Chemistry in the Bottle: Oxidation, Exhaustion, Precipitation and Contamination

“The developer was bad” is four different statements, and the four have different causes, different signatures on the film, and different answers. Two of them are irreversible and one is often nothing at all.

• Oxidised: oxygen has spent the developing agents while the bottle stood.
• Exhausted: film has spent them, and the reaction’s own products are now restraining it.
• Precipitated: something has come out of solution, which may or may not matter.
• Contaminated: something else got in.

This page separates them. It matters because a great many faults that look like development faults were created days earlier, on a shelf, by nobody.

Aerial oxidation in storage

Part III owns the chemistry: oxygen is an oxidising agent that you did not add and cannot exclude, the sulfite in a developer is a preservative spent on purpose to protect the agents from it, and when the sulfite is gone the agents are next.

What this page adds is the practical consequence: it is the air in the bottle, not the age of the solution. Oxygen reaches the solution across the surface where the two meet, so a bottle with a centimetre of air over a litre of developer ages slowly and the same bottle half empty ages fast.

The manufacturers publish the arithmetic, and they agree with each other in a way that is worth setting out in one table because it is one of the few numerical patterns in the whole of darkroom keeping:

Solution	Full container	Half full	Ratio
Kodak D-76 stock	6 months	2 months	3×
ILFORD ID-11, MICROPHEN, PERCEPTOL stock	6 months	1 month	6×
ILFORD ILFOSOL 3 concentrate	24 months	4 months	6×
ILFORD BROMOPHEN stock	6 months	3 months	2×
ILFORD ILFOSTOP concentrate	5 years	12 months	5×
ILFORD ILFOTOL concentrate	3 years	12 months	3×

Six products from two makers, and every one of them loses between half and five-sixths of its life to a half-empty bottle. Kodak’s own sentence beside its table says why: “solutions in full bottles have a longer shelf life; partially filled bottles allow some oxidation of the solution.”

Two further numbers from the same sheets are worth having because they extend the argument to a tank and a tray. ILFORD give their stock developers 4 months in a deep tank with a floating lid and 1 month without one — a floating lid being nothing but a way of removing the air-liquid interface. Kodak give D-76 as a working solution 24 hours in a tray and 1 month in a tank with a floating lid. And both makers say the same thing about a diluted developer: ILFORD’s 1+1 and 1+3 solutions “should not be kept for more than 24 hours”, Kodak’s D-76 at 1:1 is “diluted just before you use it” and discarded after one batch.

The colour sequence, and how far it can be trusted. As a metol-hydroquinone developer oxidises it goes from water-clear to straw to brown, because the agents’ oxidation products are coloured while the starting materials and the sulfite’s own oxidation product are not.

✕Colour is a late symptom, and clarity is not evidence

Part III states the course’s position and this page follows it: a stock developer well protected by sulfite can be almost water-clear on the day its protection runs out. The sulfite is spent first, and it is spent invisibly. Only when it is gone do the agents begin to make coloured products in quantity.

So the colour sequence is a real observation and a one-way test: brown convicts, clear acquits nothing.

The course labels the sequence honestly. The general chemistry — that oxidised developing agents are coloured and that the sulfite’s oxidation product is not — is established, and Kodak’s primer describes the coloured oxidation products of pyro and their deposition in the film. The specific clear-to-straw-to-brown sequence for a metol-hydroquinone developer is stated here as practitioner observation consistent with that chemistry, not as a figure read off a manufacturer’s sheet, because no sheet in this course’s corpus publishes it.

The smell is not a keeping test at all. A sharp smell from a fixer or a stop bath is sulfur dioxide from acidified sulfite, and Part III treats it as a ventilation warning rather than as evidence about the bath’s condition.

Oxidised or exhausted: two signatures on the film

This is the distinction the page exists for, and Part XXVII’s exhaustion experiment measured it rather than asserting it. Its finding, in the shape of the curves:

Sequential use — real exhaustion — lowers contrast index steadily. Each film through the bath consumes agent and releases bromide, and both effects accumulate. The contrast index falls gradually and roughly in proportion to the loading, so the failure is a slope. Added bromide, separately, holds back the toe and costs speed while leaving the straight line nearly alone — which is why bromide accumulation and agent consumption are two different signatures even though they arrive together.

Oxidation is a plateau followed by a collapse. An oxidising bath tracks a fresh one closely while the sulfite still holds, and then falls away sharply. The failure is a knee, not a slope.

That difference is diagnostically useful in a way a number is not. Ask what the last four films looked like, not just this one:

• Contrast drifting down, film by film, over a bath’s life → exhaustion, and the answer is a capacity log and Kodak’s 15 per cent time increase after every four rolls per gallon.
• Three normal films and then one that failed abruptly, from a bath that has been standing → oxidation, and the answer is a fresh bottle.

Oxidation can also stain, because oxidised agents are coloured and are deposited along with the silver. Exhaustion does not stain; it just does less.

=What the colour and the smell can tell you before you commit a film

They can tell you to stop. They cannot tell you to proceed.

A brown developer is disqualified and you have saved a film for the price of looking. A clear developer of unknown age is not qualified, for the reason above.

The test that does qualify a bath is a clip test: a scrap of the same film, exposed to room light, developed in the suspect bath for the normal time. It should go to full black. If it does not, the bath is not delivering. It costs a few centimetres of film off the leader and thirty seconds, and it is the single most useful habit on this page.

A clip test does not measure contrast, so it cannot detect a bath that has drifted 15 per cent low — that needs a control strip and a densitometer, which is Part XIV’s subject. It detects the failure that ruins a film, and that is the one you are asking about.

Precipitation and crystallisation

A stock solution in a cold room throws crystals. Solubility falls with temperature, and a concentrated stock made warm and stored at 5 °C can put a layer of crystals on the bottom of the bottle. Carbonate and borax are the usual candidates in a classic developer, because both are present near their solubility limits.

The question that matters is whether it will redissolve. Warm the bottle gently to 20 °C, stand it for an hour and swirl it. If the crystals go back into solution and the volume is unchanged, nothing has happened chemically and the bath is what it was. If they do not, or if there is a fine haze or a sludge rather than crystals, something has changed and the bath’s composition is no longer the formula.

Two things that are not this. ILFORD note that when their powder developers are made up, “it is normal for a few grains of powder to remain un-dissolved”, and that “discoloured or darker particles” in the white powders are normal and will not affect development. Neither is a fault. Reading them as one and discarding a good litre is a real and avoidable waste.

The dissolution order, and what the standard explanation gets wrong

Every classic formula gives its ingredients in an order and says to follow it. The usual explanation offered for metol going in before the bulk of the sulfite is that the sulfite would otherwise let the metol oxidise. Kodak’s own primer gives a different reason, and it is worth correcting because the usual one is repeated everywhere and is rarely sourced.

⚗It is a solubility effect, not an oxidation effect

The primer’s general rule is indeed that “the preservative should be dissolved first” — to protect the agent from oxygen. It then states an exception for metol (which Kodak called Elon), and the exception is where the received wisdom comes from and where it goes wrong.

Metol, says the primer, is “readily soluble in warm water (about 125 °F) (52 °C)” and does not oxidise rapidly. So the reason for putting it in first is not oxidation. It is this: “if the sulphite is dissolved before the Elon, as is the case with developers such as hydroquinone, a white precipitate often appears especially if the sulphite solution is concentrated. This precipitate forms because Elon is a combination of an insoluble base with an acid which renders it soluble. When the acid portion is neutralized by a weak alkali such as sodium sulphite, the insoluble base is precipitated.”

Metol is a salt — the sulfate of a base that is not itself soluble. Sulfite is a weak alkali. Put the salt into a strong sulfite solution and you liberate the free base, which comes out as a white precipitate.

The asymmetry is the practically useful part: “when once the Elon is dissolved, it takes a fairly high concentration of sulphite to bring it out of solution again, though only a low concentration of sulphite is required to prevent the Elon from dissolving.” Getting it in is the hard part; keeping it in is easy. Hence the compromise the primer records as common practice — dissolve a little of the sulfite first to protect the metol, then the metol, then the rest of the sulfite.

And the reassurance nobody passes on: if a precipitate does form, “this will usually redissolve on adding the carbonate and no harm has been done”, because the base forms a soluble sodium salt with carbonate.

A separate rule about the alkali, which is often merged with the one above and should not be. The primer: dissolve each chemical completely before adding the next, because “if the alkali is added before the crystals of the developing agent are dissolved, each crystal becomes oxidized at the surface and the resulting solution will give fog.” That one is an oxidation rule, it is about the alkali rather than the sulfite, and its consequence is fog rather than a precipitate.

Mixing temperature is specified for a reason and the reasons differ by ingredient. The primer’s 52 °C for metol is a solubility figure. ILFORD’s modern instruction for their powder developers is to dissolve part A in warm water “at about 40 °C/104 °F”, add part B gradually while stirring, then make up with cold water and allow to cool to 20 °C before use. Too cold and the powder does not go in; too hot and you accelerate the oxidation you are trying to avoid, and in a hardening fixer you precipitate the alum.

One more instruction from the same sheet, easy to skip and easy to follow: “as most water drawn from pressure mains is highly aerated, we advise that users draw off the water they need and leave it to stand for a few minutes before using it to make up developers.” You are dissolving a preservative whose job is to consume dissolved oxygen. Letting the water stand costs nothing and spends less of it.

Hard water and sequestering agents

Calcium and magnesium in the supply react with carbonate and with sulfite to give insoluble salts. The results are a sludge in the bottle, a scum on the material, and — where the precipitate forms in the emulsion rather than on it — a fault that does not wash off.

The manufacturers’ answer has been the same for eighty years and comes in two parts. For the solution, a sequestering agent: Kodak Limited’s 1944 formulary lists Calgon among its products, “for addition to developing solutions to prevent the precipitation of calcium salts when using hard water”. A sequestrant works by binding the calcium and magnesium ions into a soluble complex so that they are still in the bath but no longer available to precipitate. For the final rinse, distilled water: the same formulary instructs that where the water is very hard and tends to deposit a scum on drying, “give the final rinse in distilled water containing this addition of wetting agent”, and the modern Kodak processing sheets say the same in one line.

When distilled or deionised water is worth the cost is a judgement, and it is a smaller cost than it looks because the volumes differ enormously. A litre of developer needs a litre of water; a final rinse needs a few hundred millilitres; a wash needs tens of litres. So the sensible order of spending is rinse first, then mixing, and never the wash — the wash is the one place where the water’s job is to be plentiful rather than pure.

!What the course cannot tell you about your water

Water hardness is local and this course has no way of knowing yours. The manufacturer sheets in the corpus name hard water as a cause and prescribe distilled water and sequestrants as the answer, but none of them states a hardness figure above which the problem appears, and the course will not invent a threshold.

What you can do instead costs one session: mix one litre with tap water and one with distilled, process two identical strips, and look at them dry. If you cannot tell them apart, your water is fine for mixing and you have settled the question for your address. If you can, you have found a real difference and you know what it is worth.

Contamination, listed by direction

The direction matters, because the consequences are entirely different.

Fixer into the developer. The worst of the four, and out of all proportion to the quantity. A trace of thiosulfate in a developer acts as a silver-halide solvent in a bath that is trying to reduce silver halide, and the result is fog, stain and a drop in activity together. ILFORD’s first-film guide is blunt: buy beakers in different colours, use a different one for each solution, because “even a trace of fixer can contaminate the developer, and possibly ruin your next film”. Kodak’s Z-133 lists “developer contaminated with fixer or stop bath” among the causes of an abnormally light film, and its corrective action is not to adjust anything: replace the contaminated developer, and wash the mixing equipment thoroughly before use.

Stop bath into the developer. Acid into an alkaline bath lowers the pH, and pH is what sets a developer’s energy — Kodak’s primer states that the quantity of alkali governs the energy. The result is a slow, low-contrast developer, and the effect is proportional to how much got in rather than catastrophic for a trace. It is the least damaging of the four and the easiest to detect, because a pH meter or even an indicator paper will show it.

Developer into the fixer. This one is unavoidable in principle — every film carries some — and the question is only how much. Carried-over alkaline developer neutralises the fixer’s acid, which shortens its life, and it is the specific condition the fixing page names for dichroic fog and for aluminium sulphite sludge. The control is a rinse or a stop bath, which is what a stop bath is for: ILFORD’s own sheets describe it as reducing carry-over of excess developer into the fixer and so prolonging the fixer’s life.

Fixer onto dry paper before exposure. A splash, or a hand that has been in the fixer and then in the paper box. Where the drop landed, the halide is dissolved before any exposure or development can happen, so the print has a white spot that no exposure can fill. It is not a bath fault at all — it is a bench fault — and it is unrecoverable in the print.

How contamination actually gets in

Kodak’s Z-133 names four routes, and the list is worth reading as a description of a home darkroom rather than a laboratory:

1. Mixing equipment that has not been thoroughly cleaned.
2. Dry chemicals that become airborne during mixing and settle in an adjacent solution.
3. Pipes and tanks made of material that reacts chemically with some solutions.
4. Solution splashed or dripped into another solution.

Route 3 is the one people forget, and Kodak’s 1928 primer gives the specifics: “tin, copper, and zinc, or alloys of these metals will usually produce bad fog and stain with photographic developers, and are also unsatisfactory for use in fixing baths.” A galvanised bucket is zinc. A brass fitting is copper and zinc together.

To which a home darkroom adds its own list:

• A shared graduate, measured out for the fixer and then for the developer.
• A tong used in two trays, which also puts fixer on prints as tong marks.
• A funnel not rinsed between decanting the fixer and decanting the developer.
• A thermometer moved between baths without a rinse — small, but it goes into every bath, every session.
• A bottle refilled from the wrong jug, which is a labelling failure rather than a chemistry one.
• A wet work surface, on which a bottle base picks up whatever the last spill was and carries it to the next shelf.

Two published procedural controls close most of that off. ILFORD’s SIMPLICITY instructions: “to minimise problems with cross contamination, make the solutions in order Developer > Stop > Fixer. Wash any measuring cylinders / stirrers in-between each make and after use.” Making the cleanest bath first means that whatever traces survive a rinse travel in the direction that costs least: a trace of developer in the fixer shortens the fixer’s life, while a trace of fixer in the developer is the fault the makers tell you to discard the bath over. And their chemical sundries sheet: wash all utensils and vessels after use, and “wherever possible use dedicated equipment for making up developer solutions”. Kodak’s Z-133 says the same in a laboratory register — use separate mixing tanks for developers and fixers.

!A number the course does not have

The manifest for this page asked for the quantity of thiosulfate contamination that measurably affects a developer, so that the page could say how small a trace matters rather than implying that any trace is fatal.

No source in this course’s corpus states one. ILFORD say “even a trace”; Kodak say to replace the developer and wash the equipment; neither publishes a concentration, a volume or a threshold.

So the page states the qualitative warnings as the manufacturers state them, and declines to invent a figure. What follows practically is not weaker than a number would have been: because the threshold is unknown, the sensible practice is the one that does not depend on knowing it — dedicated vessels, the mixing order above, and a clip test on any developer you have reason to doubt. If you want the number for your own darkroom you can bracket it: add a measured 0.1 mL, 1 mL and 10 mL of working fixer to three 500 mL portions of fresh developer, run a strip through each and a fourth as a control, and read the base plus fog. That is a Part XXVII experiment and the course would be glad of the answer.

The tests a home worker can run, and what each is worth

Five checks, cheapest first, and what each one can and cannot settle

1. 1Colour, against a white card, in one lightFree, instant, one-way. Brown convicts. Clear acquits nothing, because sulfite protection fails before colour appears.
2. 2pH, against a calibrated meter or a narrow-range paperDetects acid contamination and gross mis-mixing. ILFORD publish fresh figures for their own stock developers — PERCEPTOL 7.68 to 7.82, ID-11 8.60 to 8.70, MICROPHEN 8.67 to 8.93 — and add that users should measure their own fresh solutions for later comparison, because published laboratory figures may differ from yours.
3. 3A clip test on a scrap of filmA few centimetres off the leader, exposed to room light, developed for the normal time. Should reach full black. Detects a bath that is not delivering; does not detect a bath that has drifted 15 per cent.
4. 4A clearing-time test on the fixerA scrap of film dropped on the bath. Discard when the clearing time exceeds twice the fresh value — which you must have measured when the bath was fresh, because it cannot be recovered later.
5. 5A control strip, read on a densitometerThe only one of the five that measures contrast rather than detecting failure. Part XIV builds the apparatus and the discipline; it is what turns "seems all right" into a number with a tolerance around it.

The order is by cost, and the first four cost almost nothing. Note that each detects something the one above it cannot: the ladder is not four ways of doing the same test.

Storage and labelling, which is the real prevention

Full bottles. The table at the top of this page is the argument. Where a stock will be used over months, decant it into several small bottles filled to the neck rather than leaving one large bottle to empty gradually.

Small bottles, for the same reason, and glass or a rigid plastic that does not breathe.

Dated labels. The date the bottle was made up, not the date it was bought — those are different numbers and only one of them starts the clock.

A stated shelf life recorded from the manufacturer rather than from hope. Write the maker’s figure on the label beside the date, so that the bottle answers the question without you having to remember which sheet it came from. Kodak’s Z-133 gives the principle in a sentence: “to minimize solution storage and keeping problems, mix only the amount of solution that you will use during the recommended keeping time.”

Dedicated vessels per bath, marked, and ideally in different colours as ILFORD suggest.

✦The one-shot decision, made on evidence rather than on thrift

ILFORD’s own account of reusing a stock developer is unusually candid and worth having in full, because it names four distinct penalties rather than one: reuse “lowers image quality slightly and increases the risk of physical damage”; the developer “oxidises with reuse and storage”; “the risk of contamination is increased”; “precipitates may be formed”; “tiny particles of emulsion from the films processed previously may be held in suspension”; and there is “a risk of miss counting the number of films that have been processed”.

Their conclusion: one-shot processing “is recommended when image quality, reliability and consistency are more important than economy”, and diluted developers should never be reused.

That is a manufacturer telling you what reuse costs, in their own product, so that you can decide. Much of this page becomes irrelevant if you work one-shot — which is exactly why the recommendation exists.

The honest ending

Some bottles are discarded. That is not a failure of technique; it is the price of a solution that reacts with the air.

The arithmetic is worth doing once. A litre of film developer is a few pounds. A roll of film is a few pounds, and the pictures on it are usually worth more than either. A session that begins with a doubtful developer usually ends with a lost film that costs more than the chemistry, and it costs an evening as well.

So the decision rule is short: if the bottle is past the maker’s stated life for the way it has been kept, or if a clip test does not go to black, pour it away. Do not compromise, do not extend the time to compensate for a bath you have not measured, and do not put a film you care about through a bath you are arguing with.

Atlas entries this page owns

Seven entries in the troubleshooting atlas belong here: oxidised developer, exhausted developer, precipitate in a stock solution, undissolved chemical, developer contaminated by fixer or stop, fixer contaminated by developer and hypo spots on unexposed paper.

The chemistry of oxidation is Part III’s, the preservative is Part VIII’s, storage and incompatibility rules are Part II’s, and the measured curves for exhaustion against oxidation are Part XXVII’s. This page owns the signatures and the decision.

Recap

Four failures share one name. Oxidation is spent by air across the surface of the liquid, so a half-empty bottle costs between half and five-sixths of a solution’s published life, and it fails as a plateau followed by a collapse. Exhaustion is spent by film, and it fails as a steady slope in contrast index. Colour is a one-way test — brown convicts, clear acquits nothing — and the smell is a ventilation warning rather than a keeping test. A crystal in a cold bottle may be nothing; the question is whether it redissolves. Metol goes in before the bulk of the sulfite for a solubility reason and not an oxidation one, and Kodak’s own primer says so and adds that a precipitate usually redissolves when the carbonate goes in. Hard water precipitates with carbonate and sulfite, and the answers are a sequestrant in the bath and distilled water in the final rinse. Contamination is worst in the direction fixer-into-developer, where the manufacturers say to discard rather than adjust; the published controls are dedicated vessels, mixing developer first and fixer last, and washing between. And the honest ending is that a doubtful bottle costs less to replace than the film it will spoil.

Check your understanding

Question 1. A developer mixed three weeks ago is brown and smells. What would you test before using it, what result would let you proceed, and what would you do with the bottle if it fails?

• Test the pH; if it is within a tenth of the published figure, proceed
• Nothing needs testing: brown convicts on sight. The bottle is disqualified, and the useful question is why — three weeks is well inside Kodak's six months for a full bottle of D-76 stock and ILFORD's six for ID-11, so either the bottle was half empty (where the same makers give two months and one) or it was warm, or it was contaminated. Pour it away, record which of those it was, and change that.
• Run a clip test; if the strip reaches black, the developer is fine and can be used
• Dilute it further and extend the development time to compensate

Show the answer and why

Answer: Nothing needs testing: brown convicts on sight. The bottle is disqualified, and the useful question is why — three weeks is well inside Kodak's six months for a full bottle of D-76 stock and ILFORD's six for ID-11, so either the bottle was half empty (where the same makers give two months and one) or it was warm, or it was contaminated. Pour it away, record which of those it was, and change that.

The point of the colour test is that it is one-way: it can disqualify but cannot qualify. Once it has disqualified, no further test is worth running on that bottle, because the answer will not change what you do. What is worth doing is the diagnosis of the *storage*, since three weeks is far short of every published figure for a full container and a long way inside even the half-full ones — so something specific happened, and it will happen again to the next bottle unless it is found. The smell, incidentally, is not the evidence: a sharp smell is sulfur dioxide and is a ventilation matter.

Question 2. Why does a few millilitres of fixer in a litre of developer cause damage out of proportion to its concentration?

• Because thiosulfate is a strong acid and collapses the pH
• Because thiosulfate is a silver-halide solvent, so it dissolves halide in a bath whose whole function is to reduce halide selectively — attacking the unexposed grains as well and producing fog, stain and lost activity together, rather than simply diluting the developer by half a per cent
• Because it precipitates the sulfite
• Because it raises the temperature of the bath

Show the answer and why

Answer: Because thiosulfate is a silver-halide solvent, so it dissolves halide in a bath whose whole function is to reduce halide selectively — attacking the unexposed grains as well and producing fog, stain and lost activity together, rather than simply diluting the developer by half a per cent

It is a difference in kind rather than in degree, which is what "out of proportion" means here. A half per cent dilution of a developer is nothing; a half per cent of a substance that dissolves the material the developer is selecting between attacks the selectivity the whole process depends on. Note the manufacturers' corrective action, which tells you how they regard it: Kodak's Z-133 does not offer an adjustment, it says replace the developer and wash the mixing equipment. The course also states plainly that no source in its corpus publishes a threshold quantity, so "even a trace" is as precise as the evidence allows.

Question 3. Two negatives are both flat. Film A came from a bath that has processed six rolls, film B from a bottle that has stood half-empty for four months and processed nothing. What difference in the record separates the two causes?

• There is none; both are simply weak developers
• The shape of the failure over time. Exhaustion falls gradually — the four films before A would each have been slightly flatter than the last — while oxidation holds close to fresh and then collapses, so the films before B would have looked normal and B failed abruptly. Part XXVII measured both shapes.
• The colour of the two solutions, which is decisive
• The pH, which falls in one case and rises in the other

Show the answer and why

Answer: The shape of the failure over time. Exhaustion falls gradually — the four films before A would each have been slightly flatter than the last — while oxidation holds close to fresh and then collapses, so the films before B would have looked normal and B failed abruptly. Part XXVII measured both shapes.

This is why the diagnostic question is "what did the last four films look like?" rather than "what does this one look like?". A single flat negative is consistent with either cause; a sequence is not. Colour helps only in one direction and only if the oxidised bath has gone far enough to show it. The practical consequence differs too: exhaustion has a published compensation — Kodak's 15 per cent time increase after every four rolls per gallon — and oxidation has none, because what has been lost is not recoverable by more time.

Question 4. What does Kodak's 1928 primer give as the reason for dissolving metol before the bulk of the sulfite, and how does it differ from the explanation usually offered?

• The same reason: sulfite would let the metol oxidise
• A solubility reason rather than an oxidation one. Metol is the salt of an insoluble base with an acid that makes it soluble; sulfite is a weak alkali which neutralises the acid part and precipitates the base. The primer states that metol is readily soluble in warm water and does not oxidise rapidly — and adds that a precipitate which does form usually redissolves when the carbonate is added, with no harm done.
• That metol dissolves only above 52 °C and the sulfite lowers the temperature
• That the sulfite reacts with metol to form a coloured product

Show the answer and why

Answer: A solubility reason rather than an oxidation one. Metol is the salt of an insoluble base with an acid that makes it soluble; sulfite is a weak alkali which neutralises the acid part and precipitates the base. The primer states that metol is readily soluble in warm water and does not oxidise rapidly — and adds that a precipitate which does form usually redissolves when the carbonate is added, with no harm done.

The usual explanation attaches the general rule — dissolve the preservative first, to protect the agent — to the one ingredient the primer names as an exception to it. Getting the mechanism right changes the practice: it explains why a pinch of sulfite before the metol is a workable compromise (a low concentration protects without preventing dissolution), why the order matters much less once the metol is in, and why a white precipitate at the mixing stage is usually recoverable rather than a ruined batch. A separate rule, often merged with this one, really is about oxidation: adding the alkali before the agent's crystals have dissolved oxidises each crystal at its surface and gives a fogging solution.

Check my answersClear