Assignment: Assessing a Historical Process for Hazard

Five lessons in this part handed you five finished verdicts. Somebody had already read the classifications, already found the exposure limits, already decided which single substance settled the argument, and already written down what a reader would need to have for the procedure to be reasonable. That is the easy half, and it is easy because the answers were not in doubt.

The situation this assignment is written for is the other one. A reprint of a nineteenth-century manual falls open at a formula nobody has classified. A supplier’s list carries a salt whose period name you half recognise. A blog reproduces a bath from 1888 without a word about what is in it. In none of those cases will a page arrive to tell you the answer, and the useful skill is not a list of forbidden words — it is a method that produces a level, with a named reason, from evidence you can show somebody else.

That method is what you practise here. Three period procedures are set, each with a page reference into this course’s research corpus. You assess each one, classify it against the course’s A to D rubric, name the single item that decided it, and say what would have to be true for the level to fall. Then you write one process-atlas entry to the course’s own format, and identify three museum objects from their surfaces. The whole of it is reading, arithmetic and writing: no darkroom, no bench, no chemicals and no facility of any kind is needed to complete this part, which is the one place in this course where a reader without a laboratory is at no disadvantage whatever.

What you are asked to produce

Four documents, and one prohibition that governs all of them.

1. Three hazard assessments, one per set procedure, on the record described below.
2. One process-atlas entry for a process from this part, written to the course’s format before you read the published entry, and then compared against it.
3. Three identifications of objects in an open-access museum collection, with the reasoning written out and the rights metadata recorded.
4. Optionally, a fourth assessment on a procedure you found yourself. This is where the marks for distinction are, because a procedure nobody has classified for you is the real case.

☠The prohibition, and it is the reason this is a Level D assignment

No deliverable may contain a working procedure. Not a quantity to weigh. Not a working strength or a dilution. Not a temperature and a time. Not a sequence of steps in an order that could be followed. Not a substitution offered as making something safer. This applies to all four procedures, including the one you judge least hazardous and including the one this course publishes in its own formulary as something a reader may mix.

Two reasons, and the second is the one people miss. First, a partial procedure is worse than none: it reads as permission while omitting the controls that would have mattered. Second, an assessment is a classification document, not a working document, and the moment quantities appear in it the reader of the document cannot tell at a glance which procedure they belong to. A submission containing a procedure is returned unread. That rule is in the marking scheme below, with the rest of the scheme, because a prohibition nobody can see the enforcement of is only a wish.

The rule is enforced on this course’s own pages by machine as well as by editing: npm run validate:labs fails the build if any page whose front matter says safetyLevel: D carries a Procedure, Preparation or Build heading. The full policy is the Level D page, and it is worth reading before you write, because it also says what a Level D document may contain, which is a great deal.

The method, as a repeatable sequence

One assessment, in the order the steps have to happen

1. 10. Translate the nomenclaturePeriod names first, before anything else, because you cannot look up a classification for "bichromate of potash" and because two period names sometimes point at the same modern substance and sometimes do not
2. 21. List every substance the procedure namesIncluding the ones mentioned in passing, the ones in a variant offered as an alternative, and the products of any step that brings two of them together. The list is of substances, not of bottles
3. 32. Find each one’s current classification and exposure dataA GHS classification, a workplace exposure limit or a supplier’s safety data sheet, from a named source, with the date you read it. The period manual is evidence of what was believed and is never evidence of what is true
4. 43. Identify the energiesHeat, open flame, ultraviolet, mains electricity, pressure, and mechanical energy such as grinding. A procedure with no energy input is a real finding and should be written down as one
5. 54. Identify the waste streamsWhat is in the tray, the filter and the bottle when the work is finished, and whether any lawful domestic route exists for it where you live
6. 65. Identify the failure modesNot what the procedure does when it works. What a splash, a spill, a wrong bottle, an acid in the wrong tray, a broken vessel or an unventilated hour would actually do to a person
7. 76. Classify on the worst single item, and name itThe level is the highest criterion met by anything in the procedure. Never an average, never a majority. An assessment that does not name the deciding item has not finished
8. 87. Ask what would move it down, and whether you could assemble thatA named control against each hazard, then the honest question of whether it is available to you and what facility would be needed if it is not
9. 98. Answer the historical halfWhy was the procedure used, what problem did it solve, and what solves that problem now. The modern descendant is named with the lesson that teaches it, or the assessment is incomplete

Step 0, which is where most assessments go wrong

You cannot search a classification database for a name that has not been used since 1890, and a period manual will not translate itself. Worse, the translation is occasionally not one-to-one: the period word cyanide covers both the free salt and the complex hexacyanoferrates that this course handles at Level A, and those two are not the same hazard at all. Do this step first, in writing, and carry the modern name and the CAS number forward.

What the manual prints	What you look up	Where the course’s entry is
bichromate of potash, potassium bichromate	potassium dichromate	potassium dichromate
hyposulphite of soda, hypo	sodium thiosulfate	sodium thiosulfate pentahydrate
sulphocyanide of ammonium	ammonium thiocyanate	ammonium thiocyanate
cyanide of potassium	potassium cyanide	potassium cyanide
yellow prussiate of potash	potassium ferrocyanide, hexacyanoferrate(II)	potassium ferrocyanide
red prussiate of potash	potassium ferricyanide, hexacyanoferrate(III)	potassium ferricyanide
corrosive sublimate, bichloride of mercury	mercury(II) chloride	mercury(II) chloride
nitrate of uranium	uranyl nitrate	uranyl nitrate hexahydrate
muriatic acid, spirit of salt	hydrochloric acid	hydrochloric acid
protosulphate of iron	iron(II) sulfate	iron(II) sulfate heptahydrate
lunar caustic, nitrate of silver	silver nitrate	silver nitrate
pyroxyline, gun-cotton	nitrocellulose	nitrocellulose
chloride of gold	gold(III) chloride	gold(III) chloride

Every row of that table is taken from the alternative names carried on the course’s own chemical encyclopaedia entries, which is where to go when a name is not on it. Where a period name resolves to nothing you can find, that is a finding: record it as an unidentified substance and treat the procedure as unassessable until it is identified, because you cannot classify what you cannot name.

Step 1, and the half of it people forget

List the substances the procedure names, then list what any two of them make when the procedure brings them together. A bath is not only what went into it; it is also what is in it afterwards, and the products are frequently the interesting part. Three that recur in this corpus:

• Free cyanide plus any acid gives hydrogen cyanide gas, at once, with no heating. Kodak Limited printed that warning in 1949 on the same page as a formula that offered a cyanide bath.
• Complex cyanides — the ferri- and ferrocyanides — are not free cyanide, but Princeton’s guidance for photography states the conditions under which they release hydrogen cyanide anyway: heat, hot acid, or strong ultraviolet, with poisonings on record from acidified Farmer’s reducer.
• Thiosulfate plus a strong acid decomposes, precipitating sulfur and releasing sulfur dioxide, which Kodak’s 1928 primer sets out in its chapter on fixing.

The incompatibility matrix carries the course’s own pairs, and chemical storage and incompatibilities is the lesson that taught you to read them.

Steps 2 and 3, where the evidence standard bites

For each substance: the notified or harmonised GHS classification, the hazard statements with their agreement figures where the source gives them, any workplace exposure limit, and the date you read it. Three sources cover almost everything in this corpus: the ECHA Classification and Labelling Inventory as aggregated on PubChem, HSE’s EH40 table of workplace exposure limits, and the NIOSH Pocket Guide. A supplier’s safety data sheet beats all three for a product you actually hold, and recording a safety data sheet is the procedure for capturing it so the citation survives a year.

Read an exposure limit as a measure of how much control the substance is judged to need, not as a concentration you are invited to work up to. And read EH40’s own caution as part of the method: the absence of a substance from its list does not indicate that the substance is without risk. Where the corpus holds no hazard record at all, say so in the assessment; a gap is a finding, and a hazard nobody has characterised is one against which no control can be specified.

Energies are the step that gets skipped because the answer is often none. Write the none down anyway. It is what allows the mercury toning bath below to be classified on its chemistry alone, and it is what distinguishes that bath from the wet-plate darkroom, whose governing hazard is a volatile flammable solvent rather than anything in a bottle.

Steps 4 and 5

Waste is the step that most often changes the answer, because a chemical hazard usually has a named control and a waste problem frequently has none. Ask what is in the tray at the end, what is on the filter paper if you tried to precipitate it out, and whether a lawful route exists from your own address. GOV.UK routes household hazardous waste to the local authority’s collection service rather than to the bin or the drain, the Environment Agency’s WM3 guidance carries the photographic waste codes, and what may lawfully be done differs between jurisdictions and changes. The disposal caveat is not a formality: the chemistry in this course is general and the law where you are is local, and you must check your local regulations.

Failure modes are what an assessment is actually for. A procedure that works as written harms nobody; the question is what the third-worst afternoon looks like. Splash on skin, splash in an eye, a bottle knocked over, a bath poured into the wrong tray, an unlabelled container found by somebody else, an hour longer over the bench than intended, and — the one this part keeps returning to — an acid two trays away from something that must never meet one. The unlabelled container procedure exists because that last case is common rather than exotic.

Step 6, and the rule that makes the whole thing work

!The worst single item decides it, and it has to be named

A procedure’s level is the highest level any one of its items reaches. Not an average, not a majority verdict, not a judgement about the procedure’s general feel. One potassium dichromate in an otherwise Level A bath makes the whole bath a chromium(VI) operation, and nothing else in it can bring that back down.

The second half of the rule is procedural and is worth as many marks as the first: name the item. An assessment that concludes “Level D” without saying which substance or which step produced the D cannot be checked, cannot be argued with, and cannot be revised if better data appears. Naming it is also what makes the assessment useful later — because when the deciding item is removed or replaced, you already know what to reassess.

Steps 7 and 8

Step 7 turns the verdict into something actionable: for each hazard, what is the recognised control, and could you assemble it? “Be careful” is not a control. Extraction is a control; a written procedure and a second competent person is a control; a licensed waste contract is a control. Where the answer is no, name the facility that would be needed, because that is the difference between Level C, where somebody can do this properly and it is not you at home, and Level D, where the controls are not ones a page can assume anybody reading it has.

Step 8 is the historical half, and it is what turns the exercise into photography rather than paperwork. Nobody used these procedures out of carelessness. Each solved a problem: a plate that had to be finished before the collodion dried, a negative too thin to print in an era with no variable-contrast paper, a blue line drawing wanted as a positive rather than a negative. Say what the problem was, and then say what solves it now, naming the lesson in this course that teaches the modern route. For a surprising number of these the modern descendant is not another bath at all: it is a paper grade, a different developer, or a decision made at exposure.

The record you keep, in one format and not a fourth one

Every field below already exists somewhere in this course. Nothing here is a new form.

Block	Fields	Taken from
Identification of the source	Author, title, edition, year, page reference, where you read it, date read	The provenance discipline of the formulary, which records ref, section and retrieved for every claim
One row per substance	Period name, modern name, CAS number, form or hydrate, classification with signal word and statements, exposure limit, source, version or revision date, date you read it	Step 4 of recording a safety data sheet, which names exactly those six document fields, plus the CAS number and hydrate from its step 2
Products and pairs	What any two named substances make, and under what condition	The incompatibility matrix
Energies	Heat, flame, ultraviolet, mains, pressure, grinding — each present or absent	The requiresDarkroom, requiresUV and requiresMains flags every page in this course carries
Waste	What is left, which stream it belongs to, whether a lawful domestic route exists where you are	Chemical waste and silver waste
Failure modes	The event, the consequence to a person, the control that would prevent it	The If a step fails sections of the SOP library
Verdict	Level, the single deciding item, the criterion of the rubric it met	The safety classification
Controls	Per hazard: the recognised control, available or not, facility needed if not	The controls array of each level in src/data/safety-levels.json
The historical half	Problem solved then, modern descendant now, with the lesson slug	The Why we study it and do not reproduce it section of every process-atlas entry
Signature	Your name, the date, and what you could not find out	The laboratory notebook convention that an entry records its own gaps

The last field is not decoration. An assessment that does not say what it failed to establish is indistinguishable from one that established everything, and the two are very different documents.

A worked assessment, so that you can see the shape of one

The procedure worked here is not one of the three you are set. It is Wall’s mercury toning bath for bromide and gaslight prints, Photographic Facts and Formulas, 1924, page 243: a mercuric chloride and potassium bromide bleach, followed by a second bath chosen for the colour wanted, of which Wall offers several — greyish-black, grey-violet, brown to violet-black and brownish-violet. Wall closes the section with the observation that all mercury toning gives intensification, which tells you something about the mechanism before you have looked anything up.

Step 0. Corrosive sublimate and bichloride of mercury are mercury(II) chloride, HgCl₂, CAS 7487-94-7. Bromide of potassium is potassium bromide. The blackening baths name ammonia, sulfite, a sulfide and — in the related intensifier on page 130 — silver dissolved in potassium cyanide.

Step 1. The bleach is a displacement, and its products matter more than its ingredients:

2 Ag + 2 HgCl₂ → 2 AgCl + Hg₂Cl₂

What is in the tray afterwards: silver chloride, mercury(I) chloride, and unreacted mercury(II) chloride in the bath

Nothing has been removed from the sheet — which is why Wall’s closing sentence is true, and the mechanism is Part XX’s to explain. For the assessment the point is narrower: the used bath contains mercury in two oxidation states and dissolved silver, and the print itself now contains mercury. Both are waste questions.

Step 2. Mercury(II) chloride’s notified classification, aggregated from 218 reports across 12 notifications to the ECHA inventory, is signal word Danger with H300, fatal if swallowed, in 100 per cent of reports, H310, fatal in contact with skin, in 26.6 per cent, and H314, H317, H318, H341, H361, H372 and H410 alongside. HSE’s EH40 gives mercury and its divalent inorganic compounds, naming mercuric chloride specifically, a long-term limit of 0.02 mg/m³ measured as mercury, with a biological monitoring guidance value attached. NIOSH sets a ceiling of 0.1 mg/m³ for mercury compounds with a skin notation, an immediately-dangerous-to-life value of 10 mg/m³ as mercury, and a symptom list that is largely neurological — tremor, insomnia, irritability, indecision, headache and lassitude — with the central nervous system and the kidneys among the target organs. Read 2026-09-06.

⁇The period sentence that this assignment exists to teach you to handle

On page 130 Wall states that mercuric chloride is extremely poisonous when taken internally, but that absorption by the skin, even in the case of cuts and abrasions, is “practically harmless”. The second half of that claim is not supported by anything in the current record: the notified classification carries H310, fatal in contact with skin, and NIOSH attaches a skin notation to the whole family.

What to do with a sentence like that is a defined operation, not a matter of taste. It is evidence of what was believed in 1924, and as such it belongs in the assessment, because it explains how the procedure came to be worked the way it was. It is not evidence of what is true, and it may not be carried into the hazard section. Every assessment you write will meet a sentence of this kind, and the marking scheme awards for spotting one, quoting it, and saying which of the two things it is.

Step 3, energies. None. Room temperature, no flame, no ultraviolet, no mains equipment, no grinding. That is a real finding and it is worth writing: this procedure is dangerous entirely because of what is dissolved in the tray, which makes it the cleanest possible test of whether you have assessed the chemistry properly rather than the atmosphere.

Step 4, waste. The used bleach carries mercury and dissolved silver. There is no domestic treatment, and precipitating the mercury out moves the problem onto a filter paper rather than solving it. In England and Wales this is household hazardous waste and goes to the local authority’s service; the Environment Agency’s WM3 guidance lists photographic bleach and fixer solutions among the hazardous entries in its chapter 09. Elsewhere, check your local regulations.

Step 5, failure modes. A splash onto broken skin, against a statement that says fatal in contact with skin. A tray tipped into a domestic drain. The light-sensitive bleach solution — Wall says to keep it in the dark — left in an unlabelled bottle for somebody else to find. And the variant on page 130 that blackens with silver dissolved in potassium cyanide, which puts free cyanide in a room that also holds an acid stop bath.

Step 6, the verdict. Level D. The deciding item is mercury(II) chloride, and the criterion it meets is the first Level D criterion in the rubric. What decides it is not the H300 — plenty of Level B substances are toxic if swallowed — but the combination of a fatal-by-skin-contact statement with an exposure limit that the regulator does not trust air measurement alone to police, which is what a biological monitoring guidance value means. The cyanide variant would decide it by itself if the mercury did not.

Step 7, controls. Extraction where a solution is made up, splash protection rated for a fatal-by-skin substance, and a licensed waste route. The first two are conceivable in a specialist laboratory; the third is a contract rather than a piece of equipment. None of the three is something this course can assume a reader at home has, and the biological monitoring implied by the EH40 entry is not something a reader can arrange at all. No available control moves this down.

Step 8, the historical half. Wall’s bath solved a real problem: it gave a choice of print colours and extra density in one operation, at a time when the alternatives were fewer and the papers less forgiving. What solves that now is sulfide and sepia toning, which converts the image silver to silver sulfide rather than adding a second metal to it, and which this course teaches as a Level B laboratory — bleach and redevelop a sepia print — using the rehalogenating ferricyanide bleach in place of the mercuric one. Selenium toning covers the colder end of the same range. The modern descendant is named, with its lesson.

=The same method, ending at Level A

Run the identical eight steps on the second of the three fixing solutions Towler groups at page 121 of The Silver Sunbeam, the one that uses hyposulphite of soda, and the machine produces a different answer. One substance: sodium thiosulfate, for which 277 of 281 aggregated reports state that the substance does not meet GHS hazard criteria — and the encyclopaedia entry says what that does and does not mean, because an absence of an agreed classification from a thin evidence base is not a finding that nothing can go wrong. Products: with a strong acid it decomposes and precipitates sulfur, so the pairing is worth recording, and it is exactly why this course keeps the acid stop bath and the fixer as separate operations. Energies: none. Waste: silver-bearing, which is a recovery question rather than a toxicity one, and Part XII teaches the route. Failure modes: a spoiled negative, mostly. Verdict Level A, and the deciding item is nothing in the bath — the level is set by the operation, a tray of a dilute solution at room temperature. Modern descendant: it is the modern route, published as the plain hypo fixing bath and explained in how fixer works.

Two formulas, one book, one section, 1864, and the method separates them without any appeal to how old they are. That is the whole argument of this part, done in one page-turn.

The three procedures you are set

Each is a real procedure in a real book in this course’s research corpus, with a stable page reference. Read them; do not reproduce them. Your deliverable describes and classifies; it never transcribes.

#	Procedure	Where	What it is
1	Pizzighelli and von Itterheim’s working modification of Pellet’s positive process	Lietze, Modern Heliographic Processes, 1888, pages 69 to 73	A gum-and-iron sensitiser in which light destroys the image, developed in yellow prussiate so that the blue forms where the light never reached, then cleared in a dilute acid bath
2	Kodak Formula IN-4, the chromium intensifier	Kodak Limited, Formulæ and Notes, 1949, page 34, and the same formula in Eastman Kodak’s 1928 primer at page 59	A stock solution of potassium bichromate and hydrochloric acid, used dilute, to bleach a thin negative so that it can be redeveloped denser
3	Fixing solution Formula No. 1, with cyanide of potassium	Towler, The Silver Sunbeam, 1864, page 121	The first of three fixing solutions Towler prints, the other two being hyposulphite of soda and sulphocyanide of ammonium

They were chosen so that the honest answers land at three different levels of the rubric, so that none of them can be classified by reputation, and so that all three do a job you already understand from earlier parts of this course. Two of them come from books that also print, on the same or a neighbouring page, the procedure that supersedes them — which is the historical half of the assessment handing itself to you.

The optional fourth. Find a procedure nobody has classified for you. Period manuals in the bibliography are full of them, and so are reprints and supplier listings. The marks are for the same eight steps, plus one more: say how confident you are and why, because a procedure with no course page behind it is the case where an assessment has to carry its own uncertainty.

The marking scheme

Published here rather than held back, because a scheme the student cannot see cannot be worked to.

Criterion	Marks	What a full mark looks like
Nomenclature translated	5 per assessment	Every period name resolved to a modern name and a CAS number, and any name that could not be resolved recorded as unresolved rather than guessed at
Substances and products listed	10	Every named substance, every variant’s substances, and the products of any pairing the procedure creates
Classification and exposure data sourced	15	Named source, the statements that matter, the exposure limit where one exists, and the date read. A period manual cited as a hazard source scores zero for this criterion
Energies and waste	10	Both present, both answered even where the answer is none, and the waste answer engages with the law where you actually live
Failure modes	10	At least three, each with the consequence to a person and the control that would have prevented it
Verdict, with the deciding item named	20	The level, the single item that decided it, and the criterion of the rubric it met. A verdict without a named deciding item scores zero here however correct the letter is
Controls and the facility question	10	A named control per hazard, an honest availability answer, and the facility named where the control is not available
The historical half	15	The problem the procedure solved, and the modern descendant named with the lesson in this course that teaches it
Fact separated from interpretation	5	A period claim quoted, identified as evidence of belief, and kept out of the hazard section

☠How the prohibition is enforced

A submission containing a quantity, a working strength, a dilution, a temperature and time, or a followable sequence of steps for any of the four procedures is returned unread and scores nothing. Not marked down: returned.

This is not a technicality. The reason the rest of the scheme can afford to reward describing the chemistry precisely is that the description stops short of the recipe. If the deliverable may carry quantities, then the more thorough the assessment the more complete the working procedure it contains, and the assignment turns into the thing it exists to prevent. The rule applies to procedure 1 as strictly as to procedure 3, even though the course publishes procedure 1’s own formulary entry, because a marker reading a page of assessments cannot tell at a glance which quantities came from which book — and neither can the next person who picks the document up.

Naming a substance is not a quantity. Saying that a bath was used dilute is not a working strength. Saying that a plate was exposed in sunlight is not a time. The line is sharp and it is easy to stay the right side of once you have written one assessment.

Model answers

!Write yours first

What follows is compressed to the verdict, the deciding item and the historical half. Everything between — the substance rows, the products, the failure modes — is the part that carries the marks, and reading a conclusion before you have reached one costs you the exercise. Come back after you have three assessments on paper.

1. Pellet’s positive process, Lietze 1888, pages 69 to 73

Substances. Gum arabic; ammonium iron(III) citrate; iron(III) chloride; an organic acid, citric or tartaric; potassium ferrocyanide as the developer; hydrochloric acid in the discharging bath, made from the concentrated acid as sold at the usual 36 to 37 per cent w/w; sulfuric acid in Pellet’s own lower-fuming variant; potassium alum in the final rinse.

Products and pairings. Prussian blue in the print, which is the image. The pairing to notice is the complex cyanide and the strong acid, which appear in the same procedure two steps apart. Lietze’s sequence washes and brushes the surface film off between them, and Princeton’s conditions for hydrogen cyanide release from a ferrocyanide are heat, hot acid or strong ultraviolet. This assessment’s reading — which is also the reading behind the course’s own Level B ruling on the formulary entry — is that a cold, dilute, separately-washed acid step supplies none of the three, so the pairing is a failure mode held in check by the sequence rather than the deciding item. State that as a reading rather than as a fact, because it is one, and say what would change it: a warm bath, a stronger acid, or the two brought together without the wash between. Pellet himself was concerned enough about the acid fume to recommend replacing part of the hydrochloric with sulfuric to reduce it, and Lietze passes on the instruction that the acid goes into the water and into a bottle rather than a tray.

Energies. Sunlight, an exposure of minutes; no heat, no flame, no mains.

Waste. Iron salts, Prussian blue and dilute acid. Neutralisation and collection is a domestic possibility rather than a specialist one, subject to your local rules.

Verdict: Level B. The deciding item is concentrated hydrochloric acid as purchased — H314, causes severe skin burns and eye damage, in 99.9 per cent of the reports that carry hazard codes, with EH40 setting the gas and its aerosol mists at 1 ppm long-term and 5 ppm over fifteen minutes and NIOSH setting a ceiling of 5 ppm. The Level B criterion it meets is the handling of concentrated acids, and the operation that carries the risk is the dilution rather than the printing. Splash protection, ventilation, eyewash within reach and acid-into-water discipline are all controls a domestic reader can assemble, which is what keeps it at B rather than C.

Historical half. The problem was a positive from a positive: a blueprint of a line drawing that reads dark-on-light rather than light-on-dark. What solves it now, in this course, is the cyanotype family worked in the ordinary negative-working direction with a digital or paper negative, taught as a laboratory in coating and printing a classic cyanotype, with the variants and their trade-offs in the new cyanotype and its variants. Pellet’s own process is not superseded so much as made unnecessary, and the course publishes it as a formulary entry at Level B.

2. Kodak IN-4, the chromium intensifier, 1949, page 34

Substances. Potassium dichromate; concentrated hydrochloric acid; and a non-staining redeveloper, for which the handbook names D-72.

Products. The bleach converts the image silver to silver chloride, and the chromium(VI) is reduced in the process. The used bath and the wash water carry chromium.

Verdict on the rubric: Level C. Potassium dichromate’s notified classification runs to twelve statements, including H340, may cause genetic defects and H350, may cause cancer, each in 99.8 per cent of 491 reports, and H360 in 78.4 per cent. EH40 sets chromium(VI) compounds at 0.01 mg/m³ as chromium with the Carc and Sen annotations and a biological monitoring guidance value; NIOSH treats chromic acid and chromates as a carcinogen at 0.001 mg/m³ as chromium. The rubric’s Level C criteria name “chromium(VI) compounds, formaldehyde and similar substances in any quantity”, so the rubric read literally puts this at C, and the deciding item is potassium dichromate.

Verdict in this course: Level D, as a decision rather than a rubric reading. The course is written for a home darkroom and cannot assume a fume cupboard, a licensed waste route or a supervisor, so it presents the dichromate chemistry at Level D and says that this is its own decision. The full argument is the chromium policy and it is restated in the part overview. An assessment that reaches C on the rubric and then records the course’s departure, with the reason, is the full-mark answer; one that reaches C and stops has read the rubric correctly but missed that a classification is a decision about a reader as much as about a substance; one that reaches D without noticing that the rubric says C has got the right letter for an unexamined reason.

Historical half. The problem was a thin negative in an era with no variable-contrast paper and no second chance at the plate, and Kodak’s 1928 primer records that this intensifier had found increasing favour for the ease and certainty of its operation and the permanency of the result. What solves it now is, in order of preference: a negative that does not need intensifying, which is what the characteristic curve is for; a paper grade or a filtration that absorbs the contrast deficit; and, where a transparency genuinely has to be made denser, Kodak IN-5, the silver intensifier printed on the very next page of the same handbook, which this course publishes at Level B because every one of its reagents has an encyclopaedia entry at Level A or B.

3. Towler 1864, page 121, fixing solution No. 1, with cyanide of potassium

Substances. Potassium cyanide and water. Hardwich, writing in the same year, records that the commercial salt arrived as fused lumps often contaminated with a large percentage of carbonate of potash — which is a reminder that the substance in the bottle is not always the substance in the formula, and that the assessment should say so.

Products. With any acid, hydrogen cyanide gas, immediately and without heating. A darkroom is a room with an acid stop bath and an acid fixer in it.

Energies. None. This is the case that proves why step 3 has to be asked anyway: the procedure needs no energy input at all, and it is the most hazardous of the three by a wide margin.

Waste. Silver-bearing cyanide solution, for which no domestic route exists.

Verdict: Level D. The deciding item is potassium cyanide, and the criterion it meets is the first Level D criterion, which names cyanide fixing explicitly. The classification is signal word Danger with H300, H310 and H330 — fatal if swallowed, fatal in contact with skin, fatal if inhaled — each in 99.8 per cent of 575 reports; EH40 gives hydrogen cyanide 0.9 ppm long-term and 4.5 ppm over fifteen minutes with the Sk notation; NIOSH sets a ceiling of 5 mg/m³ over ten minutes and an immediately-dangerous-to-life value of 25 mg/m³ as cyanide. But note what actually decides it, because this is the subtlety of the case: the incompatibility rather than the toxicity. A substance whose governing hazard is what it does on meeting something already standing in the room cannot be controlled by handling discipline alone.

Historical half. The problem was speed and a whiter plate: the collodion photographer had to finish while the plate was wet, and Hardwich calls cyanide a far more energetic solvent of the insoluble silver salts than hyposulphite of soda. What solves it now is thiosulfate, which does the identical chemistry with a different ligand — how fixer works and fixer formulations and types — and Towler prints it himself as the next of the three fixing solutions in the same section. The full evidence on the cyanide bath, and the correction that the period’s so-called cyanide toning bath was a thiocyanate, is the formulary entry and the cyanide lesson.

The atlas entry

The second deliverable is one process-atlas entry, written to the course’s own format. Choose a process from this part — the daguerreotype, the ambrotype, the tintype, wet-plate collodion, the uranium-toned print, heliography, the carbon print, the Woodburytype, collotype or photogravure — and write your entry before you open the published one.

Six sections, in this order, which is the order every entry in the atlas uses:

Section	What it has to do	The commonest failure
The chemistry	The reactions, with mechanism, and an honest gap where the mechanism is not established	Repeating a period explanation as though it were a modern one
Historical workflow	What was done, in what order, in the third person and the past tense, at the resolution a museum label needs	Slipping into the second person, at which point it has become a procedure
The image	The property a conservator identifies the object by, and the chemistry that produced that property and not another	Adjectives. “Beautiful tonality” identifies nothing
Permanence	How the object fails, over what timescale, on what evidence	Asserting archival permanence, which is a claim requiring evidence
Hazards	The substances, from named sources with dates, and what they do to a person handling the object now	Confusing the hazard of making the object with the hazard of holding it
Why we study it and do not reproduce it	The deciding item, the control that is missing, and the modern route this course teaches instead	Ending on a warning instead of on an alternative

The atlas also carries structured fields in its front matter — family, inventor, year, status, safetyLevel, image colour, surface, tonal scale, light source, sensitivity, permanence, cost and difficulty — and filling those is where your hazard assessment and your atlas entry meet: the safetyLevel field is the verdict from step 6, and status is either practised or historical-study depending on it.

Then open the published entry and write a short comparison: where you agreed, where you differed, and which of you has the better evidence. Differing is not automatically an error. The atlas entries carry their sources, so a disagreement can be traced to a document rather than argued about, and finding one the course has got wrong is worth more than agreeing with it.

The identification exercise

Three objects, identified from what you can see, with the reasoning written out.

Choosing the objects. Use a collection whose reuse terms you have actually read. This course has read two: the Met’s open-access policy and the Rijksmuseum’s information and data policy, which marks material free of copyright, or over which the museum has waived its rights, as Public Domain or under the Creative Commons Zero dedication, licenses some material CC BY 4.0 with attribution, and keeps some items restricted with a copyright holder named separately. Other collections have their own terms and they are not interchangeable; the Library of Congress publishes a guide to assessing exactly this question for its own prints and photographs. Read the terms of whichever you use.

Search by date and holding institution, not by process. If you search for “daguerreotype” you have been told the answer before you have looked at anything. Search a date range instead — roughly 1840 to 1860 for the cased objects, 1855 to 1900 for the paper prints — and cover the medium field of the catalogue record before you look at the image. Reveal it only when your reasoning is written.

Record the metadata block for each object, in this order, which is the order this course’s own image-credit policy uses: creator; title; date; institution; accession or object number; the rights position; and the source URL. For the rights position, prefer the standardised vocabulary of RightsStatements.org — twelve statements in three groups, In Copyright, No Copyright and the undetermined ones — or the collection’s own words quoted exactly. Recording a rights position you have read is part of the exercise, and it is the same discipline the brief requires of this course: no image is ever used here without a creator, a licence and a source.

Then identify, from four kinds of evidence.

Evidence	What to look for	What it distinguishes
Support	Glass, iron, paper, or a metal plate in a sealed case	A tintype is on iron and has no glass to look through, which is the shortest confirming test in the atlas; an ambrotype is on glass
Surface	Specular mirror, glossy coated layer, matte fibre, or relief you could feel in raking light	A daguerreotype is the polished plate itself; an albumen print has a coated layer; a salted paper print is in the fibres; a carbon print stands in relief
Colour	Neutral, reddish or light brown, purplish brown, blue, or a pigment colour that no silver process makes	Untoned salted paper runs light brown to reddish brown; gold-toned albumen is a rich purplish brown; a carbon print is whatever colour the tissue was made in
Polarity behaviour	Whether the picture inverts as the object is tilted, and what happens if a backing is removed	The daguerreotype is ambivalent — highlights scatter, shadows reflect — and reads positive or negative with the viewing angle; the ambrotype does not, but flips if its black backing comes off

Write the reasoning, not the answer. What you saw; what each observation ruled out; what remained; and how confident you are. Then reveal the catalogue attribution and record where you were wrong, which is the part of the exercise that teaches.

!The honest limit of an identification from a screen

A photograph of an object is not the object. Specular behaviour, surface relief, the exact tarnish of a plate and the weight of a support are the evidence a conservator uses and are the evidence a digital image loses first. Where the hard cases are settled in a laboratory they are settled instrumentally: the Getty’s atlases identify processes by X-ray fluorescence and infrared signatures, not by eye.

So an identification made this way is a hypothesis with a confidence attached, never a determination — and saying so in your write-up scores marks rather than losing them. This is the same discipline as the hazard assessment’s last field: a document that records what it could not establish is worth more than one that quietly does not mention it.

What to submit

• Three hazard assessments, on the record above, one per set procedure.
• One process-atlas entry in the six-section format, plus a short comparison against the published entry.
• Three identifications, each with its full metadata block, its reasoning, its confidence and its correction.
• Optionally, a fourth assessment on a procedure of your own finding.
• A one-page covering note listing everything you could not establish, and where you looked.

What you keep afterwards is worth more than the marks. Three completed hazard assessments and one atlas entry are the same documents a working laboratory produces before it starts anything, and the habit of writing them before rather than after is the difference between a discipline and a reputation for being careful.

Where this goes next

The method does not stay in this part. In Part XXVII you will apply the same eight steps to a developer you have designed yourself, where the awkward difference is that no period manual and no museum has assessed it for you and the substances are ones you are actually going to handle. In the capstone the same assessment is run over a whole workflow, which is where it stops being an exercise and becomes the document you write before you start.

And when you meet an unfamiliar formula in a reprint years from now — which you will — the sequence is the same one, in the same order, and the only thing that will have changed is that the classifications you look up will have been revised. Which is why every assessment carries the date it was read.

Check your understanding

Question 1. An assessment concludes: "Level D — the bath contains several substances of concern, and taken together the procedure is unacceptable for home work." What is wrong with it?

• Nothing; that is the correct form for a Level D verdict
• It should have averaged the levels of the individual substances rather than taking the worst
• It does not name the single item that decided the level, so it cannot be checked, argued with, or revised when better data appears
• Level D verdicts must always cite an exposure limit, and this one cites none

Show the answer and why

Answer: It does not name the single item that decided the level, so it cannot be checked, argued with, or revised when better data appears

The rule is that the worst single item decides the level, and the discipline is that the item is named. "Taken together" is the shape of an averaged judgement, which the rubric does not use. Naming the deciding item is what makes the assessment auditable and what tells you exactly what to reassess if that item is removed or the classification behind it changes.

Question 2. Wall's 1924 handbook makes the period claim that skin absorption of mercuric chloride, even through cuts and abrasions, is "practically harmless". Where does that sentence belong in an assessment?

• In the hazards section, with a note that older sources disagree with newer ones
• Nowhere; a claim contradicted by current data should be left out of the document entirely
• In the historical section, as evidence of what was believed in 1924 and therefore of how the procedure came to be worked as it was — and never in the hazards section, where the notified classification carrying H310 governs
• In the hazards section, because a working photographer's observation is Tier 2 evidence

Show the answer and why

Answer: In the historical section, as evidence of what was believed in 1924 and therefore of how the procedure came to be worked as it was — and never in the hazards section, where the notified classification carrying H310 governs

The brief's seventh rule separates established chemistry from historical evidence, and this part is where that separation earns its place. A period assertion is excellent evidence of belief and worthless as evidence of toxicity. Removing it altogether would be a different error: it explains why operators handled the salt the way they did, which is part of the history the assessment is supposed to tell.

Question 3. Kodak IN-4 is a potassium dichromate and hydrochloric acid bleach. What is the fully correct verdict in this course?

• Level C, because the rubric names chromium(VI) compounds in any quantity as a Level C criterion, and the rubric is the authority
• Level D, because chromium(VI) is a carcinogen and carcinogens are always Level D
• Level C on the rubric, recorded together with the course's own decision to present it at Level D because a course written for a home darkroom cannot assume a fume cupboard, a licensed waste route or a supervisor
• Level B, because the working dilution is one part of stock in ten of water

Show the answer and why

Answer: Level C on the rubric, recorded together with the course's own decision to present it at Level D because a course written for a home darkroom cannot assume a fume cupboard, a licensed waste route or a supervisor

Both halves are needed. The rubric read literally puts chromium(VI) at C, and an assessment that reaches C has read it correctly. The course then departs to D and says openly that this is its own decision about its own readership rather than a rule of the rubric — which is exactly the kind of distinction the seventh rule of the brief asks for, and exactly what a reader needs in order to disagree with the decision if their circumstances differ.

Question 4. Pellet's process brings a potassium ferrocyanide developer and a dilute hydrochloric acid bath into the same sequence. Why is that pairing not the deciding item in its assessment?

• Because complex cyanides cannot release hydrogen cyanide under any conditions
• Because the conditions on record for hydrogen cyanide release from a hexacyanoferrate are heat, hot acid or strong ultraviolet, none of which a cold dilute bath reached after an intervening wash supplies — so the pairing is a failure mode held in check by the sequence, while the deciding item is the concentrated acid being diluted
• Because the ferrocyanide has already been consumed by the time the acid bath is reached
• Because Prussian blue is inert once formed, which neutralises the acid

Show the answer and why

Answer: Because the conditions on record for hydrogen cyanide release from a hexacyanoferrate are heat, hot acid or strong ultraviolet, none of which a cold dilute bath reached after an intervening wash supplies — so the pairing is a failure mode held in check by the sequence, while the deciding item is the concentrated acid being diluted

The distinction between a hazard and a controlled failure mode is what an assessment is for. Princeton's guidance names the three conditions under which a ferricyanide or ferrocyanide will release hydrogen cyanide, and records real poisonings from acidifying Farmer's reducer, so the pairing is genuinely worth writing down. What decides the level, though, is the item with the worst classification handled in the worst operation: concentrated hydrochloric acid, H314, at the moment it is diluted. Lietze passes on Pellet's own instruction to pour the acid into the water and into a bottle, which is the period recognising the same thing.

Check my answersClear