function of every ingredient

Potassium sodium tartrate tetrahydrate, 100 g — the whole of the developer.

What it is. Rochelle salt, KNaC4H4O6·4H2O, formula weight 282.22, CAS 6381-59-5, and one of the few reagents in this formulary sold in a food aisle as well as a laboratory catalogue: its encyclopaedia entry records it as the food additive E337. It is a double salt: one tartrate dianion, one potassium ion, one sodium ion and four waters of crystallisation. Photographers’ Formulary’s sheet writes “Rochelle salts” and footnotes the systematic name, and specifies no hydrate at all — see Storage for why that matters.

Why it is here. Because the developer of a siderotype is a solvent for iron(II) oxalate and nothing else, and a tartrate is a competent one. Tartrate offers an iron centre two carboxylate groups and two alpha-hydroxyl groups; it is a chelating ligand, wrapping the metal in a ring rather than touching it at one point, which is what makes its complexes stable enough to pull iron out of an insoluble salt and hold it in solution. The historical pedigree is older than the kallitype: Willis’s 1880 patent for platinum printing names “the tartrate or citrate of soda, of potash, or of ammonia” among the salts that will serve as an improved developer, though he adds “I prefer however to use the citrate of soda”.

What it does chemically. Three things, in this order. It dissolves the photochemically formed iron(II) oxalate by complexing the iron(II), which is the developing action. It then holds the iron(III) produced when that iron(II) reduces the silver, keeping it soluble instead of letting it hydrolyse into the paper. And over the remaining minutes it complexes and removes the iron(III) in the parts of the sheet the light never reached, which is a clearing action happening early. Not one of those steps involves the tartrate being oxidised or reduced. There is nothing in this bottle that can reduce silver; the reducing agent was manufactured by the exposure.

What the photographic consequence is. A developed silver image where the light fell, in a warm brown that no other bath in the kit produces, and a sheet whose unexposed areas have already begun to give up their iron before the clearing bath sees them.

What happens with more of it. Nothing dramatic, and that is worth knowing. The bath is at about a third of saturation, so there is a great deal of room. More tartrate means more ligand for the same quantity of iron, which should mean a faster and more complete removal of the unexposed iron and no change in the amount of silver reduced, because that is fixed by the exposure. What it will not do is make the print warmer: the sheet’s own brown-tone developer has twice the tartrate of the sepia one and is colder, because it also has borax. Strength and proportion are different levers and the sources only document the second.

What happens with less of it. Below some strength the ligand runs out before the iron does, and the symptom is the one Wall names — iron left in the paper, a stain in the masked areas, and a print whose permanence is compromised before it is dry. No source read establishes where that threshold is. The three published tartrate developers span roughly two and a half to five and a half per cent, and nothing read goes below it.

What it interacts with. Borax, which is the whole subject of the neighbouring page. Silver(I), with which tartrates give sparingly soluble silver tartrate — the basis of Wall’s silvering process and a nuisance anywhere a sensitiser splashes into a developer bottle. Calcium in hard water, which competes for the ligand. And strong oxidisers, in the ordinary way of any salt of an organic acid, which is one more reason the dichromate this course refuses is a bad neighbour as well as a bad hazard.

Distilled water, 200 mL at 52 °C — not an inert ingredient. The sheet requires distilled water in capital letters for every solution in the kit, and the mechanism is the calcium reaction under Mixing: tap-water calcium precipitates as calcium oxalate in the tray and drives iron down the hydrolysis chain the developer exists to prevent. The temperature is a rate control, not a solubility one, and the water’s own contribution ends there.

What is deliberately absent. There is no developing agent, because the reducing agent is made by light. There is no sodium sulfite or other preservative, because there is no easily oxidised organic reducer to preserve. There is no separate alkali, and its absence is the formula’s active decision rather than an oversight — the borax that would supply it is sold in the same kit, in the other bottle. And there is no restrainer, because there is no fog reaction to restrain: unexposed silver nitrate simply stays as silver nitrate until the fixer removes it.

Interactions

Tartrate and iron, in both oxidation states. This is the interaction the formula is built on, and its double duty is what makes a single-salt bath sufficient. A ligand that held only iron(II) would develop the print and leave every unexposed area loaded with iron(III); a ligand that held only iron(III) would not develop anything at all.

Tartrate and borax — the interaction this page is defined by not having. The two stocks are miscible in any proportion and the sheet publishes three of those proportions. The consequence of adding borax is a colder image, a shorter development time and a bath that wants heat; the consequence of leaving it out is this page. The full argument about what each does to the silver and to the iron is under The mechanism.

Tartrate and silver(I). Silver tartrate is sparingly soluble. In the tray it costs nothing — the sheet is already surrounded by dissolved silver — and it may even be part of the residue described under Behaviour; in the stock bottle it is a contamination signal. If a bottle of Rochelle salt stock clouds after a session, suspect a pipette or a funnel that also touched the sensitiser.

The developer and the first rinse, where two published rules pull in opposite directions. King’s instruction after his citrate developer is unambiguous: “if the first rinse is alkaline, ferrous hydroxide compounds may be formed in the paper, making complete clearing difficult or impossible” — so rinse in neutral or slightly acidic water. Ware’s argument is equally unambiguous in the other direction: nitrate dissolves colloidal image silver “especially under acidic conditions”. Both are about the same wet print in the same minute. The course states both and picks neither, because nothing read tests them against each other; what it will say is that King’s rule is stated for a working process by someone who prints it, and Ware’s is stated as chemistry by someone who studies it, and that they are not necessarily in conflict, since “not alkaline” and “not strongly acidic” leave a window between them.

The developer and the clearing bath. In this kit the clearing bath is potassium oxalate, 60 g in 500 mL, five minutes at 20 °C. Carrying developer into it on a poorly drained print dilutes it and carries iron with it; the sheet’s own instruction is to drain the print, rinse it quickly, and move on.

The developer and the fixer. The kit’s fixer is 50 g of sodium thiosulfate pentahydrate and 12 mL of 28 per cent ammonia in a litre, used for not more than five minutes, and the sheet’s reason for the limit is important: “a longer soak will cause the print to fade. Do not use a standard photographic fixing bath; the finely divided, unprotected silver metal will be etched from the print.” A kallitype’s silver is colloidal and unprotected by gelatin, and an acid hardening fixer will take it away.

The developer and chromium(VI). The one interaction this course declines. See Safety.

Variants

Wall’s four developers of 1912 — the historical record, not a procedure. Wall’s ninth edition prints four kallitype developers, each in ten fluid ounces of distilled water, and says that “on the salt used the tone of the finished print will depend”. Two of the four are tartrate baths and both carry a dichromate:

Wall’s formula	Tone	Rochelle salt	Borax	1 per cent potassium bichromate
No. 1	pure black	¾ oz	1 oz	7 to 9 drams
No. 2	sepia	½ oz	none	4 to 5 drams
No. 3	warm maroon	none	none	none — sodium tungstate 1 oz
No. 4	purple	450 grains	120 grains	10 drams

Formula No. 2 is the direct ancestor of the bath on this page, at essentially the same tartrate strength, and the difference between them is the chromium. The course prints Wall’s composition as history and does not offer it as an instruction; the modern kit reaches the same colour without a chromium(VI) salt, which is why the modern kit supplies the quantities in the object above. Wall’s No. 3 is a curiosity worth noticing: sodium tungstate alone, no tartrate, no borate, and a fourth colour again — a reminder that the tartrate-versus-borate account is a description of one family of baths and not a law.

Thomson’s developer, 1924. Wall’s later collection gives it as equal parts of a 5.2 per cent Rochelle salts solution and a 9.4 per cent borax solution “with the addition of 0.02 to 0.8 per cent potassium bichromate, which keeps the whites pure and acts as a restrainer”, and directs thirty minutes in it. Two things are worth taking from it even though the course does not reproduce it. The first is the thirty minutes, three times Wall’s own 1912 figure and six times the modern sheet’s, which shows how unsettled the development time was. The second is the description of the dichromate as a restrainer rather than a contrast agent, which is a different account of the same additive from the one the modern sheets give and is recorded here as the source’s wording.

Hall’s formulas, 1924. Wall prints a kallitype developer of an entirely different construction: sodium acetate 125 g and tartaric acid with a dichromate solution, in a litre. It is worth a line because it is the tartrate idea approached from the acid rather than the salt, and because its clearing bath is sodium citrate with citric acid — the same chemistry that eventually replaced all of these.

What the trade did next. Photographers’ Formulary’s own later kallitype kit carries no Rochelle salt at all: it replaces the three tartrate-and-borax developers with a single 20 per cent sodium citrate bath. Bostick and Sullivan’s kit supplies a proprietary black-tone developer whose composition is not published. King’s published method recommends one developer and states the reason: “there is really no reason to use any other developer unless you want an unusual color that cannot be rendered through toning with gold, platinum or palladium.” That last clause is the argument for keeping this page. A tartrate bath is exactly the case he leaves open.

No course variant of these quantities is offered. There is nothing here to make safer by changing the weights: the salt is not a hazard, the hazards in this process live in the sensitiser, the clearing bath and the fixer, and a weaker tartrate bath would simply be a worse one.

☠The dichromate contrast control, which this page also refuses

Every source read that gives a contrast control for a Rochelle salt kallitype developer gives a chromium(VI) salt. Photographers’ Formulary supplies 5 g of potassium dichromate for a 10 per cent solution and directs 5 to 20 drops of it per 500 mL of mixed developer, in bold type, into the developer and not the sensitiser. Wall’s 1912 developers each carry a 1 per cent bichromate solution. Thomson’s 1924 formula carries 0.02 to 0.8 per cent. The Getty atlas records the same practice for the platinotype, with potassium chlorate as the other historical oxidant.

The course’s chromium ruling is that chromium(VI) is not used, at any level, anywhere. Potassium dichromate’s notified classification carries H340 for genetic defects, H350 for cancer, H360 for reproductive toxicity, H330 for fatal by inhalation and H410 for long-lasting aquatic toxicity, and its own page classifies it safety level D. The sheet that supplies it concedes that “all chromium compounds are potential carcinogens” and that a spillage on skin “will cause a chemical burn, which will appear as ulceration” — and then directs the excess down a drain with copious water, which this course does not follow.

So this page publishes no contrast control. What is left is the negative. The sensitiser page collects the specific figures each supplier gives and the reasons the course declines them; no chromium-free substitute is invented here, because no source read establishes one.

Safety

Safety level A. The bath is a food-additive salt dissolved in distilled water and handled at darkroom temperatures, and it is classified A because the hazard assessment of its only ingredient comes back empty in three independent places.

• PubChem holds no GHS classification for the tetrahydrate at all — the classification heading on CID 165453 returns nothing. The classification quoted on the chemical page is that of the anhydrous salt, against which three separate ECHA C&L entries are filed and all three report that the substance does not meet GHS hazard criteria.
• HSE’s EH40 has no entry for tartrates, and neither does the NIOSH pocket guide.
• The salt is E337, a permitted food additive.

What that does not mean. EH40 states in its own paragraph 6 that absence from a list does not indicate that a substance is safe, and the ECHA data is thin — only a small fraction of companies supplied information under the main entry. A food-additive designation is a statement about a regulated use at regulated concentrations, not a hazard assessment of a jar of the crystals. The supplier’s safety data sheet for the product you buy is the authority for it.

What the real hazards of this operation are, and where they come from. None of them is the tartrate.

• Hot water. The stock is mixed at 52 °C/125 °F. That is a scald temperature for a slow spill. Mix in a sink, in a vessel you can hold, and do not carry a full open bowl of it across a darkroom.
• The sheet in the tray is not clean. It arrives carrying silver nitrate and ferric oxalate, and the developer becomes a solution containing both once a few prints have been through it. Silver nitrate stains skin brown-black and the stain is silver bound to protein, which does not wash off, and oxalates are toxic — which is why the same sheet singles out the potassium oxalate clearing bath below. Handle prints with tongs.
• The baths on either side. In this kit the clearing bath is potassium oxalate, which the sheet itself calls “rather toxic” and an anticoagulant, and the fixer is made with 28 per cent ammonia, which the sheet says must be mixed in a well-ventilated area and never sniffed. Those hazards belong to their own pages, and they are in the same session as this tray.
• Chromium(VI) is not used. If your kit contains the dichromate packet, it still has to be labelled, stored away from anything reducing, and disposed of through a licensed route rather than down a drain.

Personal protective equipment. Nitrile gloves — HSE’s COSHH essentials sheet P1 takes single-use nitrile gloves 0.2 mm thick as splash protection where the safety data sheet gives no more specific advice — and safety glasses, with splash goggles while the stock is being mixed with hot water. Weigh the crystals without raising dust, which is ordinary powder discipline rather than a response to any classification.

Ventilation. Nothing in this bath produces a vapour, and ventilation is not among the controls for the developer itself. It is among the controls for the session: the ammonia that goes into the kit’s fixer is the reason the darkroom needs air moving in it.

Storage

The dry salt. Dry, closed and labelled. The Merck Index, by way of Haz-Map, records that the crystals effloresce slightly in warm air, and the direction of that drift matters: a jar kept somewhere warm loses water of crystallisation and therefore gains tartrate per gram. For a developer measured in hundreds of grams the effect is small, but it runs towards a stronger bath rather than a weaker one, which is the direction that costs nothing.

Which hydrate did you weigh? Neither the sheet nor Wall names a grade. The tetrahydrate at 282.22 and the anhydrous salt at 210.16 differ by 34 per cent in tartrate per gram: 100 g of the tetrahydrate is 0.354 mol of tartrate and 100 g of the anhydrous salt is 0.476 mol, 34 per cent more. The encyclopaedia’s entry, and the article a photographic supplier sells, is the tetrahydrate. If you have bought the anhydrous grade, say so in the notebook rather than adjusting the weight, because no source read publishes a conversion for this formula and the course will not invent one.

The stock solution. In any closed, labelled bottle, and no source read publishes a keeping time. What can be said is what to look at rather than what date to write: a tartrate solution that has grown crystals has been cold, and warming and stirring will redissolve them; a stock that has clouded without being cold has met something, and the likeliest something is silver.

The working developer. The sheet says to store it in any convenient container and gives no life for it either. It is not the tartrate that ages — it is the iron and the silver the bath accumulates, and the symptom is a clearing problem, not a development one.

Incompatibilities

• Chromium(VI) in any form. Refused under the chromium policy, and separately a poor chemical neighbour: a strong oxidiser in a bath whose whole purpose is to carry a reducing iron(II) complex.
• Silver salts, in the bottle. Tartrates give sparingly soluble silver tartrate. In the tray this is part of the process; in the stock it is contamination.
• Strong oxidisers generally, the ordinary incompatibility of a salt of an organic acid.
• Hard water. Calcium competes for the ligand and precipitates as calcium oxalate in the presence of the oxalate coming off the sheet. Distilled water throughout, as the sheet requires.
• An acid stop or an acid hardening fixer. Not used in this process at all. The sheet is explicit that a standard photographic fixing bath will etch the colloidal silver out of the print, and Ware’s account explains why: nitrate dissolves colloidal silver most readily under acidic conditions.
• Borax, when you did not intend it. Not a hazard, but the one contamination that will silently change what the page is about. Keep a dedicated funnel, graduate and bottle for each stock; a few millilitres of the borax stock in the tartrate bottle is a colour shift you will chase for weeks.

Waste

The tartrate itself is close to the least of the darkroom’s problems: it is a food additive, and the potassium and sodium that come with it are fertiliser salts. What makes the spent developer a waste stream is what it has taken out of the print — iron from the sensitiser and silver that has diffused off the sheet, and the black residue that settles out of a used bath is very likely silver-bearing.

Bottle the used developer, label it and hold it with the other iron- and silver-bearing baths from the same session, which in this kit means the clearing bath and the fixer as well. Kodak’s J-52 publication gives 5.6 to 9.4 as the pH window sewer codes most often set, and a tartrate developer sits inside it, so pH is not the governing consideration here — the metal is. ILFORD’s guidance for domestic users is to bottle wastes separately, label them and take them to a household waste and recycling centre. Photographers’ Formulary’s own sheet tells the reader to “consult with local sewer and water authorities regarding proper disposal of darkroom chemicals in your area”, which is the right instruction. Local regulation governs, and this page does not give a jurisdiction-specific rule.

The dichromate packet, if your kit has one, is not part of this. It is a chromium(VI) waste and goes through a licensed route.

Troubleshooting

A yellow or brown stain in the areas the negative masked. Residual iron, and it is the classic kallitype failure — Wall calls it “a most prolific cause of failure” in 1912 and Ware finds it in the few surviving hundred-year-old specimens as “pronounced image fading and seriously yellowed highlights”. Develop the full ten minutes rather than pulling the print when the image looks complete; use a bath that has not already taken a dozen prints; and check the clearing bath, which is the step designed to finish the job.

The print looks colder than sepia. Something has put an alkali in the bath. Check the graduate and the funnel, check that you drew from the right bottle, and check the sizing — the sheet says gelatin sizing gives a blue tone where arrowroot gives brown, so a change of paper can look exactly like a change of developer.

A tide-line or a band of different density across the print. The sheet went into the developer in two stages. Wall’s method for a large print is to immerse it face down and turn it over at once, breaking bubbles with a fingertip; for a small one, pour or slide it in a single movement.

The image is weak and got weaker after the fixer. Two different faults with the same symptom. Underexposure is the first candidate, because a kallitype’s density is fixed by how much iron the light reduced and no amount of development will add to it. The second is bleaching in the fixer: Bostick and Sullivan warn that “kallitype prints will bleach in the fix, so either tone before fixing or overprint by a stop or two”, and Photographers’ Formulary caps the fixing at five minutes for the same reason.

The stock will not dissolve, or has crystallised in the bottle. Both are temperature. The dissolution absorbs heat and cools its own water; the sheet says the mixture “turns cloudy but, in time, will clear” and that the bowl may need warming. Warm it and keep stirring. A bottle stored cold will drop crystals and take them back on warming.

The bath has gone cloudy or thrown a dark deposit. Expected, and described under Behaviour. Agitate so it does not settle on the highlights. Cloudiness in the stock bottle is a different matter and suggests silver contamination.

Prints take longer and longer to clear as the session goes on. The developer is loading with iron. Neither this sheet nor Wall publishes a replenishment rate, so the honest advice is the conservative one: mix a fresh 500 mL rather than pushing a tired bath, and keep a note of how many prints each tray took before the stain appeared. That note is data this formulary does not have.

Experiments

Each of these changes one variable and keeps a control. Coat and expose all sheets from a single sensitiser batch on a single paper with a single sizing, print from one negative, and process everything after the developer identically — the clearing bath and the fixer change the print as much as this tray does.

1. The colour ladder. Hypothesis: image colour is set by the proportion of borax to tartrate, not by the amount of tartrate. Control: the sepia bath as published, 48 mL of stock in 500 mL. Variable: the A-to-B proportion. Make up five working developers at 500 mL each: the three the sheet publishes, plus one at 48 mL of A with 32 mL of B and one at 48 mL of A with 96 mL of B. Develop for a matched time, clear, fix and wash identically, and judge the prints dry and side by side under one light. The sheet’s rule predicts a monotonic warm-to-cold sequence ordered by borax; the brown-tone bath, with twice the tartrate of the sepia one, is the test case that separates proportion from amount.

2. Does the tenth minute earn its place? Hypothesis: development beyond the point where the image stops changing removes residual iron and shows up as a cleaner masked border, not as more density. Control: a sheet developed the published 10 minutes. Variable: development time, at 2, 5, 10, 20 and 30 minutes. Mask a strip of each sheet during exposure so every print has a border that saw no light. Clear and fix all five identically and compare the borders, not the images. Wall’s claim and Thomson’s thirty minutes are what is being tested; King’s separate claim, that much of the residual iron leaves in the developer, is the mechanism it would support.

3. Room temperature against warm. Hypothesis: the tartrate bath is used cool because it is slow, not because warmth harms it. Control: 10 minutes at whatever your darkroom’s room temperature actually is — measure and record it, since the sheet does not state one. Variable: bath temperature, one sheet at room temperature and one held at 38 °C in a water bath, both for 10 minutes and both for 5. Record the image colour as well as the density: if warmth alone moves the colour, the account under Image characteristics needs revising.

4. The pH this page could not publish. Hypothesis: the tartrate working bath is mildly alkaline and very poorly buffered, and the borax bath is strongly buffered. Control: freshly mixed sepia developer. Variable: added acid. Measure the pH of the fresh sepia bath, of the fresh black-tone bath, and of each after adding successive small measured volumes of a dilute acid, and plot pH against acid added. The two curves are the buffer argument made visible, and the first number is one the course would like to have.

5. Distilled against tap. Hypothesis: calcium in tap water measurably worsens clearing. Control: the developer as published in distilled water. Variable: the water. Mix a second 500 mL in your tap water and develop matched sheets. Judge by how long each print takes to clear and by the masked border afterwards, and record your water’s hardness from your supplier’s published figure.

📓What to record

Paper and its sizing. Sensitiser batch and coating date. Exposure source and time. Which stock bottles, mixed on which date, and the hydrate on the jar. The working bath’s composition in millilitres of each stock. The actual temperature of the tray, measured. Development time. Clearing time to visually clean. Fixing time. The wet colour and the dry colour, judged under a named light. And the number of prints this tray of developer had already taken — because that is the figure no published source for this bath supplies, and a formulary is only as good as the observations that come back to it.