If you live in the UK and have a serious interest in photography, there's a good chance that you have already visited the National Museum of Photography, Film & Television at Bradford - if not, then remedy the omission soon! The permanent displays offer a fascinating insight into the history of photography and the changing exhibitions reflect the best of photographic art and practice, both contemporary and traditional. But few of the three-quarters of a million visitors each year are aware that behind the public attractions, the museum houses another world: the strong-rooms that hold the 'crown jewels' of photography.

Among these national treasures are the first silver photographs to be made on paper - products of the genius of William Henry Fox Talbot during the years 1835-41. His earliest prints have a strange beauty, due in part to their extraordinary colours, and they deserve to be seen by a wider public than a handful of scholars. But is it safe to expose them to the hardships of public exhibition? That was the question Roger Taylor, Senior Curator of Photographs, put to me in the autumn of 1992.

Accreditation

My previous experience of early photographs had been with the iron-based non-silver processes of the nineteenth century so I couldn't be described as a silver halide emulsion scientist, and it seemed to me doubtful that I was the right person to try to answer this question. Talbot's procedures, however, involved chemistry rather different from that of the development emulsions of the modern photographic industry, and mainstream photographic science has quite understandably not paid much attention to these obsolete printing-out processes for nearly a century. So perhaps there was a role here after all for a 'latterday alchemist', although there was still a problem of credibility in one quarter: the choice of a chemist to study this material was causing consternation among some photograph conservators. Maybe they were suffering nightmare visions of a prototypical mad scientist plopping priceless Talbotypes into beakers of boiling nitric acid! Some reassurance was needed that chemical science today is a touch more sophisticated than that.

Stabilization

The technical background to the problem is easily appreciated by any practising photographer: Talbot did not fix his earliest photographs (he called them Photogenic Drawings) with 'Hypo' which, as you know, dissolves out all the unexposed silver chloride; rather, he stabilised them by treatment with a strong solution of common salt, which left silver chloride behind in the highlights of his finished pictures. This treatment certainly diminished their light-sensitivity, but they might still fog in a strong light with some ultraviolet content. (For the sake of readibility, I've simplified the situation a bit - Talbot also used bromides and iodides in his sensitizers and 'fixers' - but chloride was the commonest.) The main issue then became: how sensitive are these stabilised photogenic drawings to light?

Investigation

There was no question at the outset of experimenting directly on Talbot originals, because the items in the Collection were sacrosanct. In deference to the preciousness of the material, I had to rule out even the so-called 'non-destructive' techniques of analysis, such as X-ray fluorescence spectrometry, which my research student at the V&A, Jacqui Rees, had been applying with such success to the study of nineteenth-century platinotypes.

How then to assess the vulnerability of the Talbots? Lacking this knowledge of the material, its curators could not make a properly informed decision whether it was safe to exhibit - or even to look at! I would be working with one hand tied behind my back, because I was deprived of the scientist's preferred tool - direct experiment, but I could see three remaining routes into the problem.

First, I should be able to calculate how quickly silver chloride would fog under gallery illumination, from the known photosensitivity of this substance. Of course, this wasn't at all the same thing as a complex Talbot photogenic drawing, but it would at least provide a starting point.

The second line of enquiry was more problematic. The consternation about the sensitivity of this material evidently had some basis in fact: I heard dark rumours that priceless original Talbot prints had suffered damage by being exhibited. If the relevant facts could be teased out from the natural reticence of those in the know, they would provide vital 'experimental' data, but he sensitivity of this subject was clearly less photochemical than diplomatic!

Thirdly, by following Talbot's procedures I could replicate his photogenic drawing paper, and then test it destructively without ethical qualms - assuming again that the resemblance to original material was significantly close. Unknown to me, this approach had already been tried in the 1980s by photograph conservators, yielding results that gave cause for concern, but in 1992 this evidence still remained unpublished.

Calculation

My first option, however, was the 'armchair research' - well suited to a fireside winter in snowy Buxton - dabbling with some straightforward quantum theory. Armed with the photographic scientists' bible, The Theory of the Photographic Process, together with copies of The Museum Environment and Calvert and Pitts Photochemistry, I eventually emerged from a mess of algebraic calculation with a rough numerical result which expressed the sensitivity in practical terms that might be useful to a curator. This number was the length of time that the object could be exposed, under the most stringent standard of gallery illumination (50 lux tungsten light), before suffering a change that was perceptible to the unaided human eye. The Threshold Exposure Lifetime (TEL, for short) seemed an apt name for this time span. Any exposure longer than one TEL could justly be argued to have crossed the threshold of perceptible damage to the object, which amounted to a change of about 0.01 in its optical density.

The TEL for pure silver chloride turned out to be about three hours (given the likely errors in my estimate, this means the figure probably lies between one and ten hours). That was the first piece of warning evidence for Roger Taylor at the Museum. But this finding didn't dash all our hopes of an exhibition: it is well-known that silver chloride is only sensitive to ultraviolet (UV) and violet light, so it might be possible to protect the photogenic drawings by very judiciously filtering the gallery illumination, which would not spoil the rendering of their subtle and beautiful colours. Such a safeguard is no more than a refined version of the principle of the safelight that we all use in our darkrooms. Fortunately, the protection of sensitive exhibits by UV-absorbing filters had already been investigated by a scientist at the National Gallery, Dr. David Saunders, and our problem with early photographs just called for an extension of his previous calculations, which he readily and generously carried out. He concluded that we could absorb the UV and violet-blue portions of the spectrum with steep-cut interference filters over the gallery lamps without significantly spoiling the colour rendering. That should give us a protection factor of about a thousand times. It looked as if the idea of an exhibition might, after all, be on. But then came the bombshell.

Revelation

My probing for information flushed out a transatlantic revelation: it was privately disclosed that a Talbot photogenic drawing had indeed suffered severe fogging while exhibited in 1989 under impeccable environmental conditions. In the interests of scientific objectivity, I should add that the details of this contretemps were finally made public four years after the event.

Obviously the display of similar material was now out of the question, and it seemed there was nothing left but to tidy up the loose ends. From the description of the ill-fated Talbot, the total light exposure and density change could be estimated and the TEL calculated: again, about three hours. This looked like nice agreement between theory and 'experiment' - however unintentional the latter may have been! Or was it nice agreement? Something must be wrong with my theoretical assumptions, because I was told that the Talbot had been exhibited under illumination from which all the UV had been removed; moreover it was glazed with UV-absorbing plastic. 'Belt and braces', indeed! And yet the print had still fogged quite rapidly. Evidently we could not assume that a photogenic drawing has the same spectral light sensitivity as pure silver chloride. But why not?

Sensitization

To satisfy my scientific curiosity, I had to dig deeper, and soon the answer popped out of a rarely-visited corner of the technical literature. It was the Becquerel Effect. When pure silver chloride, which is colourless, is exposed to UV light, it acquires a purplish-lilac tint due to the release of tiny particles of silver metal trapped inside the crystals. This colour means that the crystal is now absorbing light in the middle of the visible spectrum and the energy so absorbed can cause the formation of yet more silver particles. So there is an ongoing print-out by visible light, because the silver chloride crystal has become panchromatically sensitized by the colloidal silver generated within it in the first instance. It's a bit like the fable of the Trojan Horse. In fact, this panchromatic sensitization was actually employed to photograph light of long wavelengths, in the red and even infra-red regions of the spectrum, before Vogel invented dye-sensitized emulsions in the last century.

Now, any Talbot photogenic drawing has almost inevitably seen a bit of UV light from the sun or daylight during its 150 year history - a few minutes is enough to impart a rather pretty lilac 'veil' to the high values, which Talbot called 'sunned highlights' and considered to be one of its attractions. The Becquerel effect then makes the image sensitive to light across the whole visible spectrum, so it cannot be protected simply by removing the UV and blue radiation. The only complete answer is total darkness.

This was a disappointing and unwelcome conclusion to relay back to Bradford, but at least the curators would now know the worst and could take steps to use facsimiles for the purposes of exhibition and scholarship, whenever possible. Even the photographing of the material holds dangers, but it can be performed safely by minimising the light exposure with well-filtered flash, rather than photofloods.

Fixation

By now, it may have occurred to some photographically-experienced readers, that a photogenic drawing could be protected from further fogging by re-fixing it in 'hypo'. This would indeed dissolve away the excess silver chloride but, alas, it would also transform the image beyond recognition - the beautiful lilac highlights would disappear and the magenta tones in the shadows would also be profoundly shifted towards dull brown. To work such a transformation on a precious object would be quite unacceptable. These colours, by the way, are a consequence of the silver particles being much smaller than the wavelength of light, unlike those in a modern 'black and white' photo, but the full explanation of colour in colloidal silver is too indigestible a topic for this article.

While thinking about these early photographs, another significant fact that came to light was the extraordinarily small amounts of silver they contain - about one tenth of that in modern prints. I estimated that the image in a typical photogenic drawing is constituted of no more than one thousandth of a gram of silver. Considering how susceptible this metal is to attack, as a chemist I marvel that any such images have survived for as long as 150 years; we have cause to be thankful.

Experimentation

Having outlined my approaches to the problem via 'theory' and 'case history', I'll briefly describe the third route involving experiments on replica material. As a prerequisite, naturally, I had to read and digest all that Talbot himself had written about his working methods - not only the published accounts in the Proceedings of the Royal Society and elsewhere, which were often lamentably sketchy in their technical details, but also his personal Notebooks, which are held in the Talbot Collection at Bradford. Fortunately for me, these had recently been transcribed by the American scholar and photohistorian, Dr. Larry Schaaf, who very kindly let me use a pre-publication copy, so I didn't have to struggle with deciphering Talbot's handwriting - just his archaic chemical terminology! It became evident that Henry Talbot was a prolific experimentalist, and had tried dozens of variations on the process of making silver images. We could never be absolutely sure that any photograph from the Collection had been made in a particular way, because he did not annotate his experiments. Despite this uncertainty, it seemed worth proceeding by replicating his most popular material.

Having prepared, exposed, and chloride-stabilised some photogenic drawing paper according to Talbot's recipes, I took it to the Kodak Research Laboratories at Harrow, where I was welcomed by Hilary Graves, recently chairperson of the Imaging Science and Technology Group of the RPS and a recipient this year of the Society's Fenton Medal. With the aid of Hilary's colleagues, the paper was tested in one of Kodak's state-of-the-art densitometers, from which within minutes I was able to plot its rate of fogging. After making some corrections for high-intensity reciprocity failure, the TEL could be calculated: it came out again at about three hours. Bingo! A full house. All three independent paths of investigation had led to roughly the same conclusion. It was also easy to confirm the operation of the Becquerel effect in the chloride-fixed photogenic drawing paper by putting a yellow filter in the light path. The rate of fogging only dropped to half its previous value, which clearly demonstrated the sensitivity of this material to light of long wavelengths.

Although these experiments were only a preliminary skirmish in the battle for an understanding of this material, they could pave the way for a full study of the feasibility of interactive testing - if support for such a programme could ever be found. Given instruments sensitive enough to measure the rate of fogging invisibly, a technique for testing original material could become ethically acceptable.

Publication

As the outcome of this little piece of photochemical detective work, a number of findings have come to light - even if the Talbots may never do likewise! Some of the observations, relating to the procedures of early photography, may interest photohistorians; other results are more relevant to the curators and conservators who are responsible for the care of similar vulnerable early pictures elsewhere.

It was agreed at the Science Museum that the conclusions should be opened to debate by a wider public by preparing a publication. With Roger Taylor's encouragement, I undertook the further work needed to broaden the original remit and expand it into a small book. Here you can (if you are so minded!) find a full account of the chemistry of early silver photographic processes and the problems of their identification and deterioration, including recommendations for their conservation; together with details of the calculations underpinning the photochemistry and some attempt to understand in scientific terms why Talbot's photographs look now as they do.

First published in The Photographic Journal, 135 (1), 38-40, (January 1995).

My book 'Mechanisms of image deterioration in early photographs - the sensitivity to light of W.H.F. Talbot's halide-fixed images 1834-1844', is published by the Science Museum and the National Museum of Photography, Film & Television, 1994.
ISBN 0 901805 78 5, paperback.

To purchase 'Mechanisms', please visit the Siderotype.com website.