Egg Albumen: A Protein Binder and What It Does to Silver

Every instruction in albumen printing that looks like superstition is an instruction about protein. Beat the whites to a stiff froth and let them fall back. Put the salt in before you beat, not after. Add two millilitres of acid. Leave it a week in the cold until it smells. Never let it get hot. Never let it dry too far. Do not put a second coat on until the first one has been made insoluble. Each of those is a controlled attack on a folded molecule, and by the end of this page you should be able to say which molecule, what happens to it, and what the photograph looks like as a result.

The reward for that work is what Reilly states as a general law of printing papers in his first chapter: an image that has penetrated into the paper fibres has a lower maximum density and a matte surface, because the light reflected from the paper is scattered and diffused by the fibres; an image produced “in a compact layer resting on top of the paper fibers” scatters much less and reaches a much greater maximum density. Albumen is how the nineteenth century got the image out of the paper and onto it, forty-five years before baryta did the same job with barium sulfate.

What egg white actually is

Not one substance. The photographic word albumen means the clear white of a hen’s egg taken as a whole, and Reilly says so explicitly — “many specific proteins can be identified in egg white, but when used collectively they are referred to as albumen.” That collective noun hides the fact that the components behave differently from each other, and the whole preparation exists to stop them doing so.

The numbers must come from a food chemistry source rather than from a photographic one, because the photographic literature was written before anybody could separate the proteins. Li and colleagues’ 2022 review gives the composition as about 88 per cent water, 11 per cent protein, 0.2 per cent fat and 0.8 per cent ash, and the protein fraction as follows, each figure a proportion of the total egg white protein:

Protein	Share of total protein	Molecular weight	Isoelectric point
Ovalbumin	54 %	45 kDa	4.5
Ovotransferrin (UniProt’s alternative name for it is conalbumin)	12–13 %	77 kDa	6.0
Ovomucoid	11 %	28 kDa	4.1
Lysozyme	3.4–3.5 %	14.3 kDa	10.7
Ovomucin	1.5–3.5 %	very large and variable	4.5–5.0
Ovomacroglobulin	0.5 %	760–900 kDa	4.5–4.7
Avidin	0.05 %	68.3 kDa	10.0

Four things in that table matter to a printer.

One protein dominates. Ovalbumin is more than half of the protein, and Li’s review names it as “a major contributor to the foaming, gelling and emulsifying properties” of egg white. When you beat a white to a froth, ovalbumin is most of what you are beating; when the coating on your paper turns insoluble, ovalbumin is most of what has turned.

The isoelectric points are scattered from 4.1 to 10.7. That is the chemical statement of Reilly’s practical complaint that the proteins “all have different viscosities”. At the pH of a fresh white, some of these molecules carry a large net negative charge and some are near their isoelectric point and barely charged at all, so they behave differently in solution and settle out of a coating differently. Lowering the pH with acid, as the albumen solution does, moves all of them in the same direction at once.

One of them binds metals for a living. Ovotransferrin, the second largest component, has two ligand centres that in Li’s words “can bind to metal ions such as Fe, Cu or Zn”. Nothing in that review says anything about silver, and this page will not pretend that it does — but a printer who is about to pour a metal salt onto a protein layer should know that metal binding by egg white proteins is ordinary rather than exotic.

And two of them are the reason albumen prints yellow. Egg white protein carries sulfur, in cysteine and methionine side chains. The reference record for ovalbumin, UniProt’s reviewed entry P01012, gives a 386-residue chain of 42,881 Da carrying six cysteines and seventeen methionines, with exactly one disulfide bond annotated — the bridge between Cys-74 and Cys-121 — which leaves four free cysteine thiols per molecule, unpaired and available. Mike Ware states the comparison that makes this matter in a footnote: “egg albumen contains a much higher concentration of the thio-aminoacid, cysteine, than does gelatin.” That single difference between the two protein binders of nineteenth-century photography is most of the deterioration lesson.

A protein, in the terms this page needs

A protein is a chain of amino acids folded into a specific shape, and the shape puts particular side chains on the outside where they can meet other molecules. That is the whole of the picture this course needs, and three consequences follow from it.

The surface is chemically mixed. Some exposed side chains carry a negative charge at working pH (the carboxylates of aspartate and glutamate), some carry a positive one (the ammonium of lysine, the guanidinium of arginine), some are polar but uncharged, and some are greasy hydrocarbon or aromatic groups that would rather not be in water at all. The net charge of the whole molecule is zero at its isoelectric point and grows as you move the pH away from it — which is why the table above lists a pI for every protein and why moving the pH is one of the levers in the preparation.

The fold is held together by weak forces, and they can be broken. Hydrogen bonds, salt bridges between oppositely charged side chains, and the tendency of the greasy groups to hide from water in the interior are what keep the chain folded. Reilly gives the operative statement for photography in one line: “the chemical forces which bind together the enormous molecules of protein grow weaker as the pH is lowered, and the physical properties of the substance change as a result.” Add acid and the fold loosens.

When the fold opens, the greasy interior comes out — and finds other molecules. An unfolded chain exposes patches that were buried, and those patches associate with the same patches on the neighbouring chain. The molecules link into a network. That is what makes a beaten white stand up in peaks, what makes a cooked white opaque and firm, and what makes a dried albumen coating survive a water bath instead of dissolving off the sheet.

One protein chain, three states

1. 1Native — folded; greasy groups buried, thiols mostly inaccessible; the molecule is soluble and the solution is thin
2. 2Denatured — unfolded; buried groups now exposed. The chain is the same chain — no bond in the backbone has been broken
3. 3Coagulated — unfolded chains associated into a network. Insoluble, and Reilly says the change is complete and irreversible
4. 4The four agents this process uses — beating, heat above 65 °C, alcohol, and a metal salt — silver nitrate is the metal salt, and it arrives at the sensitising step

A schematic of the state change, not a structure. No source read for this course publishes a molecular picture of the coagulated albumen layer of a photographic print; what the sources establish is the four agents, the irreversibility, and the insolubility that follows.

Denaturation, coagulation, and why the coating survives a water bath

Reilly’s chapter two gives the four agents that turn albumen insoluble, and they are worth listing together because every one of them appears somewhere in the two labs of this part.

Heat. “Albumen is also coagulated by temperatures above 65 °C.” That is the number to keep, and it explains two apparently contradictory instructions. Every wet bath in the printing session is held near 21 °C and nothing is ever warmed, because heat coagulates a wet layer unevenly and blisters it. But Bostick & Sullivan harden a dry coating deliberately with a dry-mount press set to 250 °F — about 121 °C — or a clothes iron on its highest setting, for two minutes between coats. Heat above the coagulation point is the tool when the layer is dry and the enemy when it is wet.

Alcohol. “Albumen is insoluble in alcohol, and in fact alcohol will coagulate albumen, a property that is useful to obtain multiple coatings of albumen on a single sheet.” This is the whole basis of the double-coating procedure, and the reason the bath is 70 per cent rather than neat is discussed below.

Salts of metals. Reilly lists coagulation “by contact with salts of metals” among albumen’s ordinary properties, and then draws the photographic conclusion that most beginners get wrong:

The reason why albumen does not dissolve off the sheet during processing is because contact with silver nitrate in the sensitizing bath coagulates it and forms a new insoluble silver-albumen complex called silver albumenate.

Mechanical work. Beating denatures, which is why the whites are frothed. This one is different from the other three in that it does not by itself make an insoluble film — the froth falls back to a liquid, and that liquid is what you coat with.

⚗What is and is not established about coagulation here

Established, from Reilly: albumen coagulates above 65 °C, is coagulated by alcohol, is coagulated by salts of metals, and is coagulated by silver nitrate specifically, which is why a single-coated sheet does not wash off during processing.

Established, from Li and colleagues: avidin, a trace protein, is irreversibly denatured at 70 °C, and lysozyme is not denatured by 1 to 2 minutes at 100 °C in acid solution. The review states no denaturation temperature for ovalbumin or for ovotransferrin, does not describe the structural mechanism of heat denaturation, and does not discuss what mechanical whipping does at an air-water interface.

Not established anywhere the course has read: a molecular account of what silver albumenate actually is, a bond dissociation figure, a maximum safe water temperature for a wet albumen layer, or a rate for any of these processes. The nineteenth century worked these out by operation rather than by measurement, and the twentieth century moved on to gelatin before anybody measured them. Where this page gives a chemical picture beyond the four agents, it is marked as an inference.

There is one more mechanism, and it belongs to the exposure rather than to the coating. Ware’s account of printing out states that a silver chloride paper only continues to darken if something absorbs the chlorine that light liberates, or it simply re-oxidises the silver you have just made. In plain salted paper, he writes, the chlorine absorber is “a combination of water, which is present in the paper fibres, and excess silver nitrate”. In the albumen print, “the albumen itself acts as a scavenger for the chlorine.” The binder is not a transparent bystander over the reaction; it is one of the reagents.

2 AgCl + light → 2 Ag + Cl₂

What light does — and what makes it stick is what happens to the chlorine next

Reilly attributes the standing explanation to Hermann Vogel: as light dissociates each unit of silver chloride, the liberated chlorine unites with the excess silver nitrate to make new silver chloride, which light breaks down in turn, “and the cycle begins again, to be repeated over and over. When excess silver nitrate is available, more image silver will be formed and a greater maximum density attained.” Ware adds the albumen to the same job. Both belong to Part XXII’s chemistry lesson, which owns the mechanism; what is new here is that the binder is inside it.

Why it is beaten, and why the froth is not what you use

The beating is the first and strangest step, and Reilly is unambiguous about its purpose:

Albumen is never used in the strictly native condition; before any photographic use may be contemplated, the egg whites must be beaten to a froth and allowed to settle back to a liquid state. This beating process denatures the various proteins — all of which have different viscosities — and results in a homogeneous liquid which will form an even layer on the sheet of paper.

Homogeneity is the goal, not aeration. A native white is stringy and unevenly viscous, and a stringy liquid cannot be floated into an even film. Beating breaks the difference between the components down until the liquid behaves as one substance.

What keeps the thick material out is the separation and the muslin, not the froth. The chalazae — the twisted cords that hold the yolk centred — are excluded at the moment the egg is cracked. Reilly’s instruction is that only the clear white is saved, “without the slightest contamination of yolk, blood or the stringy tissue known as the chalazae”, and he recommends separating over a small bowl and pouring each white into a larger one, so that one bad egg never has to be picked out of a litre of good ones. The filtration is the muslin, and it happens twice: the mixture is strained through it after the settling, squeezed through under some pressure if it will not run, and filtered through it again immediately before the albumen goes into the coating tray.

It would be tidy if the froth did some of that work itself, catching the thick material in the foam so that only clear liquid drained back. No source read for this course says that it does, and the course does not assert it. Reilly gives the beating exactly one purpose, which is homogeneity; he gives the exclusion of the chalazae to the separation and the filtration to the muslin. The inference is available to you and it is not evidence, which is worth noticing here because it is the shape of most of what gets repeated about this process.

The salt goes in before the beating, and that is not arbitrary. Reilly gives the reason and it is a chemical one rather than a matter of convenience: “the chlorides added are vital to the formation of the photographic image, but they also help to denature the albumen; one of the effects of adding chlorides is to reduce the volume of froth produced during the beating step.” So the chloride is doing three jobs — making silver chloride later, denaturing the protein now, and controlling how much froth you have to wait for.

The Getty’s process description matches, from an independent direction: a solution of sodium or ammonium chloride is added to the egg white, the mixture is beaten to a stiff froth, and “left to stand overnight, the froth would liquefy, resulting in a much more homogenized and uniform solution of salted albumen that was usually filtered.”

What each step of the preparation is for

1. 1Separate, keeping every trace of yolk, blood and chalaza outYolk is fat and lipoprotein. It does not belong in a film that has to dry transparent and lie flat, and it cannot be removed later
2. 2Dissolve the chloride, and the acid, in a little water firstReilly: the chloride "should be dissolved in a minimum of water and added to the egg white before the beating process". An undissolved crystal sitting in egg white coagulates the protein around it locally
3. 3Beat to a complete frothMechanical denaturation. Evens out the viscosity of components that behave differently, and traps the chalazae and membranes in the foam
4. 4Stand, covered, 24 hoursThe froth liquefies and drains. What collects beneath it is the working albumen; what stays in the foam is what you did not want
5. 5Strain through muslinThe third filter. Reilly notes the liquid may have to be squeezed through with some pressure
6. 6Age one week, refrigeratedChemical denaturation continuing slowly. The end point is stated as an appearance and a smell, not a time: a yellowish homogeneous liquid with a slight aged odour

The sequence and every quantity are on the albumen solution's formulary page; this diagram carries only what each step is for.

Ageing and fermentation: the documented and the folkloric

This is where a reader has to be careful, because the practice is real, the effect is real, and the mechanism is almost entirely unpublished.

What the sources establish. Reilly’s account is that during the 1850s “it was found by many experimenters working independently that partially decomposed — in chemical terms ‘denatured’ — albumen yielded a glossier and more even coating”, and that decomposed albumen “passes into an acid condition and forms a homogeneous mixture without the uneven viscosity and stringiness of native egg white.” Some albumenisers went further and let it ferment at elevated temperatures for several days using naturally occurring bacteria. That became standard in the Dresden factories from the early 1870s, and Reilly’s aside is the memorable one: “Dresden paper could readily be identified by the smell.” The Getty confirms the effect independently, listing albumen paper “prepared using aged or partially putrefied albumen” among the things that “produced higher-gloss albumen prints.”

Reilly’s own working end point for a modern batch is a description rather than a time: after settling for twenty-four hours and a week in a refrigerator, “it is a yellowish homogeneous liquid with a slight ‘aged’ odour that signals its readiness for use”, and it stays useful for several weeks after that. Deciding when a batch has gone too far is likewise sensory: “the smell of the albumen and its color, sedimentation, etc., will reveal when it has decomposed too badly to use.”

What the sources do not establish. Not one of the documents read for this part gives a chemical account of the fermentation: no organism is named, no product is identified, no pH is measured before and after, no viscosity is reported, and no experiment separates the effect of the falling pH from the effect of whatever the bacteria are making. Reilly himself writes that “the origins of the fermentation technique are uncertain.” What the literature contains is a practice, a stated visual and olfactory end point, and an agreed effect on gloss and evenness.

⁇Fermentation: a documented practice with an undocumented mechanism

The disagreement is not about whether it works. Reilly, the Getty and the nineteenth-century trade all say that aged or fermented albumen coats more evenly and prints glossier, and Reilly adds that fermented albumen “provided glossier paper that toned more easily”. Nobody read for this course contradicts that.

The disagreement is about what to call the explanation. One reading, which Reilly supports directly, is that the whole effect is denaturation by acidification: decomposition drives the pH down, low pH weakens the forces holding the fold together, and the liquid becomes homogeneous — in which case bacteria are simply a slow way of adding acid, and the two millilitres of glacial acetic acid in his own sample preparation are a fast one. The other reading, which the period practice implies and no source read here states, is that something the organisms produce does more than lower the pH, since the trade went on fermenting deliberately at elevated temperatures long after acetic acid was cheap and available.

What the course uses, and why. It teaches the acid explanation as far as Reilly’s stated mechanism goes, and marks everything beyond it as unestablished. It does not teach a deliberate warm fermentation, for two reasons: no source read gives a temperature, a duration or an end point that could be followed, and the microbiology of deliberately incubating a protein liquid at elevated temperature is outside anything the course has read an assessment of. Reilly’s refrigerated week is what the labs use, because it is the only ageing regime in the corpus that comes with numbers.

The salting is inside the binder, and that is the whole difference

In Part XXII the chloride is dissolved in water, the paper is floated on it, and the sheet is dried; the salt ends up in and among the fibres, and so does the silver chloride that forms there when the sheet meets silver nitrate. In albumen printing the chloride is dissolved in the egg white before the paper is ever touched. When the sensitiser arrives, the silver chloride is precipitated inside the protein layer, because that is where the chloride is.

NH₄Cl + AgNO₃ → AgCl + NH₄NO₃

The same double replacement as a salt print — but happening in a protein film rather than in cellulose

Where the halide is, in the two processes, at the same magnification

1. 1Salted paper — chloride applied to the sheet in water; silver chloride forms among and inside the fibres, over a depth nobody read for this course has published
2. 2Albumen paper — chloride dissolved in the binder; silver chloride forms inside the protein film, in a plane rather than in a volume
3. 3Diffuse reflection — why a salt print looks matte and why its maximum density is limited: reflected light is scattered by the fibres
4. 4Specular reflection — why an albumen print is glossy and reaches a higher maximum density: a smooth transparent top surface, and less scattering in the whites as well

Drawn to show the arrangement, not to scale, and the depth of the salted-paper image zone is drawn rather than measured because no source read for this course publishes one. The albumen layer's own thickness is not published separately either; the Getty measured a whole unmounted albumen photograph, support included, at 0.204 mm.

How much chloride, and what changes if you move it. Reilly’s figures are the ones the formulary carries, and they describe a shallow optimum. Papers with a low chloride content of 1 to 1.5 per cent “are less sensitive and tend to produce slightly more contrasty prints from thin negatives” than papers at the normal 1.5 to 2.5 per cent — “this slight gain in contrast is at the expense of a rich, dense image, however, so it is best in ordinary circumstances to keep the chloride at 1.5 % or above.” Going the other way is simply waste: “the use of more chloride in the formula than is necessary only results in higher silver consumption without conferring any additional benefit.”

Which chloride is a smaller decision than it looks. Ammonium chloride was the commonest in the nineteenth century; sodium chloride works, and so does a mixture, and Reilly says “prints with similar color and contrast may be expected with the use of either of these chlorides, or both used in the correct amount.” The cation is a spectator in the double replacement above, and it leaves as a soluble nitrate.

🧮Calculation: how much chloride is on a square decimetre, and how much silver arrives to meet it

Every number below is traced to a source or is marked as the course’s own inference. The conclusion is an order of magnitude, and the assumptions are the point of the exercise.

Chloride laid down by the coating. Reilly’s albumen carries 15 g of ammonium chloride in a litre of egg white, which the formulary page reads as 1.5 per cent w/v and explains why. Ammonium chloride is NH₄Cl, molar mass 53.49 g/mol, of which chlorine is 35.45 g/mol:

15 g/L × (35.45 ÷ 53.49) = 9.94 g Cl⁻ per litre = 0.280 mol/L

Chloride in the coating solution

How much of that solution reaches a square decimetre. Reilly gives one figure that lets this be computed, and it is an industrial one: coating a ream of albumen paper — 480 sheets, each 46 × 58 cm — took 9 litres of albumen solution.

480 × 0.46 m × 0.58 m = 128.1 m²

Area of a ream

9 L ÷ 128.1 m² = 70.3 mL/m² = 0.703 mL per square decimetre

Coating volume per unit area

0.703 mL × 0.280 mol/L = 0.197 mmol, or about 7.0 mg of Cl⁻

Chloride per square decimetre

How much silver arrives. Reilly’s replenishment example loses 150 mL from a litre of bath during one printing session, and elsewhere he says a 12 per cent bath stays serviceable for “perhaps two or three batches — say a total of 30 11 x 14 prints”. Those two figures are in different paragraphs and combining them is the course’s inference, not Reilly’s statement. An 11 × 14 in sheet is 9.94 dm², so thirty of them are 298 dm²:

150 mL ÷ 298 dm² = 0.50 mL per square decimetre

Sensitiser taken up per unit area

0.50 mL × 120 g/L ÷ 169.87 g/mol = 0.355 mmol, or about 60 mg of AgNO₃

Silver nitrate per square decimetre

The comparison. About 0.36 mmol of silver arrives to meet about 0.20 mmol of chloride — a molar excess of roughly 1.8 to 1. If every chloride ion is converted, a little over 40 per cent of the silver you applied is still sitting in the layer as free silver nitrate when the sheet goes into the printing frame.

What that explains. It is the excess Talbot found he needed and Vogel explained; it is why the first wash runs milky with precipitated silver chloride; it is why every drop of that first wash is silver-bearing waste; and it is consistent with Reilly’s flat statement that “only about 6-8 % of the silver initially present in a sheet of sensitized paper is utilized to form the image.”

What it does not establish. The per-area figures are chained inferences from two industrial and two illustrative numbers, and a modern printer coating an 8 × 10 sheet by hand in a tray will not match them. Bostick & Sullivan’s own kit implies a coating consumption about four to five times Reilly’s — 1000 mL of albumen for “approximately 60 prints on 8x10 paper” is 3.2 mL per square decimetre — and the difference is presumably their double coat plus what stays behind in the tray, which is the course’s guess and not their statement. Measure your own. Weigh the albumen before and after a coating session and divide by the area you coated; that single figure turns this calculation from an argument into a fact about your bench.

Why the image is denser, and why it is warmer

Two separate effects, often run together and worth keeping apart.

Density is geometry. Reilly’s account is entirely about where the silver is and what happens to light. Confine the image to a compact layer on the surface and the light reflected from the paper is not scattered by fibres on the way out, so the shadows can get much darker than they can in a matte sheet. Add that the layer itself is “some smooth and transparent substance like gelatin or albumen” and “the scattering of light in the white areas of the print will be minimized as well, and the paper will look more ‘brilliant’ and have more contrast.”

Colour is particle size and refractive index. This is Reilly’s chapter one and it belongs to printing-out generally rather than to albumen specifically. Light acting on silver chloride makes tiny particles of metallic silver that aggregate up to a limited size — “a size that chemists call colloidal”, and the AIC gives the range as 5 to 25 micrometres in diameter. Particles that small do not absorb all wavelengths and appear black; they absorb some and not others, “depending in part on the index of refraction of the material in which the particles are dispersed.” Two consequences follow that every printer meets:

• Different binders give different colours from the same chemistry. Albumen, gelatin and starch give prints of different colours because the particles are dispersed in media of different refractive index.
• The colour changes twice during processing, and both changes are physical. Fixing dissolves the unreduced silver chloride, which lowers the refractive index of the system and lets the silver particles pack closer together, so a print that was rich purple out of the frame turns yellower and loses density in the hypo. Drying changes it again — “becoming darker and colder in tone because the index of refraction of the system and the distances between particles have changed.”

📷What you will actually see, in order

Out of the printing frame: rich purple to dark brown, and, as Reilly notes, many writers have wished the print could be kept in that state. In the first wash: it reddens, because the layer swells and swelling alone shifts the colour. In the gold toner: it moves towards lilac or purple, and you judge it by transmitted light. In the fixer: it goes yellower and weaker, dramatically, and this is not a fault. On drying: it comes back darker and colder. Reilly’s warning about judging the toning by eye follows directly — the prints “alter dramatically in fixing and drying, and end up a little darker and colder in tone than they appear in the toning bath.”

And the third effect, which is contrast rather than density. Reilly reports Hübl’s measurements and they are the sourced basis for the claim that albumen has a shorter scale than salted paper: the scale length of glossy albumen paper was equal to that of platinum paper, while salted paper exceeded platinum “by a considerable margin”. The shape differs too. Compared with salted paper as a reference, albumen showed “a slow progression of tones from the shadows to the middletones and a relatively abrupt jump from the middletones to white”, where platinum does the opposite. Hübl’s conclusion is a working instruction: albumen and platinum want negatives of the same density range, but a negative meant for albumen “should emphasize highlight detail at the expense of shadow detail, in order to compensate for relatively high contrast in the highlight end of the scale on albumen paper.”

The silver–albumen complex

This is the mechanism most often asserted without evidence in writing about albumen prints, so the evidence is worth setting out in the order it was found.

1859 — Davanne and Girard. Reilly records their communication to the French Photographic Society in December 1859: a 2 per cent solution of potassium cyanide removed all traces of silver from albumen prints, while strong solutions of hypo did not. Their own conclusion, in Reilly’s quotation, is that the results “show that it is difficult to remove every trace of silver salt contained in albumenised proofs, and consequently, explain the difficulty which photographers often meet with in their attempts to obtain proofs on albumenised paper in which the whites shall be pure and well preserved.” They also noted that cyanide fixation bleached the image and was highly poisonous, which ruled it out.

1866 — Matthew Carey Lea confirmed the presence of residual silver and searched without success for a solvent for it.

1868 — John Spiller, reading to the Photographic Society of Great Britain on 14 January, named what he thought it was: the metal “was retained in the whites of the albumen print, and indeed in all parts of the coating, in the form of an argentic organic compound, colorless, unalterable by light, and comparatively insoluble in hyposulphites and other fixing agents.” His test for it was to moisten the white surface with ammonium sulfide and watch a brown stain appear.

The mid-1890s — Haddon and Grundy measured it. They took prints that had been sensitised and fixed but never exposed, which should have contained no silver at all, and found that a thoroughly fixed and washed sheet “still contained nearly 5 % of the silver left after sensitization and before processing.” Then they proved the point theatrically: they converted the residual silver to silver chloride in chlorine water, applied potassium nitrite as a chlorine acceptor, and printed out an image on it “nearly as intense as one printed in the usual way”.

Where the chemistry connects. Reilly’s own statement of the probable origin is the one the conservation literature has inherited: “the chemical bonding of silver to sulfur-containing side groups on the protein molecules of albumen, some of which have a very high affinity for silver. The silver bonded during sensitization to these sites on the protein is so tightly held that treatment in hypo is not sufficient to remove it.” The AIC’s wiki lists that bonding among its proposed causes of highlight yellowing. Ware supplies the comparison that makes it specific to albumen: egg albumen carries much more cysteine than gelatin does. And the UniProt record supplies the count: four free thiols per ovalbumin molecule, on a protein that is 54 per cent of what is in the layer.

⚗Silver albumenate: what is claimed, and by whom

Reilly’s claim, stated as fact in his chapter two: contact with silver nitrate in the sensitising bath coagulates the albumen “and forms a new insoluble silver-albumen complex called silver albumenate. This substance itself is light sensitive and makes an important contribution to image formation in albumen paper.” His chapter one puts albumen among the “active” organic substances that “facilitate the more complete reduction of silver chloride and also themselves form light-sensitive substances when in contact with silver nitrate (silver albumenate, silver citrate, etc.)”.

The independent evidence for a retained, hypo-resistant silver compound is strong and is above: three separate nineteenth-century investigators found it, one of them measured it at nearly 5 per cent of the applied silver, and one printed a picture out of it.

The independent evidence that it is a thiolate is weaker and is an inference rather than a measurement in every source read here. Reilly, the AIC and Ware all name sulfur-containing side groups as the binding site. None of the documents read for this part reports a spectroscopic identification of a silver–sulfur bond in an albumen print, a stoichiometry, or a stability constant. The course therefore states the binding as established, the sulfur site as the standard and well-motivated explanation, and the molecular identity of “silver albumenate” as not settled by anything it has read.

Why a printer should care today. Because it is the reason the fixing and washing of an albumen print cannot be shortened, the reason no wash removes the last of the silver, and the reason a print that was properly processed can still yellow. It is also the reason the course teaches gold toning here as protection rather than as colour.

Gloss, and where it comes from

Gloss is a property of a surface, not of an image. A dried albumen film is smooth and transparent, so a large part of the light that hits it leaves in one direction — a specular reflection — instead of being scattered by fibres. That is all gloss is, and everything the trade did to increase it was aimed at making the top surface smoother, flatter or thicker.

Dilution sets it. Reilly puts the whole spectrum in one sentence: albumen “may be applied in pure form to produce a glossy paper or may be diluted to any strength, with a corresponding loss of gloss, detail, and ‘brilliance’ of image.” The Getty says the same from the object side: the major difference between published recipes “lay in whether the albumen salt solution was used as prepared, or if it was diluted with various amounts of water”, and diluted coatings gave less glossy prints. The course’s formulary carries both ends of that axis as two entries — the undiluted albumen solution and Hardwich’s three-parts-albumen-to-one-of-water albumenised salting solution — precisely because they are two points on one line rather than variants of each other.

Drying temperature sets it. “The higher the temperature of the drying environment, the faster will be the drying, and more importantly, the glossier will be the paper.” Reilly records drying rooms held at 30 to 50 °C in nineteenth-century factories and calls high-temperature drying “a way to improve the gloss and depth of single-coated albumen papers” rather than a necessity.

Ageing sets it, as above.

Coating twice sets it, which is the next section.

And after the print is finished, mechanical work sets it. Rolling and burnishing were routine. Reilly explains that heavy layers of albumen can dry with “a horny, rough surface that may obscure the finest details”, and that rolling and smoothing in a press “tends to restore some of the detail and also provides a glossier surface and increased contrast”, as well as improving adhesion to the mount. Small prints — cabinet portraits and stereo views especially — went through a heated roller called a burnisher, which “in skilled hands … could produce a mirror-like gloss”. Studios advertised the result under names like “French Enamel” and “Extra Superior Finish”. The Getty adds the consequence for identification: burnishing and heat treatment can suppress the microcrack network entirely, so a glossy albumen print made after 1870 may show none of the crazing that identifies an unburnished one.

🕰The gloss was the point, and at first nobody liked it

The Getty records that during the 1850s “the glossiness of albumen photographs was highly criticized in the photographic literature”, because the public was used to the matte character of salt prints — and that this “changed over time as the public developed a taste for the high definition and contrast of glossy albumen photographs.” Early printers deliberately diluted their albumen to avoid a gloss their customers disliked, producing prints so lean that conservators now have difficulty telling them from salt prints at all.

Taste then swung the other way and kept going: double coating after about 1870, burnishing, and finally surface varnishes of beeswax, paraffin, collodion or shellac, all of which the Getty identifies by infrared spectroscopy on surviving prints. Matte albumen paper appears in the Getty’s timeline in 1897, at the far end of the same argument, and it has its own entry in the atlas.

Double coating, in detail

The problem is stated in one line by Reilly and it is a solubility problem: without a hardening step between coats “there was no gain in thickness or amount of albumen on the sheet, because the second coating step dissolved off the albumen remaining from the first coating operation.”

Three solutions exist in the sources, and they are three ways of getting to the same insoluble film.

Time. “The simplest and most widely used method in the 19th century was to store the paper in a warm loft for six months, during which time a slow curing process sufficiently hardened the albumen.”

Steam. A current of steam, “which in effect cooks the albumen and renders it insoluble” — heat above 65 °C, applied to a dry layer.

Alcohol, which is the practical modern method and the one the formulary carries. A brief immersion in 70 per cent isopropyl alcohol, and Reilly is precise about why that number and not another: “Pure alcohol is too strong and unevenly coagulates the albumen layer, while too dilute alcohol solutions are not strong enough to coagulate the albumen before it partially dissolves into the water. Experience has shown that a 70% solution is the most effective.” He adds a second rule that shows how carefully the bath is matched to the sheet: whatever chloride the albumen contains must also be put into the alcohol, “if the albumen contains 2 % ammonium chloride, so should the alcohol solution”, or the chloride leaches out of the layer into the bath and you have quietly reduced the salting.

Single against double coating

Drawn to show the arrangement, not the depth; the two stacks are drawn one above the other for comparison and are not a section through one object. No source read for this course publishes a thickness for a single albumen layer, so the drawn ratio between layer and support carries no measurement. Layer depths are drawn to be readable, not to scale: on real film the base is roughly a hundred times the emulsion, and drawn honestly the emulsion would vanish. Any thickness given in the labels is the real one.

What the second coat buys, and what it costs. Reilly: two coatings “produce papers that are quite glossy and may even be so heavily coated that they are brittle and hard to tone”; double-coated paper “usually produces better prints from thin negatives than single-coated paper does”; and “a large portion of the albumen paper sold after 1880 was of the double-coated variety”. Against that: “double- coated papers have a greater tendency to curl, and are harder to manipulate in sensitization and printing”, and toning and fixing are both more difficult “because the albumen becomes increasingly less permeable as the coating thickness increases.” Any variation in coating thickness shows up immediately in the toning step, because the thinner areas tone faster and deeper.

One practical detail that is pure craft and pure physics. Gravity thins the coating at the top of a hanging sheet and thickens it at the bottom, every time. Reilly’s remedy is to mark the edge that hung lowest during the first drying and hang that edge uppermost after the second float, so the two gradients cancel. It costs nothing and it is the difference between a batch that tones evenly and one that does not.

✕Two different hardening questions, and only one of them has a single answer

It is easy to read the sources as saying that every albumenised sheet must be cured before use. They do not quite say that, and the difference matters at the bench.

Between two coats, hardening is compulsory and the sources agree: without it the second float dissolves the first layer and you gain nothing.

Before sensitising a single-coated sheet, the sources differ. Reilly gives no curing step at all for a single coat — the sheet goes from the drying line to the silver bath, and the silver bath coagulates it, which is his stated reason why albumen does not wash off during processing. Bostick & Sullivan harden every coat with heat before it meets anything else, including the last one. The nineteenth-century trade’s six-month loft cured whatever was in it.

What the course does, and why. The coating lab follows Reilly for a single coat and hardens between coats; it records the choice, because that is a variable a later comparison will want. What makes the difference survivable either way is the discipline in the sensitising step: the printing lab treats the float time and the bath temperature as limits rather than guidelines, because on a sheet that has not been separately hardened the silver bath is in a race between coagulating the layer and dissolving it. The failure has its own troubleshooting entry.

An industry built on eggs

The chemistry above explains why albumen paper was a handicraft product long after everything around it had been mechanised, and the scale of that handicraft is worth having in numbers.

Reilly’s figure is for one Dresden firm in one year. The Dresdener Albuminfabriken A.G. produced 18,674 reams of albumen paper in 1888. A ream was 480 sheets of 46 × 58 cm; coating a ream took 9 litres of albumen solution, “obtained from 27 dozen eggs”. Multiply it out and that single company consumed over six million eggs in the year — Reilly’s own arithmetic, and the course has checked it: 18,674 × 324 eggs is 6.05 million.

🕰What the egg figures are and are not

Numbers like “six million eggs” circulate widely and usually without a source. This one has a real one: Reilly cites a nineteenth-century reference for the ream production of a named company in a named year, and states the two conversion factors — 9 litres per ream, 27 dozen eggs per 9 litres — that turn it into eggs. That is a company figure multiplied by two trade constants, and it is not the same kind of claim as a national or global egg-consumption total.

Larger figures for the whole industry are common and the course has not verified any of them. No document read for this part gives a total for German, European or world albumen paper production in eggs, and none gives a method for arriving at one. Where you meet such a figure, ask which company, which year, and which conversion factors — and if the answer is not there, treat it as tradition rather than as data.

Two other facts of the trade are firmly sourced and are more useful than a headline number. Yolks were a serious by-product: preserved with salt and sold to bakeries or to tanners for finishing kid leathers, and Reilly says the money from them “was a large factor in the profit and loss statements of the giant Dresden albumen paper producers.” And the labour was gendered: the illustrations Reilly reproduces “show that nearly all the tasks in albumen paper factories were performed by women, and this was the case in both European and American factories.”

Why an egg-based industry produced a variable material. Every one of the variables this page has described was outside anybody’s control at scale. The eggs varied. The fermentation was run by naturally occurring organisms at whatever temperature the cask sat at. Each sheet was floated by hand, and some were floated twice. The paper itself came from only two mills in the world — Blanchet Frères et Kléber at Rives in France and Steinbach at Malmedy, sold outside Germany as “Saxe” — because only they could make a stock thin enough, smooth enough, free enough of the metal flecks that came from buttons left in the rags and from the machinery, and made with water pure enough. Rives drew its water from a mountain lake fed by melting alpine snow. The Getty adds that both mills sized their paper with a mixture of starch and resin soaps.

And the last twist is a marketing one that conservators still meet. From 1863 a very large proportion of albumen paper was tinted with aniline dyes — pink, purple and blue — mainly for portraits, with pink the most popular. The Getty is explicit that the tints were “added … to counteract the yellowing”; IPI’s Graphics Atlas describes a blue-tinted carte de visite on a 2½ × 4 inch card and notes that the technique was “introduced in the 1860s”. The dyes had poor light stability, so most of that paper is now unrecognisable — except that, as Reilly notes, a peculiar buff or chamois cast sometimes identifies a print that started out tinted. The industry knew about the yellowing, could not stop it, and dyed the paper instead.

What the labs will have to cope with

Five properties of this layer, each of which becomes a rule in the next two pages.

It is thin, and it is on thin paper. The Getty measured a whole unmounted albumen photograph stored flat at 0.204 mm, support included, and remarks that such prints “had a strong tendency to curl inside, forming tight rolls … rather fragile and difficult to handle without special treatment and conditioning.” Reilly says the force in a thick albumen layer is “quite enough, if a print has been incorrectly mounted, to tear the print in half.” Mounting is not decoration.

It is water-soluble until something makes it otherwise. That is the sensitising race described above, and it is why the float has a time limit.

It swells, and swelling alone changes what you see. The reddening in the first wash is a swelling effect, not a chemical one.

It blisters if a bath changes temperature or pH abruptly, and it does so worse the thicker the coating. Reilly’s remedy is to hold every solution at about 21 °C and to make the fixer slightly alkaline with 2 g of sodium carbonate per litre, which “promotes more effective fixation and helps to prevent the blistering of the albumen layer on heavily albumenized papers” — the reason his fixing bath has its own formulary entry.

It cracks with age. The Getty’s identification section calls a fine, rather uniform network of surface microcracks “the most typical microscopic signature of albumen photographs”, found in most prints made before 1870; the AIC attributes overall cracking of the binder to “variations in humidity that caused differential dimensional changes between the image layer and the support.” A brittle protein film glued to a hygroscopic paper support, cycled through a century of seasons, does what any rigid coating on a moving substrate does. That is the subject of the last page of this part.

An experiment you can run before either lab

🧪Two drops, one variable

Break one egg. Put a drop of the white on a clean glass slide and leave it to dry. Take the rest, beat it to a full froth with a fork, let it stand covered for an hour, and put a drop of the liquid that has drained beneath the froth on a second slide beside the first. Let both dry completely.

What to look at. The evenness of the two dried films at a glancing angle, the presence of ridges and strings in the first and their absence in the second, and whether either has dried clear. Then put a drop of water on each and watch how quickly it takes hold.

What it tests. Reilly’s claim that beating “results in a homogeneous liquid which will form an even layer on the sheet of paper”, and his claim that the dried material is a brittle transparent mass. It does not test coagulation, because neither film has met heat, alcohol or a metal salt, and both should still be attacked by water. Record what you saw; you will want it when a coated sheet comes out streaky and you are trying to decide whether the fault is in the albumen or in the floating.

Recap

Egg white is 88 per cent water and about 11 per cent protein, and over half that protein is ovalbumin. The proteins behave differently from one another, which is why native egg white cannot be coated, and the whole preparation — beat, salt, acidify, settle, strain, age — exists to make them behave the same. Denaturation opens the fold; coagulation links the opened chains into an insoluble film, and the four agents that do it are mechanical work, heat above 65 °C, alcohol, and the salt of a metal. The metal salt in this process is silver nitrate, which is why a single-coated sheet does not wash off the paper and why the sensitising float is a race.

The chloride is dissolved in the binder rather than laid on the paper, so the silver chloride forms inside the protein layer. Everything else follows: a higher maximum density because the image is not scattered by fibres, cleaner whites because the binder is transparent, real gloss because the top surface is smooth, a shorter exposure scale than salted paper of the same silver chemistry, and a retained silver that no wash reaches because the protein carries sulfur. Double coating doubles the protein and doubles the consequences — more gloss and depth, slower toning, worse curl, more crazing. The industry that made this material was an agricultural handicraft that ran on eggs, and it was variable for reasons this page has named one at a time.

Check your understanding

Question 1. Why is the chloride dissolved in the albumen rather than applied to the paper first, as in a salted paper print?

• Because ammonium chloride will not dissolve in water at the concentration required
• Because the silver chloride then forms inside the protein layer rather than in the paper fibres, which is what raises the maximum density and creates the gloss
• Because the chloride is needed to coagulate the albumen before the sheet is coated
• Because a separate salting bath would wash the albumen off the sheet

Show the answer and why

Answer: Because the silver chloride then forms inside the protein layer rather than in the paper fibres, which is what raises the maximum density and creates the gloss

The chloride is where the silver chloride will form. Put it in the binder and the image forms in a compact layer on the surface; put it on the paper and the image forms among the fibres, where Reilly says the reflected light is scattered and the maximum density is lowered. The third option is not wholly wrong — Reilly does note that the chlorides help denature the albumen and reduce the froth volume — but that is a side effect, not the reason the process is built this way.

Question 2. A printer floats a second coat of albumen onto a dried but unhardened first coat, and finds the sheet is no thicker and no glossier than a single-coated one. What happened?

• The second float was too short for the albumen to take
• The first coat dissolved back into the tray during the second float
• The albumen had aged too far and lost its viscosity
• The paper was too heavily sized to accept a second coat

Show the answer and why

Answer: The first coat dissolved back into the tray during the second float

Reilly states it directly: without a hardening step "there was no gain in thickness or amount of albumen on the sheet, because the second coating step dissolved off the albumen remaining from the first coating operation." Worse, the dissolved albumen is now in the tray, contaminating the batch. The three cures in the sources are six months in a warm loft, a current of steam, or a brief immersion in 70 per cent isopropanol carrying the same chloride as the albumen.

Question 3. Which of these are established by the sources this page cites, rather than inferred? Select all that apply.

• Albumen is coagulated by temperatures above 65 °C
• Ovalbumin carries four free cysteine thiols per molecule
• The silver retained in the highlights of a fixed albumen print is bonded to cysteine sulfur
• A thoroughly fixed and washed albumen sheet that was never exposed still held nearly 5 per cent of the silver applied to it

Show the answer and why

Answer: Albumen is coagulated by temperatures above 65 °C, Ovalbumin carries four free cysteine thiols per molecule, A thoroughly fixed and washed albumen sheet that was never exposed still held nearly 5 per cent of the silver applied to it

The third is the inference, and it is the one most often written as a fact. Reilly, the AIC and Ware all give sulfur-containing side groups as the binding site, and it is a well-motivated explanation — but none of the documents read for this part reports a spectroscopic identification of a silver–sulfur bond in an albumen print, a stoichiometry or a stability constant. The retention itself is measured, by Haddon and Grundy in the mid-1890s. The 65 °C is Reilly. The four free thiols are counted from UniProt: six cysteines, one annotated disulfide.

Question 4. Hübl measured the scale length of glossy albumen paper as equal to that of platinum paper, with salted paper exceeding both. What should a printer do with that finding when making a negative for albumen?

• Make a negative of the same density range as for platinum, but favouring highlight detail over shadow detail
• Make a negative of the same density range as for plain salted paper
• Make a negative of the same density range as for a grade 2 enlarging paper
• Nothing: scale length is a property of the paper and cannot be matched by a negative

Show the answer and why

Answer: Make a negative of the same density range as for platinum, but favouring highlight detail over shadow detail

Same total scale as platinum, but a different shape within it. Hübl found albumen has a slow progression from the shadows to the middletones and an abrupt jump from the middletones to white, where platinum is the opposite; so the density range matches platinum, and the negative should carry the extra information in the highlights where albumen has the higher local contrast. A grade 2 enlarging paper wants an exposure scale of about 0.9, less than half what any process in this cluster needs.

Question 5. Predict, and then verify after the two labs: a single-coated and a double-coated sheet from the same albumen and the same negative. Which statement is best supported by the sources before you measure?

• The double-coated sheet will be glossier, will print better from a thin negative, and will tone more slowly and less evenly
• The double-coated sheet will be glossier and will tone faster because there is more binder to hold the gold
• The two will differ in gloss only; density and toning are set by the silver bath
• The single-coated sheet will reach a higher maximum density because the layer is thinner

Show the answer and why

Answer: The double-coated sheet will be glossier, will print better from a thin negative, and will tone more slowly and less evenly

All three parts are Reilly. Two coatings give "papers that are quite glossy"; double-coated paper "usually produces better prints from thin negatives"; and toning and fixing are more difficult "because the albumen becomes increasingly less permeable as the coating thickness increases", with any variation in thickness showing up immediately because thinner areas tone faster and deeper. Write your prediction down before the labs and compare it with a densitometer reading and a raking-light look afterwards — a prediction you did not record is not a prediction.

Check my answersClear