History of Science Museum: Collection Database Search


Herschel's Photographic Experiments: General Description and Outline

Herschel's Photographic Experiments


Sir John Herschel's photographic experiments were given to the Museum in 1928 by his two surviving and youngest daughters Miss Francisca Herschel and Lady Constance Lubbock. Gunther studied the material, and several items were removed and displayed in the Museum. In 1963-64 the collection was lent to Schultze (Librarian of the RPS and Curator of the Kodak Museum) for more detailed study, and in 1965 he published a description in The Journal of Photographic Science, which remains the essential guide to the material even though it contains some inaccuracies (and omits the items extracted or displayed). Various scholars and students have studied or utilised the material since Schultze's paper made it better known, and it figures notably in works on the early history of photography by Schaaf and by Ware. Although the permanent display of early photographs in the Museum was discontinued in the 1980s (for conservation reasons), it has featured in several temporary exhibitions, notably those of 1981 (at the RPS) and of 1989 (at the Museum), both with accompanying booklet.

The Herschel material is, with the exception of the experiments of the inventor himself, Fox Talbot, the largest collection of experimental photographs from the actual time of the invention in existence. Unlike Talbot's, Herschel's experiments were very neat and systematic; they were properly conducted chemical experiments; they include 'controls' in the form of raw paper samples, prepared paper samples, sensitised papers submitted to a deliberate ageing test without exposure, as well as proper exposed photographs with images. They are not of course 'photographs' in any common, pictorial sense, they are experiments in photographic chemistry and procedure. Only four of Herschel's own experiments are camera images, and only three are contact copies of real objects (2 lace, 1 leaf), the rest being contact copies exposed in sunlight of small engravings. What Herschel's approach sacrifices in visual variety and curiosity to Talbot's camera images and botanical specimens, it gains in scientific credibility, for a camera image has too many variables and is unrepeatable; Herschel's dogged concentration on contact copying engravings makes them efficient and repeatable chemical experiments.

The work consisted of fundamental experiments (by which I mean, essential to the viability of Talbot's invention) in sensitising, fixing, bleaching, and positive-printing which Herschel conducted (in regular communication with Talbot) from January 29 (a few days after the invention was announced) to November 30, 1839; and then (after an intermission caused by his moving house from Slough to Hawkhurst) a second phase of experiments exploring alternative (non-silver) sensitising agents and procedures during 1841, 1842, and 1843 (only one experiment in the collection is dated 1844).

In the first phase in 1839 Herschel invented and perfected the procedure of fixing with hypo (hyposulphite of soda, now called sodium thiosulphate) which was universally used for all silver-based processes throughout pre-digital photography; he established the procedural techniques and requirements for negative-positive printing; and he made significant improvements to sensitising, paper selection, and washing procedures (or treatment with the chemical H2O, water); he conducted extensive fixing and image-cleaning trials with other chemicals alongside the hypo experiments, and he anticipated several important subsequent discoveries and techniques (such as the creation of blue images from potassium ferrocyanide, the waxing of negatives for enhanced translucency, the practice of retouching, and even the latent image phenomenon, officially discovered by Talbot in September 1840).

In the second phase of 1841-43 occur the experiments by which Herschel invented several alternative processes, ranging from the scientifically interesting but impractical phytotype (using light-sensitive juices or dyes extracted from flower petals) and amphitype (sensitised by mercury salts), via the beautiful purple chrysotype (iron-based, but deriving its colour from gold chloride), to the eminently practical and subsequently widely used cyanotype (or blueprint, also iron-based instead of silver, and deriving its colour from potassium ferricyanide). The core of the Museum's Herschel collection consists of the original experimental trials for all these (including failures as well as successes), and many variants and byways. Curious byways include 'invisible writing' experiments, one of them conducted for the entertainment of his children; and 'vapour' images, a latent image process brought out by breathing on them.

The collection consists essentially of numerous small pieces of paper (generally about 5 by 4 inches or less), which were stored for many years very compactly in a most inappropriate little box about 8 inches square and only 3½ inches deep. An exact count has never before been attempted (or anyway Schultze did not publish a figure), "over 500" was my 1989 estimate. As a result of the present cataloguing project, we now know that there are 697 items (pieces of paper) in total, which includes all papers original to the collection, including Herschel's wrappers to the various packets, some manuscripts notes, a couple of drawings, the engravings used as test objects, larger sequences of raw and prepared papers, and a core 329 papers which are actually photographic or photo-chemical experiments. The latter figure includes 9 by Talbot and 6 by Hunt which were given to Herschel in 1839 by those experimenters. I am not sure whether any other examples of Hunt's original 1839 experiments are known to exist at all. A statistical breakdown of the entire collection is appended below.

Except for a small number of loose items, and two groups that were loose or without packets because of their slightly larger size, they are contained in 32 packets or wrappers, all original and bearing Herschel's inscription or title. As their arrangement had evidently been disturbed, Schultze listed them in a rational sequence of his own, by type and date, and the Museum numbered the packets in Schultze's sequence. There are 39 'packets' or groups counting those without actual wrappers, and also counting the several extracts restored to the collection which Schultze was unaware of. Several packets are (and always were) empty (raw and prepared paper samples only); a few consist of just one or two items; but the 'ageing test' packet contains 102 papers (69 sensitised papers and 33 scrap-paper spacers), and the so-called 'Christie letter' packet contains 62 actual experimental photographs. The present cataloguing project assigns an inventory number (a random five-figure number) to each item, consistent with the Museum's computerised cataloguing procedure, uniquely identifying the item and its catalogue entry; but the old numbering system (by packet and item-within-packet) remains essential to its location or physical place within the sequence.

Herschel's thumbprint is to be found on item 75495.


actual experimental photographs by herschel 308
actual experimental photographs by talbot 9
actual experimental photographs by hunt 6

invisible writing experiments by herschel 3

prepared paper samples by herschel 55
prepared paper samples by herschel constituting an experimental ageing test begun in september 1839 (the world's longest running scientific experiment) 69
prepared paper samples by hunt 9

raw paper samples 99

scrap paper used as spacers (some containing manuscript, eg the torn-up letter) 33
manuscript notes 11
original wrappers (formerly) containing the collection 33*

sheets of mica 3 (and 3 more actually attached to engravings)

engravings (or other prints) used as test objects 54

specimens of photolithography 2
printed instruction sheet for hunt's process 1

drawing of fir trees 1
drawing of Greek temple 1

= 697 pieces of paper

* note that I say 32 in the text above ; 32 is right (for separately wrapped packets), but one of them has a second internal wrapper, so 33 is right here!

Other narratives:

Related Objects: