Cyanomicon II

Cyanomicon
History, Science and Art of Cyanotype:

photographic printing in Prussian blue
Mike Ware
MA DPHIL CCHEM FRSC
Honorary Fellow in Chemistry, University of Manchester, UK

Scientific Advisor to the National Gallery of Art, Washington, DC
Revised & Illustrated 2016

©Mike Ware 2017 Cyanomicon_II 2
Table of Contents
Warning and Disclaimer ............................................................................................ 6
Preface .............................................................................................................................. 7
1 Introduction in Monochrome ............................................................................. 8
1.1 The legacy of John Herschel ........................................................... 9
1.2 Equivocal blue images ................................................................. 11
1.3 Acceptance of blue photographs.................................................. 14
1.4 Commercial use of blueprint ........................................................ 16
1.5 Artistic renaissance of cyanotype ................................................. 18
1.6 A rare colour in nature ................................................................. 18
1.7 Symbolic blue in art and religion.................................................. 22
1.8 Liberation from blue preciousness ............................................... 26
1.9 Colour language and coordinates ................................................ 29
2 The Invention of Cyanotype............................................................................. 33
2.1 A brief history of Prussian blue .................................................... 33
2.2 Light-sensitivity of iron salts ....................................................... 36
2.3 The first silver photographs on paper .......................................... 39
2.4 Herschel’s research records ......................................................... 40
2.5 Anthotype and phytotype............................................................. 42
2.6 Discovery of cyanotype ................................................................ 43
2.7 Invention of siderotype ................................................................ 47
2.8 Publication of siderotype ............................................................. 51
2.9 Misapprehension of Herschel’s processes .................................... 55
2.10 An unaccountable omission ....................................................... 58
2.11 Pursuit of photographic permanence ......................................... 61
2.12 Misappropriation of cyanotype................................................... 62
3 The Molecular Basis of Blueprinting ............................................................ 67
3.1 Chemistry of Prussian blue .......................................................... 67
3.2 Molecular structure of Prussian blue ............................................ 71
3.3 Prussian blue for ion exchange .................................................... 72
3.4 Prussian blue for electrochromic displays .................................... 75
3.5 Chemical history of siderotype..................................................... 75
3.6 Photochemical principles ............................................................. 76
3.7 Chemistry of blueprinting ............................................................ 78
4 Historical Variations on Cyanotype .............................................................. 80
4.1 Herschel’s formulae ..................................................................... 80
4.1.1 Proto-cyanotype .................................................................... 80
4.1.2 Negative-working cyanotype ................................................. 82
4.1.3 Hydrargyro-cyanotype ........................................................... 84
4.1.4 Positive-working cyanotype ................................................... 84
4.1.5 Anachronistic ‘Herschelotype’................................................ 85
4.1.6 Light-bleaching of Prussian blue ........................................... 86
4.2 Pellet’s process ............................................................................ 87
4.3 Survey of negative-working formulae .......................................... 90
4.4 Modifications of the cyanotype sensitizer .................................... 93
4.4.1 Robert Hunt’s ‘chromo-cyanotype’ ........................................ 93
2
4.4.2 John Mercer’s ‘chromatic photographs’ ................................. 93

4.4.3 Charles Burnett’s ‘cobalt cyanotype’ ...................................... 95
4.4.4 Other iron(III) sensitizers ....................................................... 96
4.4.5 Terry King’s ‘cyanotype rex’ .................................................. 98
4.4.6 Cyanotype using uranium and vanadium salts ..................... 100
4.5 Cyanotype diapositives on glass and film ................................... 102
4.6 Cyanotype negatives made in camera ........................................ 104
4.7 Printing cyanotypes by projection .............................................. 106
4.8 Cyanotypes on ceramics ............................................................ 106
4.9 Cyanotypes on fabrics and textiles ............................................ 108
4.10 Gel lithography: ‘true-to-scale’ process .................................. 110
4.11 Commercial blueprint papers ................................................... 112
4.12 Type B blueprint papers-Blue Layer ......................................... 116
4.13 Type AB blueprint paper .......................................................... 118
5 Studies in Blue 1842-2013 ........................................................................... 120
5.1 Herschel’s use of engravings .................................................... 120
5.2 Identifying the sources of engravings ....................................... 122
5.3 Herschel’s cyanotype tests ........................................................ 128
5.4 Cyanotypes of British algae by Anna Atkins............................... 134
5.5 Albums by Anna Atkins and Anne Dixon ................................... 141
5.6 An anonymous cyanotype album of 1843 ................................. 145
5.7 Later botanical cyanotypes........................................................ 147
5.8 Julia Herschel’s lacemaking handbooks..................................... 149
5.9 Early pictorial studies in cyanotype ........................................... 150
5.10 Documentary and topographical cyanotypes ............................ 153
5.11 Cyanotype currency notes and postage stamps ........................ 160
5.12 Linley Sambourne’s cyanotypes ............................................... 163
5.13 Some contemporary cyanotypists ............................................. 164
6 Methodology of Contact-printing Practice ............................................. 168
6.1 Diaphanes: a neologism............................................................. 168
6.2 Skiagraphy of photograms ......................................................... 170
6.3 The processing environment ...................................................... 174
6.4 Materials and equipment ........................................................... 174
6.4.1 Paper characteristics............................................................ 175
6.4.2 Using glass rods to coat paper ............................................ 178
6.4.3 Contact-printing frames ...................................................... 181
6.4.4 Ultra-violet light sources ..................................................... 183
6.5 Making suitable silver-gelatin negatives .................................... 185
6.6 Inkjet-printed digital negatives ................................................. 188
6.6.1 Summary and purpose ......................................................... 188
6.6.2 Calibration of equipment and materials ............................... 189
6.6.3 Fine-tuning negative opacities ............................................ 191
6.6.4 The 'gamma slider' for tonal correction................................ 192
6.6.5 Workflow for digital negatives ............................................. 194
6.6.6 Short version of the workflow .............................................. 200
7 Formulary for Cyanotype Sensitizers ........................................................ 201
7.1 The traditional cyanotype process.............................................. 201
7.1.1 Making up the sensitizer solutions ...................................... 202
3
7.1.2 Mixing and coating .............................................................. 202

7.1.3 Printing exposure ................................................................ 203
7.1.4 Wet processing and reoxidation........................................... 203
7.2 An improved cyanotype recipe ................................................... 203
7.3 Process shortcomings and their remedy..................................... 204
7.4 The new cyanotype process ....................................................... 209
7.4.1 Making up the sensitizer ..................................................... 210
7.4.2 Use of dichromate ............................................................... 214
7.4.3 Use of surfactant ................................................................. 214
7.4.4 Choice of papers ................................................................. 214
7.4.5 Drying ................................................................................. 215
7.4.6 Citric acid prevents fogging ................................................. 216
7.4.7 Negatives and exposure ...................................................... 216
7.4.8 Wet process and contrast control ......................................... 216
7.4.9 Washing and reoxidation ..................................................... 217
7.4.10 New cyanotype: workflow summary ................................... 219
7.4.11 Effects of heat-drying and ‘darktime’ ................................ 220
7.4.12 Diagnosis of fogged highlights .......................................... 221
7.5 Cyanotype composites with other processes .............................. 223
7.6 List of chemicals and hazards .................................................... 224
7.7 Environmental issues and disposal............................................. 229
7.8 Suppliers of chemicals and equipment ....................................... 230
8 Methods for Toning Cyanotypes ................................................................ 233
8.1 Hydrolysis of Prussian blue ........................................................ 234
8.2 Ferro-gallate and tannate .......................................................... 234
8.3 Assorted beverages ................................................................... 235
8.4 Iron-mordanted dyes................................................................. 235
8.5 Black toners ............................................................................... 236
8.6 Heavy metal incorporation ......................................................... 237
8.6.1 Lead .................................................................................... 237
8.6.2 Thallium .............................................................................. 238
8.6.3 Nickel .................................................................................. 239
8.6.4 Copper ................................................................................ 240
8.7 Insoluble metal ferricyanides ..................................................... 241
8.8 Blue toning of silver prints ......................................................... 241
9 Vulnerability & Conservation of Cyanotypes ......................................... 242
9.1 Fading of cyanotypes by light .................................................... 243
9.1.1 Experimental: preparation of specimens .............................. 244
9.1.2 Experimental: fading apparatus and measurements ............. 246
9.1.3 Results: characteristic curves and the ‘fade’......................... 247
9.1.4 Results: variation of fade with initial density ........................ 249
9.1.5 Results: variation of fade with exposure time ...................... 251
9.1.6 Results: reversibility of repeated fade and regain ................. 253
9.1.7 Results: effect of conditions on fading ................................. 255
9.1.8 Conclusions: mechanism of light-fading ............................. 257
9.1.9 Regain of colour by faded Prussian blue .............................. 261
9.1.10 Recent spectrometry of Prussian blue fading ..................... 262
9.2 Bleaching of cyanotypes by alkali ............................................... 262
4
9.3 Peptization of Prussian blue ....................................................... 265

9.4 Conservation issues with cyanotypes ......................................... 267
9.4.1 Recognition of cyanotype .................................................... 267
9.4.2 Exhibition and light sensitivity ............................................. 268
9.4.3 Safe copying techniques ...................................................... 269
9.4.4 Avoidance of alkali .............................................................. 273
9.4.5 Buffered substrates ............................................................. 274
9.4.6 Storage enclosures .............................................................. 276
Appendix I Chronology of Cyanotype ............................................................ 279
Appendix II Chemistry of Prussian Blue ....................................................... 280
II.1 Methods of preparation ............................................................. 281
II.2 ‘Soluble’ Prussian blue ............................................................... 282
II.3 ‘Insoluble’ Prussian blue ............................................................ 285
II.4 Turnbull’s blue .......................................................................... 286
II.5 Electronic spectrum of Prussian blue .......................................... 288
II.6 Iron(III) aquapentacyanoferrate(II)............................................... 289
II.7 Investigation of cyanotype Prussian blue .................................... 290
II.8 Hydrolysis of Prussian blue ........................................................ 292
II.9 Prussian white ........................................................................... 293
II.10 Berlin green ............................................................................. 294
II.11 Prussian brown or yellow ......................................................... 295
II.12 Electrochemical studies of redox behaviour ............................. 295
Appendix III Photochemistry of Cyanotype ................................................. 299
III.1 Light-sensitive iron(III) carboxylates ......................................... 299
III.2 Quantitative aspects of the photochemistry .............................. 301
III.3 The Jones-Condit Equation ....................................................... 305
III.4 Photochemistry of trisoxalatoferrate(III) .................................... 306
III.5 Composition of ammonium iron(III) citrate ................................ 307
III.6 Photochemistry of citratoferrate(III) ........................................... 309
III.7 Photochemistry of hexacyanoferrate(III) .................................... 310
III.8 Photochemistry of the blueprint process: 1 ............................... 311
III.9 Photochemistry of the blueprint process: 2 ............................... 314
III.10 Photochemical reduction of Prussian blue ............................... 317
Appendix IV Units of Concentration ............................................................... 320
IV.1 Concentration as weight percent volume (% w/v)....................... 320
IV.2 Concentration as molarity ......................................................... 320
IV.3 Concentration as weight percent weight (% w/w) ....................... 321
IV.4 Concentration as volume percent volume (% v/v) ...................... 321
IV.5 Early formulations and recipes .................................................. 321
IV.6 Modern formulations and recipes ............................................. 322
IV.7 Interconversion of Cw and C ...................................................... 322
IV.8 Conversion of units in obsolete formulae .................................. 324
Notes and References ........................................................................................... 326
5
Warning and Disclaimer

The practical instructions contained in this text involve the handling of
some chemicals that are regarded as toxic or harmful. The chemical
hazards are summarised in §7.5, but it is the responsibility of the users
of chemicals to inform themselves about the risks, and to take
appropriate precautions in their handling. Reference should be made to
the Materials Safety Data Sheets (MSDS), which are accessible online:
http://www.ilpi.com/msds/index.html
The author hereby denies liability for any consequent sickness, injury,
damage or loss resulting from the use of the chemicals and procedures
described herein.
6
Preface
My first monograph on Cyanotype was published by the Science Museum
of London in 1999, but has long been totally out of print. My book was
devoted to the study of photographic printing in Prussian blue, engaging
with its history, aesthetics, practice, conservation and chemistry. Now, in
response to kind requests, I have substantially restructured this text in an
updated, extended, and illustrated edition that is made freely available as
a ~50 MB pdf download from the World Wide Web:
http://www.mikeware.co.uk/mikeware/downloads.html
With 800 references to the literature and the WWW, I hope it may serve
occasionally as a useful resource for historians, curators and conservators
of photographs, and students of siderotype - iron-based analogue
photographic printing - should any of these good folk ever find
themselves as castaways upon the strange blue shores of cyanotype.
For those other shipwrecked mariners - photographic artists exploring
cyanotype printmaking as an expressive medium - I have included full
practical instruction in the modern process.
Mike Ware
Buxton
January 2016
www.mikeware.co.uk
7
1 Introduction in Monochrome
"Blue, darkly, deeply, beautifully blue"
Robert Southey (1774-1843)
Poet Laureate
Out of the entire gamut of photographic printing techniques left to us by

the early pioneers of the medium, only one process survives unchanged
to the present day: cyanotype. Yet its images still seem unnatural to
many viewers; the colour intensity of Prussian blue invests them with an
incongruity that has curbed the scope of this process ever since it first
saw light in the dawn of photography. Cyanotype endures nonetheless,
and is still practised today in much the same way as 172 years ago. Its
strange blue aesthetic is now gaining a wider acceptance than hitherto,
thanks to the advancing sophistication of our photographic sensibility. So
the time now seems ripe for a review of cyanotype as an alternative
medium of photographic practice. The author’s only misgiving is that it
may seem over-ambitious, in this era of specialism, for any writer to
assume all three mantles of historian, scientist, and art-critic within a
single text; this presumption therefore calls for a short justification.
Historically, cyanotype is an undervalued process which the standard
works on photography usually dismiss in a paragraph. Despite its
marginalized status, there is much of interest and relevance to be found
here - a history that deserves to be more widely appreciated for the
added light it sheds on the early development of photography.
Scientifically, the photochemical theory of the cyanotype has attracted
rather little professional attention, and the published explanations do not
seem to account fully for its phenomenology. It is fortunate that an
intense research interest in the substance, Prussian blue, has grown up in
recent years among electrochemists with quite different technological
aims, but their interest has greatly improved our chemical understanding
of this fascinating historical pigment.
Artistically, there are strong signs that contemporary photographic
image-makers are now taking more interest in the process, to judge by
the growing presence of cyanotypes on gallery walls and in published
books. The aesthetics of the blue monochrome image have always
presented something of a problem for the connoisseurs of photography,
and the origins of this difficulty will provide us with an interesting subject
for discussion.
8
1.1 The legacy of John Herschel

Photography in Prussian blue was discovered on 23rd April, 1842 by Sir
John Frederick William Herschel (1792-1871), just over three years after
Louis Jacques Mandé Daguerre (1789-1851) and William Henry Fox
Talbot (1800-1877) had announced their independent inventions of
photography in silver, respectively using substrates of metal and paper.
The success of these two pioneers in finally securing photographs in
silver was the fruition of an idea that had been gestating for more than a
century in the minds and laboratories of several leading natural
philosophers.1 In contrast, the birth of the silverless cyanotype was totally
unexpected; it came, literally and metaphorically ‘out of the blue’, to a
‘single parent’. Why should the circumstances of these two major
photographic innovations have been so different? An attempt to answer
this question will be offered in the next chapter.
While Henry Talbot was creating several thousand images in silver and
assembling them into a creative oeuvre that is now recognised as the
richest vision in early photographic art,2 his fellow-scientist Sir John
Herschel, rather than making new pictures, was experimenting widely
with chemical processes that might prove photographic, for he was
motivated by a preoccupation with the medium rather than the message.
By retracing Herschel’s steps which led to the discovery of cyanotype we
will engage with just one facet of the life and work of this great polymath.
The account of his photographic researches offers us a paradigm for the
extraordinary interactions between chance, intuition, and inductive logic,
as they blended within one outstanding scientific mind of the 19th
century.
As one of the leading physical scientists of his day, Herschel was
driven by a desire to understand photochemical phenomena, and to enlist
them as tools for probing the electromagnetic spectrum beyond the
narrow optical limits imposed by human vision. He sought to venture
outside the shortwave end of the visible spectrum, into the region of the
ultra-violet or ‘actinic’ rays that had been discovered in 1801 by Johann
Ritter. At the other end of the optical spectrum, he explored beyond the
longwave visible limit, into the region of the infra-red or ‘thermic’ rays
that had been discovered in 1800 by his father, the celebrated
astronomer, Sir William Herschel. In part, the son intended his
investigations as a vindication of his father’s work:
“thus my father’s experiments have full ocular demonstration” 3
9
Herschel also clearly entertained a practical objective: the hope of

devising a direct photographic process in natural colour. There is no
compelling evidence to suggest that he was in pursuit of commercially
useful methods of reprography, unlike Talbot, whose clear intention to
multiply his photographic images in printers’ ink was ultimately realised
in his method of ‘photoglyphic engraving’. It is therefore a happy irony
that it was Herschel who became the inventor of the first process for
reprographic photocopying, which we now call “blueprint”.
Fig. 1.1 Sir John F.W. Herschel by Julia Margaret Cameron
Over the years since his death in 1871, Sir John Herschel’s original
writings, unpublished personal memoranda, experimental notes and
specimen images have become somewhat dispersed globally; but the
following chapters will attempt to re-unite those fragments of his
scientific legacy that are concerned with the cyanotype process, and to
interpret them in the light of a modern understanding of the properties
of the image substance, Prussian blue. The reader will find that every text
on ‘alternative’ photographic processes quotes a formula for the
cyanotype which is claimed to be ‘Herschel’s original’, but references are
never cited in support, because his publications are not explicit on the
exact formula. I have now resolved this issue by reference to Herschel’s
own experimental notes of 1842 (from the manuscript archive at the
Harry Ransom Humanities Research Center of the University of Texas, at
Austin) which reveal how he explored at least fifteen chemically distinct
ways of making images in Prussian blue. It will become clear that
10
limitations in the chemical knowledge of Herschel’s day pre-conditioned

his achievements to a significant extent, and that he owed much to a
young scientist, Dr Alfred Smee (1818-1877), whose chemical expertise
provided Herschel with information and actual samples of newly-
synthesised substances that would prove crucial to his success.
1.2 Equivocal blue images

The history of the cyanotype has a rather periodic character; within its
172 years (as of 2014) we can distinguish three main phases of use. At
first Herschel’s invention was taken up only by a small élite of botanists
for the purposes of plant illustration. The most notable achievement was
that of Anna Atkins who, during two decades from 1843, produced and
distributed her now celebrated and highly-treasured albums of botanical
photograms, as described in §5.4.4 Otherwise, Herschel’s cyanotype
found little acceptance among his peers, such as Sir David Brewster of St.
Andrews University, the leading scientific patron of the new art-science
of photography.
Fig. 1.2 "Cyanotype by Sir John Herschel"

from the Brewster Album, collection of the J. Paul Getty Museum.
11
In his scholarly exposition of the famous ‘Brewster Album’ in the J.

Paul Getty Musuem collection, the photohistorian Professor Graham Smith
has noted that it contains one of Sir John Herschel’s cyanotypes of 1842,
fig. 1.2, which was donated by Michael Pakenham Edgeworth to Brewster
in 1843,5 and concerning which Brewster confided, in a letter to Talbot:
“It is taken from an engraving, and is tolerably sharp but by no
means pleasing.” 6
Talbot himself made no use whatever of the cyanotype process; but,

much later in 1864, he did comment approvingly on its permanence in
regard to the botanical photograms of Anna Atkins (see §5.4), who had
presented him with a copy of her cyanotype work, British Algae:
Cyanotype Impressions.7
In the Great Exhibition of 1851, among a multitude of exhibits
illustrating the burgeoning art-science of photography, the cyanotype
process was represented by just one minor specimen.8 This evident lack
of practice is an indication of the insignificant status accorded to
cyanotype at the time, which may be attributable, at least in part, to
inadequate published descriptions of the best procedures for carrying out
the process, as we shall see in §2.9. The medium was relegated for a
further twenty years to its sojourn in disuse. Following Herschel’s death
in 1871, the cyanotype process was appropriated by entrepreneurs of a
more commercial turn of mind than its true inventor, and then exploited
as a reprographic medium, although Herschel himself had previously
demonstrated its use for copying text and images. The re-branded
‘ferroprussiate’ process also found some use by photographers as a
cheap and easy option for proofing negatives, but its major market was
for copying the plans in every drawing office of the world. The
commercialized version of Herschel’s invention - the ‘blueprint’ - held
sway as the chief industrial reprographic process for eighty years.
By 1900, the acceptibility of cyanotype as a medium for pictorial
photography had been seriously inhibited, at least in Britain, by the
intolerant responses of various critics to its powerful colour. The arbiters
of contemporary taste in ‘the art of photography’ were at the time
habituated to an aesthetic of monochrome images that were mostly
brown, so they declared the unremitting blue of the cyanotype to be
anathema. Foremost among these critics was Peter Henry Emerson,
always characteristically vituperative in his condemnations:
12
“ … no one but a vandal would print a landscape in red, or in

cyanotype.”
“ The Ferro-Prussiate printing process, of course, does not concern
us, blue prints are only for plans, not for art.” 9
Even R Child Bayley expressed his disdain for the process in this
sardonic observation from The Complete Photographer :
“…as to printing methods, there is the simple “blue print”, which
survives, as the Darwinians tell us some of the lower forms of life
survive, from the extreme simplicity of its structure. The blue print
needs nothing but a good washing in water to reveal its full vigour
(such as it is). … There are certain subjects which the colour is said
to suit, and I live in the hope that one day I may see such a
subject.” 10
As late as 1922, the prevailing view that cyanotype was fit only for
proofing negatives was reflected in E J Wall’s dismissal:
“One rarely wants to print from an ordinary negative in the vivid
blue of cyanotype, for the color is not suitable for ordinary views,
except possibly ice and glacier scenes, and these can be much
more effectively dealt with by the cyanotype toning of a bromide or
development print.” 11
But one writer who did not join the general chorus of condemnation in
Britain was the noted professional photographer, Frank Sutcliffe:
“An amateur of my acquaintance prints everything he does in
cyanotype blue, and I must confess that looking through his prints
from time to time gives me more pleasure than I get from looking
at any other workers prints. They never give me neuralgia in the
eyes.” 12
The anonymous “amateur” in question may have been Washington

Teasdale (1831-1903) of Headingley, Leeds, where Sutcliffe was also
born, and who used the cyanotype process extensively during the 1880s
– see §5.9.13
Despite Sutton’s interesting claim for their therapeutic value,
cyanotypes were held in low esteem by the British photographic
establishment, which deemed them both unacceptable for exhibition and
unworthy of acquisition. The prevalence of this prejudicial view is
apparent from historical evidence of a negative kind. If one searches the
major British collections of photographic art for examples of the process,
cyanotypes are found to be conspicuously absent. For instance, the Royal
Photographic Society’s collection, which is an outstanding representation
of 19th century work, had not admitted a single cyanotype, as of 1999.14
13
The Scottish National Portrait Gallery holds a magnificent photographic

collection, restricted neither to Scotland nor portraiture, but has acquired
only one small cyanotype.15 Even the Victoria & Albert Museum can only
boast, in addition to its modest fascicle of Anna Atkins, a relatively small
number of recently-acquired twentieth-century examples of cyanotype.16
The National Museum of Photography, Film, & Television, now rebranded
as the National Media Museum, naturally has important holdings of early
cyanotypes by Herschel and Atkins, but otherwise the later use of the
process in the history of photography is represented only by a few
anonymous architectural studies of little consequence. A similar state of
non-representation seems to prevail in most other major British
collections. It must also be considered significant that the - otherwise
admirable - Charts for the Recognition of Photographic Processes,
devised in 1976 by Messrs B Coe, T J Collings and A T Gill for the Royal
Photographic Society, completely omitted the cyanotype process from
their scheme of things.17 The third phase of cyanotype – its rehabilitation
as a medium for visual art - came about only in the last two decades of
the 20th century, as we shall see, with the rise of the movement for
‘alternative photography’.
1.3 Acceptance of blue photographs

Early prejudice against the cyanotype process did not prevail universally.
In Paris, the collections of the Musée d’Orsay and the Musée des Beaux
Arts contain substantial holdings of beautiful cyanotypes by Paul Burty
Haviland, fig. 1.3, and Henri Le Secq.
In the USA, the collection of the George Eastman House has specimens
of Le Secq’s architectural studies, among many other significant
cyanotypes. In Canada, the collection at the National Gallery in Ottawa
contains many interesting cyanotypes, including a fine series of
ethnographic studies of Native North Americans by Edward Curtis, fig.
1.4.
14
Fig. 1.3 Paul Burty Haviland, Florence Peterson in a Kimono, ca 1910
Neither was the process despised among the amateur community in

North America, as it largely was in Britain. One handbook published in
the USA refers to the process as “capable of such beautiful results”,18 and
John Tennant, the editor of the influential American periodical The Photo-
Miniature, conducted quite a spirited defence in one of his issues of
1900:
“This prejudice against the blue print because of its color is, in
itself, curiously interesting. In every-day life we are inclined to be
enthusiastic about everything blue, from the deep blue of the sea
or the deeper depths of blue in a woman’s eyes, to the marvellous
blue of old Delft ware or the Willow plates of years ago.” 19
But there was one North American photographer who did apparently
find the process unacceptable: the young Alvin Langdon Coburn.
15
According to Nancy Newhall,20 he deemed the process “villainous”, a

rather sophisticated judgement at the time - for an eight-year-old!
Fig. 1.4 Edward Curtis, Misham
1.4 Commercial use of blueprint

In Britain, the cyanotype suffered an almost total aesthetic boycott by
photographic artists, connoisseurs, and curators until the last two
decades of the 20th century; but in contrast, one can point to huge
archives of cyanotypes where its utility as a copying process was the
paramount consideration. The commercial success of the cyanotype
process was owed, not to its pictorial use, but its reprographic facility.
These applications have endowed our language with a new word:
blueprint - a word that endures long after the process it described
16
became obsolete because, as Frances Robertson explains, it has now

taken on the expanded and more abstract meaning of “master plan”.21
Fig. 1.5 Blueprint of a steam engine

(courtesy of Roger Farnham and Harry Magee, Glasgow Print Studio)
The era of the blueprint as a copying process was heralded by the

manufacture of the cheap, sensitized paper in huge quantities. By the
beginning of the 20th century its use for copying engineering and
architectural plans had become universal in drawing offices, fig. 1.5. In
1918, a 30 foot roll of blueprint paper a yard wide could be purchased
for as little as 1s. 6d.22 This was just as well, because consumption must
have been enormous: the plans for a battleship, for instance, required
11,000 square feet of the material!23 Although it faced two or three
competing processes, the blueprint was predominant for eighty years as
the foremost industrial reprographic process, and it was only finally
displaced in the mid-1950’s in the UK, first by the diazo print medium,
then by the invention of electrophotography, which enabled
photocopying by entirely dry methods.24 It is interesting to note that
commercial production and use of blueprint paper, though dwindling,
was still significant in 1972 in the USA, and production of the paper
survives there still, albeit on a very reduced scale, for leisure purposes.25
17
1.5 Artistic renaissance of cyanotype

Today we come full circle in witnessing a second revival of the cyanotype
process among contemporary photographic artists who, like the very first
practitioners, are mostly obliged to hand-coat their own paper or, in
some instances, more unusual surfaces, such as wood, bone, glass or
stone. There is a case to be made for the rehabilitation of cyanotype as a
fine print medium, despite - or perhaps even because of - its powerful
colour. It is not generally acknowledged that, if carefully executed with a
good chemical formulation, cyanotype is capable of great delicacy in
tonal gradation which can equal any other photographic printing
process.26 It combines the advantages of very low cost, low toxic hazard,
simplicity in manipulation, applicability to many types of surface, in very
large formats if desired and, thanks to Herschel’s magnanimity, all this
without restriction by patent or commercial secrecy. The well-known
vulnerability of cyanotypes to alkali is not a serious problem if they are
treated with nothing more than ‘benign neglect’. From a general
observation of historic specimens of photography in major archives, I can
justify the claim that most 150-year-old cyanotypes have survived better
than their silver counterparts of a comparable age.
Nonetheless, the cyanotype is still relegated to the role of humble
Cinderella among alternative printing processes. It is generally treated as
a poor cousin to silver, carbon, and platinum printing - although it would
be inappropriate to press this metaphor too far - I am not implying that
these other processes are the ‘ugly sisters’! The rest of this introduction
will be devoted to examining some possible explanations for cyanotype’s
lowly status. One reason may be that the process is perceived by some
influential commentators as so facile, cheap, and unsophisticated, that
they think it is only worth practising by the unskilled. A second reason
may stem from a deep-seated aesthetic rejection evoked by the
incongruity of the blue image, when this colour is so seldom to be seen
in nature. A third more sophisticated motive may be that, by inevitably
rendering all that it touches in an uncompromising blue, the cyanotype is
felt to devalue the very colour that had always been reserved historically
by artists to embody a sense of preciousness.
1.6 A rare colour in nature

It is an observable fact that blue pigments are very rare in the natural
world.27 Let us take a chromatic survey of the everyday hues in the
18
landscape, viewed through a chemist’s spectroscopic eye. The plant

kingdom is mainly characterised by the green photosynthetic pigment,
chlorophyll, enlivened here and there by the yellow, orange, and red of
carotenoids in flowers and the turning leaves of autumn. Tree trunks and
boughs display the deep browns of complex oxidised organic
macromolecules, lignins and tannins, and the soil is similarly coloured
with humic acids. Many rocks are stained with the reddish-browns of iron
oxides.
Fig. 1.6 Colours of landscape and sky

(Collection of iwallfinder.com)
Living creatures often display the dark tones of the pigment melanin
in hair and hide (Greek: μελας = black), which contrasts with the crimson
of myoglobin and haemoglobin, showing through in the pink and red of
flesh and blood.28 The reader can savour the rarity of blue among all
these familiar colours of nature, by contemplating, for a moment, the
unnatural prospect of a plateful of blue food.29 Blue dye is used to mark
condemned meat and hopefully ensure its non-consumption. It may also
be significant in this context that blue was the colour conventionally
chosen for the glass of ribbed poison bottles in Victorian times.
There is only one obvious source of blueness in the world which
accustoms us to our everyday familiarity with the colour: the light that
comes from the vast hemisphere of a clear sky. The presence of this
ubiquitous blue source does not contradict the rarity of blue as a pigment
because the colour of the sky is not due to absorption, but to a selective
scattering of light, an optical phenomenon first elucidated by John
Tyndall in 1869, and quantified mathematically by Lord Rayleigh in
19
1871.30 Expanses of water often reflect the blue of the sky; and if the
optical path length through water is very long, transmitted light does
appear blue owing to a very weak absorption band in the red end of the
visible spectrum of water.31
Fig. 1.7 Cyanometer for measuring the blueness of the sky ca. 1789
(Collection of Bibliothèque de Genève)
In 1789 the Swiss natural philosopher Horace-Bénédict de Saussure

(1740-1799) invented the Cyanometer to quantify and measure the
blueness of the sky,32 using dilutions of the pigment Prussian blue to
make a graded circular step-tablet, Fig. 1.7, as a comparator.33 He
concluded that the transparency of the sky was a measure of the water
vapour content of the atmosphere.
Later in the 19th Century, the blue of the sky became a central issue
for a number of quasi-scientific investigators, of whom the "leading
light", a retired US Civil War general named Augustus James Pleasonton,
20
published a memoir in 1877 describing his experiments with cultivation

of grapes - and later pigs - within a glasshouse incorporating blue filters:
"I may premise that for a long time I have thought that the blue
color of the sky, so permanent and so all-pervading, and yet so
varying in intensity of color, according to season and latitude, must
have some abiding relation and intimate connection with the living
organisms on this planet."34
This book led the way for the niche following of "chromotherapy" and,
true to its ideals, was printed with blue ink on blue paper.
Besides the sky, many of the other occurrences of blue in nature prove
often to be examples, not of absorption by pigments, but of
‘microstructural colour’; that is, chromatic phenomena generated by
optical physics rather than chemistry.35 The blues to be seen in some
birds’ plumage – the Indigo Bunting, Passerina cyanea is a fine example –
are not caused by pigment absorption but by selective light-scattering
from the feathers; the striking blue colour of some species of Morpho
butterfly, Fig. 1.8,36 and the iridescent blue carapaces of some beetles
are likewise due, not to the absorption of light, but to optical interference
effects, akin to those in an opal, a soap bubble, or an oil film on a water
surface.37
Fig. 1.8 Morpho butterfly
The one important exception that ‘proves’ the rule of rarity is the
occurrence of blue organic pigments - certain members of the family of
anthocyanine dyes - in some common flowers such as bluebells and
cornflowers. These we particularly admire for the vivid chromatic accents
that they bring to contrasts in the natural landscape. It is possible that
this botanical coloration has evolved as an adaptive characteristic,
endowing the plants with survival advantages by attracting pollinating
insects, which see them standing out against a common green
background. Bees, for example, are blind to red light but can see into the
21
ultraviolet; a blue flower may be even more conspicuous to them than to

us.
It is therefore comprehensible that the scarcity of blue dyestuffs
among naturally-occurring materials has always conferred high status or
ritual significance on the use of the colour in many early cultures. Indigo
has been valued since 2500 BC. The best natural source is the Indigofera
plant that occurs in India, as its name implies, although the woad
obtained by the Picts from Isatis is the same dyestuff.38 In the early
Mesoamerican cultures ca. 800 AD, indigo from Indigofera Suffruticosa
(the Anil plant) was incorporated into a clay, palygorskite, to provide the
remarkably stable pigment, Maya blue.
A close chemical relative of indigo is Tyrian purple,39 famously
reserved as the dye for the Imperial robes of Rome, and also known
biblically as the dye argaman, reserved for rabbinical vestments in the
Judaic tradition. Tyrian purple was first extracted by the Phoenicians from
a Mediterranean mollusc, Murex Brandaris,40 but can be sourced from
several other species and, because it is formed photochemically in
sunlight, it has even been employed for photographic printing.41 It is
curious to reflect that when Julius Caesar’s invading legions fought the
Picts in 58 BC, both the Emperor’s robes and the tribesmen’s skins were
coloured with chemically similar dyestuffs, but of rather different cultural
origins. With the common availability of synthetic blue dyes in modern
times, the distinction conferred by blue clothing has become socially
inverted, so we now have the bleu de travail of the French ouvrier, the
‘blue collar worker’ of British class culture, and the ubiquitous and
commonplace ‘blue jeans’ of denim.42
1.7 Symbolic blue in art and religion

Given the universal scarcity of blue pigmentation in the natural world, we
can understand why the cyanotype image might have been considered
‘unnatural’. Its low status in photographic art, however, still remains
somewhat paradoxical when we contrast it with the elevated role of blue
in the traditions of painting. Of all the colours in the spectrum, blue held
the place of honour among the pigments on the artist’s palette. In the
history of painting, therefore, blue was always reserved for a noble role.
Patrons of great paintings were inclined to limit their expenses by
stipulating how much ultramarine the artist might use in the work; for
this was the finest, most costly blue then available, ‘the queen of all the
pigments’, and it was made by grinding up the gem-stone lapis lazuli,
22
which was mined only in Afghanistan, and therefore came from ‘over the
sea’, hence ultra mare. The Ancient Egyptians also used ultramarine to
decorate the death mask of King Tutankamun.
Blue was a prized pigment for the decoration of fine ceramics, as
exemplified by the blue and white porcelain adopted in China, traded by
Marco Polo, with some specimens dated 1351 finally ending up in the
British Museum collection as the celebrated ‘David Vases’, Fig. 1.9.
Fig. 1.9 David Vases, 1351

(Collection of the British Museum)
The first synthetic pigment, Egyptian blue was used to decorate

ceramic artworks and to dye the swaddling clothes for the baby pharoahs
of ancient Egypt.
23
The association with the colour of the celestial hemisphere adds an

extra dimension to the symbolism of blue. Because it appears in the sky
after the obscuring clouds are dispelled, blue is said to be the ‘colour of
truth’. C J Jung conjectured that:
“…blue, standing for the vertical, means height and depth (the blue
sky above, the blue sea below).” 43
Jolan de Jacobi, in her study of Jungian psychology, says:

“…blue, the colour of the rarefied atmosphere, of the clear sky,
stands for thinking.” 44
In a religious context, blue is the colour symbolising some of the

loftiest sentiments: spiritual devotion, heavenly love, and innocence. In
the traditions of Western religious art, for instance, the Virgin Mary’s
mantle is invariably rendered in blue, and so is that of Christ during his
ministry on earth.45 In some of the most notable altarpieces of Christian
art, for instance the superb Wilton Diptych shown in Fig. 1.10, the entire
court of heaven is clad in blue.46
Fig. 1.10 The Wilton Diptych

(collection of the National Gallery, London)
24
In Byzantine art, the haloes of the principal figures are golden if they
are mere mortal emperors and prophets, but sky-blue is used for the
haloes of the spiritual beings. Blue is also the predominant colour used to
decorate the churches of the Greek Orthodox tradition. In Indian religious
painting, Krishna, one of the incarnations of Vishnu the Preserver, is
always represented with blue skin, Fig. 1.11:
“Since the infinite empty space appears as deep blue in colour, it is
but proper that Vishnu, the all-pervading cosmic power, be
depicted as blue in colour.” 47
Fig. 1.11 The Lord Krishna
J E Cirlot, in his work on symbols, notes that:

‘About the various shades of blue, ranging from near black to clear
sapphire, there has been a great deal of speculation.’48
He points out that there is not one perception of blue, but many.
There is also an antithesis here between darkness and light: ‘Blue is
darkness made visible.’ Cirlot reminds us that colour symbolism derives
from several sources, and we should distinguish the mystical and esoteric
from those associations forged by primitive logic and intuition, on which
modern psychology would put more weight. The diurnal life of early
humankind was dictated by the cycle of night and day; and night was
when action had to cease. According to Lüscher, the dark blue of the
25
night sky is therefore the colour associated with quiescence and

passivity.49 He goes on to claim that prolonged contemplation of pure
dark blue has calming physiological effects on the central nervous
system, reducing blood pressure and lowering the rates of pulse and
respiration. Lüscher’s colour categories list the affective aspects of dark
blue as tranquility, contentment, tenderness, love and affection.
Although there is a high degree of subjectivity in making all these
associations between colour and human feelings, there seems to be a
broad measure of agreement emerging from different sources on the
significances of blue, especially when the depth of the colour is also
taken into account. The first great antithesis of colour, according to the
aesthetic theories of the painter, Wassily Kandinsky, lies between yellow
and blue, between warmth and coolness, between advance and retreat,
between overflowing the boundaries and contracting within them.
Concerning blue, he observed:
“Depth is found in blue...The deeper its tone, the more intense and
characteristic the effect. We feel a call to the infinite, a desire for
purity and transcendence. Blue is the typical heavenly colour; the
ultimate feeling it creates is one of rest. When it sinks almost to
black, it echoes a grief that is hardly human. It becomes an infinite
engrossment in solemn moods. As it grows lighter it becomes more
indifferent and affects us in a remote and neutral fashion, like a
high cerulean sky.” 50
1.8 Liberation from blue preciousness

Blue pigments for the artist were provided historically by six substances:
ultramarine, smalt, Egyptian blue, cobalt or Thénard's blue, azurite or
blue verditer, and indigo.51 Each of these suffered from some
disadvantage: ultramarine was very expensive - it had to be ground from
the precious mineral, lapis lazuli, which was mined only in Afghanistan;52
smalt, a cobalt blue glass used in ceramics and as a paper 'whitener',53 is
abrasive and rather weak in its covering power; the recipe for Egyptian
blue, §1.7, was lost in the 9th century and only recreated in 2007 on the
basis of a recipe left by 1st century BC Roman writer, Vitruvius; 54 cobalt
blue, first synthesised by Thénard in 1802, was a latecomer on the
scene;55 azurite, a basic copper carbonate mineral, is unstable to heat
and acids, turning into green malachite; and indigo, an organic dyestuff
obtained from a plant native to India, as mentioned in §1.6, is susceptible
to fading in light.56 We can therefore understand why the accidental
discovery In 1706 of the completely new pigment, Prussian blue (see
26
§2.1) by an artists’ colour-maker, heralded the liberation of painting

from its inhibitions concerning the expense of the colour blue.
The conservation scientist, Jens Bartoll, has shown by non-destructive
X-ray fluorescence spectrometry that Prussian blue is present in
paintings made at the Prussian court as early as 1709 (Fig. 1.12). Bartoll
has also demonstrated that by 1710 the pigment had evidently found its
way from Berlin to Paris, because it was also used in works by Antoine
Watteau and his circle.57 The ‘blue liberation’ of painting was completed
in 1821 by the further discovery of a recipe for synthesising ultramarine
itself.58 Thereafter, the history of painting is witness to a spectacular
influx of blue; for instance, in the work of the French Impressionists,
especially the landscapes and figure studies of Paul Cézanne, Fig. 1.12,
and in the powerful work of Picasso’s ‘blue period’.
Fig. 1.12 Pieter van der Werff, Entombment of Christ, 1709 (earliest
painting known with Prussian blue). Picture Gallery, Sanssouci, Potsdam.
The Expressionist school of Munich painters - the so-named Blaue

Reiter - is epitomised by Kandinsky’s ‘Blue Rider’, and the blue horses of
Franz Marc, Fig. 1.13, are later manifestations of their devotion to this
colour and its significance in the portrayal of the spiritual side of life.
27
Another turning point in the artistic significance of the colour blue

came when monochrome painter, Yves Klein, told French artists’
colourmaker, Edouard Adam, of his search for the perfect blue. Adam
faced Klein with a choice: bleu de Prusse ou bleu d'outremer :
“There was,” said Adam, “no contest, and you can see why. Prussian
Blue is sombre, good for shadows. Even in this windowless room,
ultramarine glows.” 59
But the real advance came when the marchand de couleur replaced the
traditional binder of rabbit glue, which dulled the vibrancy of the colour,
with a polyvinyl acetate resin called Rhodopas M, from manufacturing
chemists, Rhône-Poulenc. Mixed with ultramarine, Rhodopas allows the
powdered ultramarine to retain its granular look, its matt depth. In 1960,
Klein patented the new pigment under the number 63471 and the name,
International Klein Blue.
Fig. 1.13 Franz Marc, The Large Blue Horses, 1911
The colour continues to exert a fascination, as may be read in the

lyrical Rhapsody in Blue of Alexander Theroux, with its wide-ranging free
associations of the artistic, historical, linguistic and sociological
connotations of blue,60 and in John Gage’s art-historically oriented
appraisal of a recent exhibition of modern paintings on a blue theme.61
To return to the aesthetic difficulties with the cyanotype: John Wood,
in his essay, The Art of the Cyanotype, has been the first to try to
discover a reason for the sense of unnaturalness or incongruity that
appears to pervade this type of image.62 Wood puts forward two
28
antithetical interpretations of the colour blue, which have some affinities

with those I have cited above. On the one hand, he asserts that blue is
‘celebratory, affirmative and joyous’; on the other, that it has a ‘dark side’
which is ‘sad, depressed, and disturbing’.63 Wood is then able to propose
a dichotomy of a ‘joyous-ominous’ colour as the reason for the ‘peculiar
magic’ of the cyanotype, although no objective evidence is offered to
support these personal interpretations of the significances of the colour.
Perhaps this idea is also encapsulated in the German concept of
Schadenfreude – taking pleasure in the sorrow of others.
Blue has even been declared as the world’s favourite colour. In recent
surveys of randomly-chosen sample populations from several nations,
the colour blue has emerged at the top of every poll. These surveys are
part of the work of two post-modernist artists in the USA, Vitaly Komar
and Alexander Melamid. Their method is to use telephone opinion polls
to furnish sets of average aesthetic parameters for statistically ‘ideal’
paintings, according to the common taste of each national group. Their
objective is to paint for each public exactly what it wants, which turns out
in each case to be blue landscapes.64 Afficionados can join a Facebook
Group devoted to the evocation of blue in painting and poetry.65 Most
recently in 2015 a global survey of ten countries across four continents
has concluded that blue is the most popular colour, rating 23% to 33%: 8
to 18 % points ahead of any other colour.66
1.9 Colour language and coordinates

In describing the various processes of cyanotype and the resulting
images, it will be convenient to have a means of referring to colour in a
way that can be conveyed unambiguously to the reader.67 I propose to
adopt here a simple notation that indicates the approximate location of
the colour in a compact colour atlas, published as The Methuen
Handbook of Colour,68 where hue, tone and intensity are the three colour
coordinates. Referring to Fig. 1.14, the first number in this compilation,
running from 1 to 30, corresponds to a page denoting the hue. Each
double page spread presents six columns of colour patches, labelled A to
F, which contain an increasing admixture of black, shading the tone from
light towards dark. Within each column there are eight colour tablets,
numbered 1 to 8 to indicate the increasing intensity or saturation of the
colour, i.e. its relative brilliance or dullness. Some of the conventional
names for the more saturated colours of blue are represented on this
scale under the Figures 1.14 below.
29
Blue is the colour we humans perceive least efficiently, a fact that is

quantified in the expression for the contributions of red, green and blue
colours to the overall luminance of an image:
RGB Luminance = 0.3R + 0.59G + 0.11B
To conclude this chromatic introduction, it is my reluctant duty to
record an etymological irony that completely undermines the very name
of the cyanotype process and, in consequence, the title of this book. The
prefix ‘cyano-’ derives from the Greek root ‘kyaneos’ (kuaneoV), which is
usually translated as ‘dark blue’; and the suffix ‘-type’ comes from
‘tupos’ (tupoV) meaning ‘strike’, i.e. ‘print’. However, in the development
of various human languages, it was commonly the case that blue was the
last of the primary colours to receive a name; even today some primitive
languages lack any expression for this colour concept.69 Recent classical
scholarship points to the likelihood that, in ancient Greek, ‘kyaneos’
simply meant ‘dark’, and the language of Homer therefore had no word
for blue.70
19C7 Royal blue

19D8 International Klein blue (French patent 63471)
30
20A8 Ultramarine (artificial)

20E8 Lapis Lazuli
21C8 Ultramarine (genuine)

21F7 Prussian blue (also called Berlin, Milori or Paris blue)
31
22B7 Cobalt blue
23A7 Cyan (azure blue), 23C7 Cerulean blue, 23D7 Sapphire blue
Fig. 1.14 Blue Excerpts from the Methuen Book of Colour
32
2 The Invention of Cyanotype

“… sie macht sich nur durch Blut und Eisen.”
(“…It can only be carried out through blood and iron.”)
Prince Otto Edward Leopold von Bismarck (1815-1898)
Prussian House of Deputies, 28 January 1886. 71
Cyanotype images are composed of the intensely-coloured substance

Prussian blue, as it is generally known - except in the German-speaking
world, where its preferred and politically-correct name is Berliner Blau
after the city of its discovery. This celebrated pigment deserves to be
placed in its historical context because, fully a century before
photography was invented, Prussian blue was already an important colour
for artists,72 and in that ensuing century it proved also to be a key
substance for science, by opening up an entirely new area of chemistry.73
Experiments on Prussian blue led to the discovery of a whole new class of
compound – the cyanides - which hover tantalisingly on the borderline
dividing organic and inorganic matter, and which were therefore central
to the debate surrounding the nineteenth-century doctrine of vitalism.74
The political connotations of the name, Prussian blue,75 are purely
accidental which, in view of Bismarck’s statement in the epigram above,
makes even more remarkable the coincidence that Prussian blue was
indeed prepared traditionally from the raw materials of blood and iron.
2.1 A brief history of Prussian blue

Prussian blue has never been found in nature. There was no knowledge of
it before the eighteenth century,76 but this substance still qualifies as one
of the earliest synthetic pigments,77 dating from its first preparation,
probably in 1706, by a stroke of serendipity that befell an artists’ colour-
maker in Berlin, a Swiss immigrant called Johann Jacob Diesbach.78 The
chance occurrence of such a spectacular hue would have instantly
attracted the attention of any colour-maker, for at the time there was a
keen awareness of the limitations or expense of all other blue pigments,
as described in §1.8.
Diesbach identified the chemical source of his adventitious blue
colour, and the profitable manufacture of Prussian blue by Berlin chemist
Johann Frisch then began in 1708.79 Diesbach announced its merits as a
pigment in 1710;80 but, no doubt in the pursuit of gain, he succeeded in
concealing the method of its preparation until 1724, when it was finally
33
disclosed in a publication by John Woodward.81 Accounts of the original

discovery therefore tend to be conjectural; the earliest, by G E Stahl, did
not appear until 1731,82 but one version has it that Diesbach was wanting
to prepare a crimson lake by adding potash to a solution of cochineal,
alum and iron(II) sulphate. Lacking any potash, he purchased some from
a disreputable alchemist called Johann Konrad Dippel,83 fig. 2.1, who had
set up a laboratory in the same building. Dippel supplied Diesbach with
an impure sample of potash that happened to be contaminated with a
malodorous distillate of animal residues known as ‘Dippel’s oil’,84 which
provided the essential element - nitrogen - for Diesbach’s serendipitous
chemistry. Thus, the first preparation of Prussian blue, like several other
notable chemical discoveries, can ultimately be attributed to the duplicity
of an unscrupulous alchemist.85
Fig 2.1 Johann Konrad Dippel
There was no comprehension whatsoever of this reaction chemistry at

the time, because nitrogen had yet to be identified as an element. It
turned out that any nitrogenous animal matter could be used as a source
for Prussian blue: hide, hair, feathers, horn, hooves or flesh. Dried ox
blood, especially, proved a popular starting material. The recipes of the
day were varied and complex, often entailing superfluous ingredients,
and all were without doubt strikingly obnoxious in their manipulation, of
which the following is a short and colourful example:
34
“Six pounds of clippings of leather, six pounds of hoofs and horns,

and ten pounds of common potash, are boiled together in an iron
pot to dryness; the residue is then mixed with two pounds of crude
tartar, and, by means of a strong fire, is brought into fusion. The
lixiviation is conducted in the usual way, and a solution of five
pounds of sulphate of iron, and fifteen of alum being added, a
precipitate takes place, which is Prussian blue.” 86
Prussian blue rapidly came to the notice of chemists. Pierre-Joseph

Macquer, author of the first chemical dictionary, was attracted to the
substance by his interests as government supervisor of the French dyeing
industry, and he showed that it could indeed provide a fast dye for silk
and wool.87 His investigation in 1752 of the chief weakness of the
pigment, the destruction of its colour by alkalies, had an important
accidental outcome in the discovery of that useful salt ‘yellow prussiate
of potash’ (potassium ferrocyanide),88 prepared from the reaction of
Prussian blue with potash (potassium carbonate).89
Macquer and other chemists of the time also recognised that a
volatile, colourless, inflammable substance, which they named ‘the
colouring principle’, could be driven out of Prussian blue by heat, leaving
a residue that contained iron. A result of great significance to chemical
theory was later provided in 1782 by Scheele’s preparation of a solution
of this ‘colouring principle’ from the action of sulphuric acid on Prussian
blue, and his re-naming it as ‘prussic acid’. It is an unwitting testimony
to Scheele’s manipulative skills, that he achieved this preparation without
also making the uncomfortable discovery that prussic acid is a speedy,
lethal poison.90 Pure prussic acid was finally isolated in 1811 by Gay-
Lussac, who determined its composition to be a simple compound of
hydrogen, carbon and nitrogen,91 formula HCN, later to be called
hydrocyanic acid.92 The German name for this colourless substance was
Blausäure – ‘blue acid’ – in acknowledgement of its source. At this time,
Gay-Lussac also coined the name that has become the common currency
in this area of chemistry, a name that forms part of the title of this book.
He called the -CN radical cyanogène, using the Greek roots
kuanoV cennao (kyanos gennao) for ‘blue-generate’.93 This reference to its
colourful historical origins established the name for a whole class of
important new compounds - the cyanides – which very often are,
ironically, quite colourless.
So, a century after its discovery, Prussian blue, which is actually one of
the most complex of these cyanides (ferric ferrocyanide), had furnished a
route for isolating the simplest members of this new class of substance,
35
whose name derives from its colour.94 Thereafter, a rational basis for the
chemistry of cyanides began to emerge, and in 1814 Gay-Lussac isolated
the parent of them all, the gas cyanogen, having the simple formula
(CN)2. With the understanding that the fundamental unit in all this
chemistry was just a simple radical consisting of carbon bonded to
nitrogen,95 the possibility was recognised of synthesising cyanides from
purely inorganic materials, thereby making the use of organic animal
residues unnecessary - most mercifully, in view of the ‘foetid vapours’ to
which they gave rise. As a mark of approval for this achievement, the
Society of Arts in 1837 conferred a medal upon Lewis Thompson for
introducing a method of making potassium ferrocyanide by roasting
charcoal with potash and iron filings at red heat, exposed to the
atmosphere which provided the nitrogen.96 Prussian blue could then be
prepared from this product by reacting it with inorganic salts of iron.
By the 1780's Prussian blue was being widely manufactured in several
places in Europe, including in Glasgow by the firm of Turnbull & Ramsay,
Chemists,97 consequently the pigment also came to be known as
Turnbull's blue. Another Scottish user was raincoat inventor, Charles
Macintosh, who established a manufactury in the alum-works at
Lennoxtown in 1808 for dyeing wool, silk, and cotton with Prussian
blue.98 By the beginning of the nineteenth century, Prussian blue had
become a popular artists’ pigment; it was also then widely used in
European papermaking as an alternative to indigo for the colouring agent
in the popular blue-toned papers,99 and it was a constituent of that
everyday blue commodity, Stephen’s Ink. In the USA it was used as a
laundry whitener – ‘Mrs Stewart’s Blue’. The Newcastle firm of T. Bramwell
& Co., were awarded a Council Medal at the 1851 Great Exhibition - the
highest award – for their production, among other chemicals, of Prussian
blue, which was exported to China for colouring green tea.100 In a curious
later inversion of this use, it is now green tea that is used for toning
cyanotypes – see §8.3.
2.2 Light-sensitivity of iron salts

The cyanotype process is founded on a chemical reaction brought about
by light acting on certain salts of iron. The notion that light can promote
chemical transformations is, of course, central to the invention of
photography, and its history has been described at length by Josef Maria
Eder, and others.101 We shall not concern ourselves here with the long-
established sensitivity of silver halides to light, which provides the
36
mainstream of the technical history of photography, but rather confine

our attention to the less widely-known phenomenon of photosensitivity
in iron compounds, which is of more recent discovery. The first such
observation, albeit a rather indirect one, is attributed to Count Alexey
Bestuscheff (1693-1768), the Lord High Chancellor of Russia.102 This
worthy is said to have devised, in 1725, a nostrum called Tinctura tonico-
nervina, of a secret formulation, but reputed to contain gold -
presumably in order to justify its exhorbitant price.103 Bestuscheff
claimed that the agency of sunlight was employed in its production – a
further alchemical reference to gold. When the secret of the tincture’s
composition was finally revealed by the Empress Catherine the Great, who
purchased it from Bestuscheff’s heirs, it turned out to be no more
precious than a solution of ferric chloride in alcohol, which had been
decolourised by exposure to sunlight.104
Fig. 2.1 Count Alexey Petrovich Bestuzhev-Ryumin

(Collection of the Hermitage Museum)
Despite the unmasking of its commonplace nature, the tonic

continued to enjoy a long-lived commercial success as ‘Bestuscheff’s
Nervine Tincture’, although we do not know whether it was the iron, or
the alcohol, that conferred the greater benefit on the patient. A very
similar secret preparation was also popular in France during the early 18th
37
century, known as the ‘Golden Drops of General Lamotte’,105 which was

introduced to the French court under the title of Elixir d’Or and earned
the favour of Louis XV, who consequently granted the enterprising
General a substantial yearly pension to manufacture his ‘drops’, for the
Pope, among others.106 As the elixir made his fortune, it may be said to
have lived up to the promise of its name.
As late as 1853, the tincture can still be found listed in the
pharmacopoeias as Spiritus Sulphurico Etherus Martiatus, or as
Ferruginated Sulphuric Ether,107 having undergone in 1782 a modification
due to Martin Klaproth, who replaced the ethyl alcohol with diethyl ether
- making an altogether more heady mixture - but the tincture was still
said to be ‘blended in light’. In 1813, a scientific investigation of the
light-induced bleaching of ferric chloride in ether solution by Henri
August Vogel, found it to be very sensitive.108 These early observations
have no probable bearing on the actual invention of cyanotype, but they
may be responsible for some of the more misleading and unhistorical
statements that have been made concerning the process, such as:
“The principles involved in sensitizing blueprint paper were
discovered in France in about 1725 and perfected by Sir John
Herschel in England in 1842.” 109
Eder’s History of Photography suggests that the first observation of

the formation of Prussian blue by the action of light was in 1783, due to
an Italian professor of chemistry, Giovanni Antonio Scopoli, who exposed
a solution of potassium ferrocyanide and acetic acid to sunlight,
observing the precipitation of Prussian blue within about fifteen
minutes.110 There is another little-known, pre-photographic observation
of Prussian blue formation, made by John Mercer (1791-1866), a
Mancunian calico-dyer and printer, who became a self-taught colour
chemist and finished as a Fellow of the Royal Society. In 1828 he noted
the formation of Prussian blue on cotton by a light-induced reaction. The
biography of Mercer by his nephew, Edward Parnell, quotes Mercer’s own
description from his experimental notebooks:
“I spotted a piece of white cloth with a solution of pernitrate of
iron, and exposed it to the sun. On testing it afterwards with a
solution of red prussiate of potash it gave a blue - but no blue
before exposure. This is worthy of note. I have not seen it noticed
by any chemical writer.” 111
The pressure of maintaining his business deprived Mercer of the

opportunity to pursue this pioneering observation any further until the
38
1840s, by which time photography had become well-established at the

hands of other, more leisured, investigators. The chance to be counted
among the ‘pre-inventors’ of photography thus passed him by, owing to
the unfortunate necessity to earn his living - a disadvantage not suffered
by financially independent Victorian gentlemen-scientists, such as Talbot
and Herschel. Mercer later made significant inventions of toning
procedures for cyanotypes, as we shall see in §4.4.2 and §8.3, but these
were only published in an insubstantial form, and subsequently re-
invented by others. If we are seeking to identify an unsung hero of early
photography, a worthy candidate might well be John Mercer.112
The light-induced decomposition of an organic salt of iron was first
recorded in 1831 by Johann Wolfgang Döbereiner (1780-1849), who
observed that irradiation of an aqueous solution of ferric oxalate by
sunlight forms a precipitate of ferrous oxalate, which was shown to be
identical with the mineral Humboldtite, and evolves a gas, which proved
to be carbon dioxide.113 As we shall see in §3.5, the published report of
Döbereiner’s finding acted as a spur to Herschel in expediting the
publication of his own current photochemical investigations of platinum
salts, but no connection was made at this time between the light-
sensitivity of ferric oxalate and the well-known properties of Prussian
blue. Such a connection could have hinted at the formation of the
pigment as an alternative means for making images photographically.
2.3 The first silver photographs on paper

In 1834, William Henry Fox Talbot, the inventor of photography on paper,
began his experiments with silver salts as light-sensitive imaging
materials. He evinced no particular interest in Prussian blue, although he
did once record in his notebooks its formation, chemically rather than
photochemically, on a sheet of paper treated with hydrochloric acid and
potassium ferrocyanide, owing to ferric iron impurity in the acid.114
In January of 1839, stimulated by Talbot’s announcement of his
invention of ‘Photogenic Drawing Paper’ based on silver chloride, Sir John
Frederick William Herschel (1792-1871) took up the study of
photographic phenomena. Within one week he had solved the problem of
silver fixation by means of ‘Hypo’ – sodium thiosulphate. But then, in
contrast to Talbot’s single-minded pursuit of the silver image, Herschel
soon began to widen his investigations in the search for other viable
photographic processes. The history of this quest, and the fascinating
story of the interaction betweeen these two gifted men and their peers,
39
has been unfolded with clarity and scholarship by Dr Larry Schaaf in his
beautiful volume, Out of the Shadows, which provides a complete
historical background to this episode, and is essential reading for anyone
with an interest in the history of photography.115 I will therefore confine
this account chiefly to the discovery of the cyanotype process, and to
those aspects of it which have not already been described in detail.
2.4 Herschel’s research records

The primary sources for this account are Herschel’s handwritten
memoranda, experimental notes and test prints, and his published
scientific papers that sprang from them. For many years Herschel
recorded his chemical experiments in a disciplined, linear chronology,
using bound notebooks, now held in the Library of the Science Museum,
London.116 Four of these cover the years 1815-1870, but there is a
lacuna in the third volume. Between two successive experiment numbers,
recorded and dated on the same page, there is an interval of four years,
from August 1841 to January 1845. For this period, which saw some
profound changes in his working life, Herschel adopted a different
approach to his record-keeping habits. After August 1841, they became
multistranded, employing a new format of unbound loose leaves, kept in
separate envelopes titled by topic. This was a convenient way of coping
with the diversity of different parallel lines of photochemical research
that he was following concurrently – although it is possible, as pointed
out by Larry Schaaf, that Herschel always intended to ‘write them up’ in
his organised notebooks, but other concerns supervened.
Herschel’s loose-leaf records were sold at auction by Sotheby’s on 4
March 1958, and little thought of at the time, but the collector and
photohistorian, Helmut Gernsheim, was shrewd enough to acquire them
for his collection. The material was subsequently donated to the Harry
Ransom Humanities Research Center of the University of Texas at Austin,
which also has Gernsheim’s copy of the original Sotheby’s auction
catalogue with his autographic annotation against the Herschel lot:
“This comprises a lot of scraps which other people would consign
to the waste paper basket. The laboratory notebook contains no
interesting data, thoughts or descriptions merely dry factual one
line statements of the obscure substances tried; only the prints
prepared with ferroprussiate of potash still show their image, those
prepared with others are a blank.” 117
40
These original loose-leaf documents covering the crucial period of 1841-

3 are held in the Herschel Collection of the manuscript archive of the
HRHRC, where the present author had the pleasure and privilege of
spending a month as Mellon Research Fellow in 1999.118 For brevity,
these documents will be referred to henceforth as the ‘Memoranda’. The
HRHRC also has 43 of Herschel’s specimen prints, fig. 2.2, which he
consigned to the Royal Society to accompany the manuscript of his 1842
paper (§2.8).
The rest of Herschel’s experimental prints and tests described in the
Memoranda are still on the UK side of the Atlantic, divided between the
Museum of the History of Science at Oxford, the National Museum of
Photography, Film & Television (now the National Media Museum) at
Bradford, and the Library of the Royal Society of London.
Fig. 2.2 Sheet of Herschel’s specimen prints 1842 in cyanotype, etc

(Collection of the Humanities Research Center, University of Texas)
By February of 1840, Herschel’s Memoranda reveal that he had carried

out nearly seven hundred photochemical experiments, using silver salts
almost exclusively. There is a chronological list of ‘prepared papers’ in
the Memoranda, in which each preparation of chemicals coated on paper
that Herschel tested for light-sensitive behaviour, was allocated a
sequential number. It will be convenient to refer to this as his ‘prepared
paper number’. Most of these coated sheets were divided up into smaller
pieces, and thereby sufficed for several different tests.
41
2.5 Anthotype and phytotype

In his quest for more colourful images, Herschel soon took leave of the
brown pictures provided by silver, and in 1839 began experimenting with
juices from the flowers of his own garden, such as red stock or poppy. He
crushed their petals in alcohol, extracting the dyestuffs (anthocyanins)
into solutions to be painted on paper.119 These dyed coatings would
slowly bleach in sunlight, so exposure in contact with printed engravings
directly yielded positive images in plant pigments which Herschel hoped
would supply the primaries for a full colour printing system. He called
this process anthotype (Greek: ανθος = 'flower') or sometimes phytotype
(Greek: φυτον = 'plant'). But the printing was very slow, demanding
protracted contact exposures to sunlight lasting for days or even weeks.
Worse, the images proved ephemeral: there was – and still is – no means
of 'fixation', because subsequent exposure to strong light inexorably
bleaches them. It is remarkable that some of Herschel’s specimens still
survive today.120 Impermanence notwithstanding, the anthotype process
was taken further by that gifted Scotswoman, Mary Somerville (née
Fairfax,
Untitled
1780-1872),121 and there is still a lively constituency of07/03/2014 12:16
practitioners today in this environmentally-friendly print medium.122
Fig. 2.3 Sir John Herschel, anthotype from an engraving

The spring weather of 1840 was remarkably brilliant, to the enduring

benefit of the embryonic science of photography. Herschel had already
42
initiated a new series of exposure tests of ‘vegetable colours’, but this

work was interrupted in March by the relocation of the family home from
Slough in Buckinghamshire to Hawkhurst in Kent. Once resettled at
Collingwood, Herschel resumed his experiments in August, but by then
he frequently found the sun to be ‘pale’ or ‘desultory’, requiring long
exposures for these very insensitive processes.123 He pursued them
nonetheless, during the very poor summer of 1841 that followed.124 It
seems plain from the nature of Herschel’s investigation of the bleaching
of these coloured plant juices by light, that he was in pursuit of a direct
positive-working colour photographic process, and he stated as much in
one of his letters to Talbot,125 and in a letter to the 1841 Meeting of the
British Association for the Advancement of Science, reporting on these
experiments:
“ ... a circumstance which ... holds out no slight hope of a solution
of the problem of a photographic representation of natural objects
in their proper colours.” 126
The most important conclusion that Herschel was able to draw from his
observations was that a dye tends to be selectively bleached by the light
of its complementary colour, which was further evidence for the
Grotthuss law of photochemical absorption (§3.6).
2.6 Discovery of cyanotype

In the early spring of 1842, Herschel discontinued his tests of plant dyes
which were proving rather unsatisfactory, in favour of further broadening
his search for new photosensitive substances for his ‘non argentine
photography’. His attention transferred from these rather evanescent
organic dyes (now appropriately called anthocyanins), to deeply-coloured
inorganic compounds. He was later to explain this shift of interest in a
paper to the Royal Society, in the following chemically picturesque terms:
“The general instability of organic combinations might lead us to
expect the occurrence of numerous and remarkable cases of this
affection among bodies of that class, but among metallic and other
elements inorganically arranged, instances enough have already
appeared, and more are daily presenting themselves, to justify its
extension to all cases in which chemical elements may be supposed
combined with a certain degree of laxity, and so to speak in a state
of tottering equilibrium.” 127
His awareness of these ‘lax’ inorganic compounds had been informed
by earlier consultations with Dr Alfred Smee (1818-1877), Fig. 2. The
important rôle played by this talented young physician and scientist in
43
enabling the discoveries that were to follow was first brought to our
attention by Larry Schaaf.128
Smee was well-grounded in chemistry, and his interests at the time lay
in the exciting new sub-science of electrochemistry, which was then
enjoying its golden era, following the invention in 1800 of the electric
battery by Alessandro Volta (1745-1827), and the thrilling use made of it
in 1807 by Sir Humphry Davy (1778-1829), who harnessed its electric
power to “tear their elements asunder” and isolate the amazingly reactive
alkali metals, potassium and sodium.
By 1840, the Experimental Researches on Electricity by Michael
Faraday (1791-1867), Davy’s assistant and successor at the Royal
Institution, had established the quantitative basis for this new branch of
science. The design of powerful batteries had been improving steadily at
the hands of Wollaston, Daniell, and Grove; and Smee himself had
introduced a further improvement in the construction of a large battery
for his own electrochemical researches.129 The electrodeposition of
metals, then called electrotyping, was becoming a craft industry,
producing such curios as the copper-plated cucumber prepared by Smee,
which was shown to Queen Victoria - whose enquiring finger promptly
poked a hole in it. Like photography, electrotyping could enable the
duplication of art works - though from moulds rather than negatives - so
it was referred to at the time as ‘the sister art of heliography’.130
Fig. 2.4 Dr. Alfred Smee
Smee also employed his battery as a new means for preparing

chemicals that had previously been rather inaccessible; one such
substance was then variously referred to as the ‘ferrosesquicyanuret of
potassium’ or ‘red ferrocyanate of potash’ or ‘red prussiate of potash’,131
44
but we can now identify it by the modern name of potassium

ferricyanide.132 Herschel was aware of Smee’s achievement, since the
latter had read his paper before the Royal Society on 18 June 1840, and
had subsequently published it in the Philosophical Magazine.133 The chief
outcome of this research was to make this interesting, bright-red,
soluble, crystalline chemical much more readily available in a pure form,
because previous methods of preparation, dating from the substance’s
discovery by Leopold Gmelin in 1822,134 had usually resulted in a product
that was contaminated by impurities.
In his 1840 paper, Smee described how he could prepare potassium
ferricyanide from the commonplace and widely-available potassium
ferrocyanide.135 He stated that this new technique of electrolytic
oxidation could also be applied in principle to other analogous inorganic
substances - a possibility that must have stimulated Herschel’s original
enquiry of him for ‘deeply-coloured salts’.
Fig. 2.5 Oxidation of potassium ferrocyanide to ferricyanide
Shortly after his appointment as surgeon to the Bank of England, early

in 1842, Smee sent Herschel a specimen of potassium ferricyanide. On 23
April 1842, Herschel made the following entry in his Memoranda:
“Photography. Photochromy.
Non Argentine, Mineral substances (F3/2CP) [136]
Smee’s Red Ferrocyanate of Potash washed on paper gives it a fine
pale green colour.
April 23. 1842.
The spectrum thrown on this paper acts slowly but about as fast as
on Guiacum … When the paper is thrown onto water the impression
becomes stronger, loses its Violet ruddiness & turns to a fine
prussian blue … a wash of very dilute acid … immediately
45
developed a strong blue impression, having the above character.

This paper will prove valuable.
Try other metallo-cyanates of Bases.” 137
This is the first recorded observation of Prussian blue being formed

for a photographic purpose by the action of light on potassium
ferricyanide, so it represents the moment of discovery of the first
cyanotype process, although this name (Greek: kuaneoV = 'deep blue') was
not coined by Herschel until four months later.138 The significance of this
observation impressed Herschel sufficiently to include it also in his
general diary entry for 23 April 1842:
“Discovered the Photographic property of the Red Ferro
sesquicyanuret of Potassm.” 139
‘Discovery’ certainly seems a more appropriate word than ‘invention’
to describe this moment of photographic innovation.140 But the next day
Herschel was already seeking ways to improve this non-silver
photographic process, and he began making images by its means, as is
evident from the following entry in the Memoranda:
“Photography. Non Argentine. Sesqui Ferrocyanate of Potash. F3/2CP
April 24/42. Various modes of improving colour & increasing
action tried.
1. Pure water destroys the Dove colr first & washes away the
yellowish pink border. It also darkens the impression to a good
blue
2. Dilute Sulc Acid developes a much stronger blue.
3. Weak sulphate of soda washes the superfluous salt well away
with no injury to the blue wh is quite insoluble in neutral salts.
4. Very weak perchloride of Iron acidulated with S. Acid, develops a
deep & most superb blue. But the ground also is somewhat blued.
Tried it for copying engravings It does them very beautifully, but
the copies are negative in the lights are blue the shades white.
They are best fixed by very weak sulphuric acid - a barely
perceptible acidulation - If Chlor Iron be added the ground gets
blued.” 141 [Underlining in original]
Can we detect in these words a hint of surprised disappointment that this

process had proved to be negative-working?
46
Fig. 2.6 Sir John Herschel, cyanotype (1842) from an engraving by H.

Cook: Still in my Teens, published in Friendship’s Offering 1838.
Although it would, therefore, be well-suited to making positive prints

from camera negatives, Herschel’s preferred modus operandi at the time
was to make photographic copies of ‘positive’ images, usually
engravings, by means of positive-working processes such as the
bleaching of dyes. The pictures resulting from a negative-working
process would require a second printing - or ‘re-transfer’ as he called it -
to restore their correct tonality and handedness, and he realised that this
would entail a loss of definition. Herschel would soon devote
considerable effort to devising a positive-working version of the Prussian
blue process, in the attempt to circumvent what he saw as this
disadvantage. But before that became possible, there was one other
chemical ingredient, previously unknown to him, that was destined to be
introduced into the process, with the consequence of greatly enhancing
the sensitivity to light and the speed of printing in Prussian blue.
2.7 Invention of siderotype

The ‘list of prepared papers’ in the Memoranda shows that Herschel
received a further sample of potassium ferricyanide from Smee on 27
April,142 and promptly employed it to make more Prussian blue ‘copies’
from engravings, one of which he sent to Smee. This brought an
acknowledgement from Smee dated 10 May 1842, which reveals that he
47
was already well aware that Prussian blue would result from the action of
light on potassium ferricyanide, especially when mixed with a ferric salt:
“Allow me to return you my thanks for your beautiful and
interesting photograph which I presume was executed with the red
salt in conjunction with a persalt of iron as that compound
invariably deposits Prussian Blue under the combined action of air
and light.
I trust that by variety of the manipulation you will be enabled to
adopt it to the camera.” 143
(italics in original)
Smee was no stranger to the newly-invented technology of

photography; as early as May 1839, only a few months after Talbot’s first
disclosure of his achievement, Smee published an account of ‘Photogenic
Drawing’ in the Literary Gazette, in which his observations on the best
proportions for the sensitizing chemicals and the choice of the paper
were both perceptive and useful.144 In view of this prior photographic
experience, and his awareness of the ferricyanide photochemistry implied
in the quotation above, we might speculate why Smee had not employed
potassium ferricyanide to make photograms at least; it appears that he
very narrowly missed the distinction of himself becoming the inventor of
the cyanotype process. It would, however, transpire that his hope it might
be useful in the camera was unfounded. Had matters rested here, it is
unlikely that Herschel’s discovery of the ‘proto-cyanotype’ process would
have developed into an important invention - the commercially-viable
blueprinting process - because potassium ferricyanide, by itself, is of
very low sensitivity, and demands inconveniently lengthy exposures. But
Smee’s letter of 10 May went on to suggest the key chemical that was to
unlock an entire photographic treasure-trove for Herschel:
“There are two salts which of late have been used in Medicine
having been vamped up by the Chemists and Druggists. The
Ammonio Citrate and Ammonio Tartrate of Iron which are perfectly
soluble and give very dark solutions. I mention them thinking it just
barely possible that they may not have found their way into your
laboratory and should my anticipation be correct it will afford me
much pleasure to send some of each.” 145
This deferential tone is what we might expect from the 23-year-old

Smee, a newly-elected Fellow of the Royal Society, addressing the 50-
year-old Herschel, already one of the most distinguished elder statesmen
of British science. Presumably, Smee did not realise that he was placing in
Herschel’s hands the key to the best method for cyanotype and several
48
other processes besides, but simply thought that the ‘dark solutions’
might lend themselves to bleaching by light. Although they were not
difficult to prepare, the ‘Ammonio Citrate and Ammonio Tartrate of Iron’
were quite new substances on the chemical scene. Interest in them had
been stimulated by the well-appreciated pharmaceutical benefits of iron
tonics, dating back to Bestuscheff’s ‘tincture’ (§2.2).
Fig. 2.7 Ammonium ferric citrate – varieties
It is evident from his reply to Smee, that Herschel had had no prior
knowledge of these salts, which is not surprising since they do not
feature in the pharmacopoeias until after 1841.146 Then they were
promoted by druggists as ‘chalybeate remedies’147 and, indeed, they are
still valued as iron tonics today.148
Finding the chemical ‘ammonio citrate of iron’ already on the
pharmacy shelves, Herschel soon followed Smee’s recommendation to try
it - photographically, rather than medicinally149 - and immediately
discovered that it was highly sensitive to light. He successfully sun-
printed with it in a variety of ways, but his tests of the substances that he
abbreviated as ‘ACI’ and ‘ATI’ are of uncertain date, because his
Memoranda relating to these experiments appear to have been written-
up retrospectively. This was an uncharacteristic lapse for such an
impeccable keeper of scientific records. Is it possible that the excitement
of the discovery was so great as to distract him from a lifetime’s
discipline of systematic note-keeping? Herschel’s letters to Smee are
lost,150 but fortunately he made an approximate copy of his reply,
‘Substance of letter to Smee’ on 15 June 1842, which clearly displays his
excitement and pleasure at the richness and variety of processes that his
discovery of light-sensitivity in ‘ACI’ had made possible:
49
“I cannot help thanking you for your mention in one of your late
notes to me of the Ammonio citrate and Ammonio tartrate of Iron
as highly coloured salts. The former of these salts I have procured
and examined, and it has furnished me with an infinity of beautiful
photographic processes, both in conjunction with your
Ferrosesquicyanate and with other ingredients. I take the liberty to
enclose you a specimen or two, to appreciate which they should be
wetted, laid on white paper and examined with a pretty strong
magnifier. This process if you should happen to attend next
Thursday’s meeting of the R.S. you will hear described under a
name I have invented to give it in imitation of Mr. Talbot’s Calotype
- viz. - Chrysotype - from the use of Gold as a stimulant to bring
about the dormant picture which the action of light produces on
paper prepared with this salt. It is one of the most striking and
magical effects which has yet turned up in photography, the
Calotype and Argyrotype (Daguerreotype) themselves not
excepted.” 151 [Underlining in original]
Thanks to Smee, Herschel had been able to devise a faster cyanotype

recipe, which has come down essentially unchanged to the present day.
Moreover, his astute chemical reasoning quickly suggested ways of
making images in noble metals, such as silver, gold, and mercury using
the same iron salt.152 Thus in 1842, just three years after the first public
announcements of silver photography, came Herschel's novel processes
based on light-sensitive salts of iron, see Table 2.1. His collective name
for these was siderotype, (Greek: sidhroV = 'iron'; tupoV = 'wrought').153
Year Inventor Process name Image Substance

1842 Herschel Cyanotype Prussian blue
1842 Herschel Argentotype Silver

1842 Herschel Chrysotype154 Gold
1842 Herschel Kelainotype Mercury
1859 Poitevin Ferrogallate Iron-gall ink
1873 Willis Platinotype155 Platinum
1877 Pellet Pellet print Prussian blue
1889 Nicol Kallitype156 Silver
1889 Arndt Van Dyke Silver
1917 Willis Palladiotype Palladium
Table 2.1. Early Siderotype processes
The chrysotype process which so enthused Herschel, although
strikingly coloured, never took its promised place in the photographic
50
repertoire; far from rivalling Talbot’s Calotype in general practice, as

Herschel had hoped, it was destined to be dismissed by all the major
photographic authorities of the nineteenth century as impractical,
unsatisfactory and obsolete.157 However, Herschel’s original conviction
concerning the excellence of printing in gold has finally been vindicated
by the writer in recent years.158
The light-sensitivity of all the siderotype processes proved to be little
greater than Talbot's first silver process of 'photogenic drawing', so they
were not seriously entertained as useful means for securing images in the
camera, once the much faster calotype process discovered by Talbot in
1840 was in general use. Of all the sun's light scattered by a
photographic subject, only a small fraction (ca. one hundredth to one
thousandth) is captured by a camera lens and projected onto the
photographically sensitized surface. It follows from the law of reciprocity
that direct exposure of the same sensitized surface to the incident rays of
the sun will take only about one hundredth to one thousandth of the
camera exposure time, to provide a clear outline of any object placed
upon it. Such 'photograms' are therefore much easier to make by direct
contact than are camera photographs.159
2.8 Publication of siderotype

Herschel submitted details of his new photographic discoveries in a
second, long and important paper to the Royal Society,160 accompanied
by 43 specimen sun-prints in his several processes, all made from
engravings, see §2.4.161 His paper was entitled ‘On the Action of the Rays
of the Solar Spectrum on Vegetable Colours, and on some new
Photographic Processes’ and was accepted on the 15th of June 1842 for
publication in the Philosophical Transactions of the Royal Society. Part of
this paper was read by Herschel before the Society on the 16th of June,
but it did not appear in print until September; it will be convenient to
refer to this seminal work as ‘the 1842 Paper’.162 It was here that
Herschel first publicly described the making of prints in Prussian blue
from potassium ferricyanide alone, as follows:
“202. …A beautiful example of such deoxidising action on a non-
argentine compound has lately occurred to me in the examination
of that interesting salt, the ferrosesquicyanuret of potassium,
described by Mr. Smee in the Philosophical magazine, No.109,
September 1840, and which he has shown how to manufacture in
abundance and purity by voltaic action on the common, or yellow
ferrocyanuret. …
51
203. Paper simply washed with a solution of this salt is highly

sensitive to the action of light. Prussian blue is deposited …. After
half an hour or an hour’s exposure to sunshine, a very beautiful
negative photograph is the result” 163
The relatively lengthy exposure needed here should be noted (indeed,

Herschel’s ‘half an hour or an hour’ tends towards an optimistic
underestimate); it contrasts with the very short exposures, of a few
minutes only, which are all that is required for a contact print when
ammonium ferric citrate is employed.
It can be inferred from the 1842 Paper that, by June, Herschel had also
made cyanotypes (but not yet named them thus) by exposing a mixture
of ammonium ferric citrate and potassium ferricyanide - the recipe that
has endured as the standard practice until the present day:
“206. If in lieu of the perchloride of iron, we substitute a solution of
that curious salt the ammonio-citrate of iron, the photographic
effects are among the most various and remarkable that have yet
offered themselves to our notice in this novel and fertile field of
inquiry. The two solutions[164] mix without causing any precipitate,
and produce a liquid of a brown colour, which washed over paper is
green (being strongly dichromatic). If this be done under the prism,
the action of the spectrum is almost instantaneous, and most
intense. A copious and richly coloured deposit of Prussian blue is
formed over the blue, violet and extra-spectral rays.” 165
[Italics in original]
Here the chronology of Herschel’s experimental Memoranda is once

again anomalous, in that they do not record such a test being made until
10 August 1842. This may indicate another lapse brought on by the
excitement of the work in hand, for there can be no doubt that Herschel
had already made the first ‘standard cyanotypes’ in June.166 By this time,
the serendipitous discovery had matured into a spirit of calculated
invention using the ferric processes. Herschel made a simple experiment
to show which of the two components was the more light-sensitive: by
deliberately isolating the ammonium ferric citrate and exposing it alone,
followed by development in a bath of potassium ferricyanide, Herschel
proved that by far the greater part of the light-sensitivity could be
attributed to the former iron salt, not the latter.
"210. It seems at first sight natural to refer these curious and
complex changes to the instability of the cyanic compounds, and
that this opinion is to a certain extent correct, is proved by the
photographic impressions described in Arts. 204 and 209, where
no iron is added beyond what exists in the ferrocyanic salts
themselves. Nevertheless the following experiments abundantly
52
prove that in several of the changes above described, the

immediate action of the solar rays is not exerted on these salts, but
on the iron contained in the ferruginous solution added to them,
which it deoxidizes or otherwise alters, thereby presenting it to the
ferrocyanic salts in such a form as to precipitate the acids in
combination with the peroxide or protoxide of iron, as the case
may be. To make this evident all that is necessary is simply to leave
out the ferrocyanate in the preparation of the paper, which thus
becomes reduced to a simple washing over with the ammonio-
citric solution." [Italics in original]
Herschel’s understanding of the photochemical basis for the

‘deoxidising’167 behaviour of ammonium ferric citrate enabled him to
make reasoned applications of this chemical knowledge in several
different ways, as exemplified by his chrysotype, argentotype and
kelainotype processes. Herschel's important observation of high
sensitivity to light in ammonium ferric citrate (his "ferruginous solution"
of Article 210) led to the invention of these new photographic processes
(Articles 212 and 226). The Memoranda reveal that Prussian blue printing
was totally eclipsed during the intervening month of July 1842 by
Herschel’s endeavours - again employing the marvellously versatile
ammonium ferric citrate - to make pictures in mercury and gold, whose
striking beauty captured his undivided attention. His account in the 1842
Paper reveals a glimpse of the seductive delights of mercury prints which
he called kelainotypes168:
“229. … affords pictures of such force and depth of colour, such
velvety richness of material, and such perfection of detail and
preservation of the relative intensities of the light, as infinitely to
surpass any photographic production I have yet seen, and which
indeed it seems impossible to go beyond.” 169
Lest the reader should feel tempted by this eulogy to cast this book
aside in favour of a search for mercury salts, it would be prudent to take
note of Herschel’s next sentence:
“Most unfortunately, they cannot be preserved.” 170
The following week in August, however, saw Herschel’s return to

printing in Prussian blue, with fresh endeavours to refine and perfect the
processes. His chemical logic suggested that a complementary, positive-
working system should be possible by employing, not the ferricyanide,
but the commonplace ferrocyanide of potassium, in conjunction with
ammonium ferric citrate. On 16 August 1842 he wrote the name
‘cyanotype’ in his Memoranda for the first time, to describe such a
53
positive-working process, but the results that it yielded, as he later

admitted in his paper of 1843, read to the Royal Society on 17th
November 1842, usually fell far short of giving satisfaction:
“232. The positive cyanotype process described in Arts. 219, 220
of my papers, though beautiful in its effect (especially during the
first few minutes of the appearance of the picture), is very
precarious in its ultimate success … this process, though
exceedingly easy in its manipulations, could not be recommended
as practically useful.” 171
[Italics in original]
By the end of August he had accumulated sufficient new results to

justify adding a substantial postscript to his 1842 Paper, which was still
awaiting publication by the Royal Society. It is in this postscript, dated
29th August 1842, that the word ‘cyanotype’ appears in print for the first
time, but by then Herschel expressed a preference for the name to be
understood as embracing all the processes, both positive- and negative-
working, which resulted in an image of Prussian blue:
“219. Cyanotype. - If a nomenclature of this kind be admitted (and
it has some recommendations), the whole class of processes in
which cyanogen in its combinations with iron performs a leading
part, and in which the resulting pictures are blue, may be
designated by this epithet.” 172
We can do no better than respect Herschel’s preference in our

nomenclature today (contrary to the view of some bibliographers who
take ‘cyanotype’ to mean only the positive-working process173) but the
epithet should be qualified when necessary to make distinction between
the different variations of cyanotype process. Herschel then goes on to
make an excusable overstatement:
“The varieties of cyanotype processes seem to be innumerable, …”
From the observations in the 1842 Paper together with the notes in
the Memoranda, we can distinguish no less than fifteen methods that
Herschel had devised for making images in Prussian blue. Some of these
will be considered more closely in §4.1.
Smee provides an illuminating postscript to this highly innovative
episode in a letter dated 29th September 1842, which was evidently his
reply to a letter from Herschel which is lost:
“The salts of iron to which you allude are neither children of mine
by birth nor by adoption and the only circumstance which induced
me to recommend them to your notice was their deep colour as
perhaps you may recollect that a long time since you asked for
54
specimens of any high coloured metallic salts not commonly met

with. I know nothing about the salts further and had not the
slightest idea that they could be used for Photographic purposes
and did not even know the purpose for which you were likely to
want dark coloured salts.” 174
The implication of this response is that Herschel, who was scrupulous
in acknowledging the priority of fellow scientists’ discoveries, must have
enquired if Smee was the discoverer of ammonium ferric citrate and
ammonium ferric tartrate, to determine if he should be accorded a
‘proprietorial acknowledgement’ or wished to make use of them
photographically himself. Smee’s disavowal of any priority has a self-
effacing tone, but one wonders if, perhaps, he knew more than he
admitted there. Whatever the case, Smee deserves great credit for
directing Herschel’s footsteps down such a fruitful path. Herschel’s
original search for “dark coloured salts” had been founded, of course, on
his hope that some of them might be susceptible to bleaching by light,
and thus provide direct positive processes. In the event, the successful
photochemistry proved to be quite the converse, but it is a mark of
Herschel’s great qualities as a scientist that he was able to turn these
unexpected serendipitous discoveries into good inventions.
2.9 Misapprehension of Herschel’s processes

Following the Royal Society’s publication of Herschel’s discoveries with
iron-based photography, several early handbooks on the newly-invented
photography republished their own accounts of his processes because
the ordinary reader did not have ready access to the Philosophical
Transactions. But, in the course of so doing, most of them managed to
misrepresent the relative significance of his different cyanotype methods
because, as we have seen, Herschel was not entirely explicit in his 1842
Paper. Thus, George Fisher’s important early manual Photogenic
Manipulation of 1843 makes no mention of the highly effective negative-
working process, but only describes the rather failure-prone positive-
working process (see §4.1.4):
“Brush the paper with a solution of the ammonio-citrate of iron.
This solution should be sufficiently strong to resemble sherry-wine
in colour. Expose the paper in the usual way, and pass over it very
sparingly and evenly, a wash of the common yellow ferro-cyanate
of potass. As soon as the liquid is applied, the negative picture
vanishes, and is replaced by a positive one…” 175
55
This same procedure is the only one described for cyanotype by W.H.
Thornthwaite in his Photographic Manipulation of 1843.176 Robert Hunt,
in the 5th edition of his Manual of Photography of 1857 likewise confines
his description of cyanotype to the positive-working process using
potassium ferrocyanide,177 whereas in the 3rd and 4th editions of 1853
and 1854 he had quoted all of Herschel’s writings in extenso. The
Practical Manual of Photography of 1845 by “a practical chemist and
photographist” does mention the negative process, but then goes on to
complicate it by the process of bleaching with mercury salts, as in the
curious ‘hydrargyrocyanotype’, §4.1.3.178 As late as 1852, in his widely-
referenced work, Photogenic Manipulation, Robert J. Bingham
foregrounds only the positive cyanotype process.179 The second edition in
1867 of the authoritative Dictionary of Photography by Thomas Sutton
and George Dawson only describes three of the more fallible cyanotype
processes: the very slow ‘proto-cyanotype’; the unreliable positive
process; and the so-called ‘hydrargyrocyanotype’ using a mercury salt
bleach.180 Significant publications from the USA erred in a similar way:
Henry Hunt Snelling’s well-known History and Practice of Photography of
1845 only mentions the positive process,181 and S.D. Humphrey’s System
of Photography of 1849 gives precedence to it over the other,182 while
John Towler’s Silver Sunbeam of 1864 surprisingly prescribes ferric
chloride as the sensitizer for negative-working cyanotype.183 These many
unfortunate instances of misdirected practical instruction may well be
responsible for the notable lack of take-up of Herschel’s best cyanotype
process – the neglected negative-working one - over the three decades
following his discovery.
Even that dignified periodical, The Athenæum, in presenting a
synopsis of Herschel’s paper, had erred by describing the process as one
of separate development, and went on to make confusion worse
confounded by conflating both his cyanotype and chrysotype
processes.184 This called forth a lengthy corrective letter from Herschel
himself a fortnight later,185 but it seems to little avail. Remarkably similar
misinterpretations of his 1842 paper persist even today: in 2007, the RPS
journal The Photohistorian published a re-examination of Herschel’s
experiments, which states that his preferred modus operandi for his
negative-working cyanotype process consisted in coating the paper with
ammonium ferric citrate solution alone, exposing it, then subsequently
‘developing’ the image with a solution of potassium ferricyanide.186
To support this claim for ‘separate development’, the last sentence only
56
of Herschel's text in article 210 (see §2.8 above) is selectively quoted

thus:
“…all that is necessary is simply to leave out the ferrocyanate in the
preparation of the paper, which thus becomes reduced to a simple
washing over with the ammonio-citric solution."
The author then goes on to allege that:
"This critical statement, which Herschel takes the trouble to
italicise, is the one either overlooked or dismissed by later
historians." 187
On the contrary, "later historians" have considered all of Herschel's

statements very carefully: while Article 210 does indeed report a rare
occasion when he did not pre-mix the two ingredients, ammonium ferric
citrate and potassium ferricyanide, Herschel's true intention becomes
plain when one reads the text that the selective quotation was careful to
omit - the beginning of Article 210, above. There he clearly states that he
intended this as an experiment to show which of the two components
contributes most to the photosensitivity of the mixture. He already knew
that potassium ferricyanide was light-sensitive by itself (Articles 203 and
204), and that its admixture with ammonium ferric citrate was very
sensitive (Articles 206 and 207). But by exposing ammonium ferric citrate
first, then "developing" the result with potassium ferricyanide, he
demonstrated that the former was by far the more light-sensitive of the
two. In fact, Herschel describes his favoured practical method elsewhere
in the same 1842 paper - in the Articles 206, 207 and 223:
“223. If paper be washed with a mixture of the solutions of
ammonio-citrate of iron and ferrosesquicyanate of potash, so as to
contain the two salts in about equal proportions, and being then
impressed with a picture, be thrown into water and dried, a
negative blue picture will be produced agreeably to what is stated
in Art. 154.”
The same formulation is given in his personal experimental Memoranda

noting his formulae for cyanotype 'prepared papers', nos. 780, 784, and
793 (MSs at HRHRC, University of Texas at Austin), which he coated two
months after writing the 1842 paper. From these sources it is plain that,
to make his negative-working cyanotype papers, Herschel always
preferred to pre-mix his solutions of ammonium ferric citrate and
potassium ferricyanide, before coating his paper with the mixture, and
exposing it. The economy of this procedure is obvious, and it results in a
better quality of print.
57
This issue has been further obfuscated by authors citing the fact that
Herschel adopted the procedure of separate exposure of ammonium
ferric citrate alone, followed by "development", in the cases of his
positive-working cyanotype process and his chrysotype process. The
former, which was never entirely successful, required development with
potassium ferrocyanide solution; the latter with neutralised gold chloride
(sodium tetrachloroaurate), which was very costly. In both these cases,
pre-mixing the two chemicals was ruled out by the speedy
decomposition of the mixed sensitizer before the paper could be coated
and exposed. These processes should not be conflated with the negative-
working cyanotype, which became the commercially successful blueprint
process, and the version of cyanotype still popular today.
A proper study of all Herschel's photographic writings - rather than a
single, misleading and selective quotation – brings one inevitably to the
unsurprising conclusion that the negative-working cyanotype procedure
which he personally preferred was the same as that used ever since by
almost everyone, including his own son Alexander. There is no case here
for attempting to re-write this episode of photographic history, but
regrettably this misrepresentation of what is now misleadingly being
called ‘the Herschel process’ is becoming a ‘folk myth’ embedded in
some of the recent popular beginners’ literature on cyanotype.188
2.10 An unaccountable omission

From our present vantage-point of chemical hindsight, we might, on the
other hand, take the uncharitable view that a man of Herschel’s
knowledge should have anticipated the photographic value of ‘salts of
iron with the vegetable acids’, like ammonium ferric citrate, long before
Smee pressed them upon him. Ten years earlier, while in Hamburg,
Herschel had become aware of researches just published by
Döbereiner,189 who had demonstrated in 1831 the decomposition by light
of ferric oxalate, a compound similar to the citrate. It is a remarkable
omission by Herschel in 1842, nowhere to have made any use of the
photosensitivity of this substance, nor ever to have referred to it, or to its
discoverer, Döbereiner.
Very much later, in 1864, on an occasion when Herschel was obliged
to defend his priority over the invention of cyanotype, he showed by his
retrospective comments on “…these ferruginous combinations with
vegetable acids”, that he was well-acquainted with the ferric oxalate
possibility, but contented himself with the dismissive observation that it
58
“…differs only from the processes described in §218 of the paper

above cited…by the substitution of the oxalic for the citric and
tartaric acids, which, from the habitudes of that acid in general,
might have been predicted.” 190
It is nevertheless very surprising that an avid experimentalist like

Herschel should not have tested the properties of ferric oxalate, of which
he must have been aware in 1842. Had he done so, he might have
devised a platinotype process thirty years earlier than its actual
development by William Willis in 1873.
Even if Herschel’s sense of propriety forbade him from taking up the
discovery of another scientist, we might at least have expected some
citation or comment from him concerning the relevance of Döbereiner’s
work on ferric oxalate. The author of the first manual on photographic
processes, Robert Hunt (1807-1887), was well-acquainted with this
photochemistry and was moved to comment tactfully in his 1844 edition
on Herschel’s omission:
“So extensive have been the researches of the distinguished
philosopher, whose labours I have so frequently quoted,
particularly into the action of the sun’s rays on the salts of iron,
that little can be added to his published information. It may not,
however, be uninteresting to add a few brief remarks on some of
the salts of iron, to which Sir John Herschel has not extended his
observations, or at least which have not been recommended by him
as photographic agents.” 191
Hunt offers no explanation for the lacuna in Herschel’s research, but

goes on to describe his own experiments with ferric oxalate, which was to
prove a very popular photosensitive salt, destined to be used successfully
by several photographic innovators in the years to come (see §4.4.4).192
A possible explanation for Herschel’s total silence on this subject may
be that his highly-developed code of scientific ethics had been offended
by earlier actions on the part of Döbereiner. In the ‘pre-photographic’
year of 1831, Herschel had discovered that a particular platinum salt
solution was sensitive to light, and formed a white precipitate of
‘platinate of lime’ when illuminated.193 Upon hearing about Döbereiner’s
parallel photochemical researches on iron salts, Herschel hastened to
publish his platinum work in order to establish his priority in this area of
photochemistry, and to announce his intention of continuing with an
analysis to identify the product.194 Herschel’s result was quickly
confirmed in Germany by repeating the experiment, and acknowledged
with a declaration that there was no intention of further trespassing on
59
Herschel’s field.195 Despite this disavowal, Döbereiner subsequently

initiated and supervised a research project into the analysis of Herschel’s
‘platinate of lime’, in order to determine a formula for the photoproduct,
which Döbereiner published in 1835, during Herschel’s sojourn in the
Cape of Good Hope.196 While it was not in Herschel’s nature to harbour
resentment, this action may have left him with a natural feeling of
antipathy that he expressed only by completely shunning the work of his
rival in photochemistry.
Without a close examination of Herschel’s experimental records in the
Memoranda, no-one can fully appreciate the multiplicity of intertwining
threads of investigation that he was following simultaneously in 1842-3.
Multitudes of domestic plants had been pressed into service for their
coloured juices, and even dog’s urine and an ‘extract’ from a boa-
constrictor! Salts of platinum, iridium, gold, mercury, iron, lead, silver
and chromium all fell within the orbit of his experiments. Many of these
led him up blind alleys; for instance, three pages of his cyanotype
Memoranda, which describe experiments following up the apparently
promising results of adding ferric iodide to the cyanotype mixture, are
struck through, endorsed with the later revocation in Herschel’s hand:
“Hunting on a wrong scent”. Fruitless attempts to render permanent
those exquisite kelainotype prints in mercury also accounted for a large
portion of his efforts at this time, and in a letter to Talbot we can hear a
hint of exasperation in his complaint about their evanescence:
“It has led me such a dance as I never before was led by any
physical enquiry and I have not yet succeeded in reducing it to a
definite and certainly successful process…” 197
Herschel directed much more of his endeavour towards perfecting a

positive-working cyanotype process than he did to the negative-working
counterpart. The latter had simply fallen into his hands with a minimum
of experimentation, but the former was always his preferred objective,
probably because he saw it as providing a blue primary for a direct colour
photographic process;198 yet it proved so variable in quality and so
difficult to fix, that he reached the verdict that it “could not be
recommended as practically useful”. In an attempt to inhibit the
spreading of the image substance, Herschel tried adding gum Arabic to
his potassium ferrocyanide developer. Later, in order to achieve fixation
without degrading the image, he enlisted the dangerous poison ‘corrosive
sublimate’ (mercuric chloride), but both toxicity and cost militate against
60
this as a practical process. It is remarkable that one or two surprisingly

fine specimens of Herschel’s positive cyanotype still survive from 1842,
especially in the collection of the National Media Museum (previously the
National Museum of Photography Film & Television) at Bradford.
It seems appropriate to end this tale of discovery and invention with a
sobering reflection on the fickle character of Herschel’s scientific muse.
By the autumn of 1842, he had performed more than a thousand
photographic experiments, most of which proved ultimately fruitless;
Herschel’s greatest and most enduring success, the negative-working
cyanotype process, was actually accomplished with just three test-
papers.199
2.11 Pursuit of photographic permanence

Following Anna Atkins’ monumental achievement to be described in §5.4,
and the more modest endeavours of her peers during the 1850s, the
cyanotype process fell into desuetude for about twenty years. During
these two decades, photographic technology made enormous strides, but
one of the most hotly debated issues was the impermanence of some
images. Silver had revealed its chemical vulnerability early on, to the
chagrin of Talbot, Henneman, and others who printed extensively in the
medium. In 1855 the Royal Photographic Society set up its aptly-named
‘Fading Committee’ to investigate the deterioration of silver images.200
The Committee’s recommendation of gold toning as a partial remedy did
much to extend the longevity of silver photographs made thereafter,201
but this measure was not universally successful in its application. The
‘Achilles’ heel’ of silver images lies in their susceptibility to sulphiding.
The necessity for using sodium thiosulphate as the fixing agent (another
of Herschel’s discoveries) in order to dissolve and remove the excess
silver halide, could, if the thiosulphate was not fully washed out of the
paper, lead to the eventual transformation of the silver image into silver
sulphide. This greatly weakened the density of colour, especially in the
early salted-paper process, which contained relatively low coating
weights of silver.202
Alternatives to printing in silver were therefore keenly sought, and
photographic innovation was spurred on in 1856 by the Duc de Luynes’
offer of substantial prizes for the invention of methods for making
permanent prints. The Duke’s gold medal was awarded in 1859 to
Alphonse Poitevin for his process of print-making by the photo-
hardening of pigmented colloids such as gelatin.203 Here the maker is
61
free to choose the most permanent pigments for the substance of the
image, such as those known to painters for a millennium. One of these,
lampblack, bestowed its identity on this carbon process, as it was called.
Following further refinement by Swan in 1864, it is still practised today.
The process is labour-intensive and demands considerable skill in
manipulating and transferring the pigmented gelatin tissues, but it still
provides the only means for making truly permanent photographs in
colour.
An alternative route to permanent monochrome prints lies in seeking
for the image a metal more ‘noble’ - that is to say, more chemically
resistant to attack - than silver. Gold and platinum share the metallic
summit of the hierarchy of elemental chemical stability. Gold was tried
with some success by Herschel, as has already been described, but his
beautiful chrysotype process of 1842 proved too expensive and
uncontrollable for general practice.204 Platinum printing remained
intractable chemically until 1873, when William Willis began to work
towards his goal of a viable platinotype process. With further refinement
over the next twenty years, this was to become the premier medium for
‘fine-art photography’. By 1900, the walls of the photographic salons
exhibited more platinotypes than any other process.205
2.12 Misappropriation of cyanotype

It is still surprising today to see what little application the cyanotype
process found in the thirty years following its invention; in explanation it
has been argued in §2.9 that Herschel’s achievement was not well served
by his bibliographers, most of whom only reported his unsatisfactory
attempts at a positive-working process. In view of the enormous
commercial importance of the negative-working medium later on for
reprography, it is interesting to note that Herschel himself had
demonstrated its usefulness for photocopying text. Herschel’s literary
work included a translation in 1867 of a Schiller poem from German into
Latin, the manuscript of which he copied in cyanotype, fig. 2.8.
Herschel’s son, Alexander, also used the process for copying his
astronomical notes. The only other persons to have made significant use
of the process over three decades appear to have been Anna Atkins and
her circle of lady amateur botanists, as will be described in §5.4.
62
Fig. 2.8 Sir John Herschel, cyanotype copy of manuscript

(courtesy of Dr. Larry J. Schaaf)
63
Despite their historic, but relatively private endeavours, the process

became largely forgotten. Perhaps for this reason, we witness more than
one instance of its spontaneous re-invention by others. In 1864, The
Photographic News carried a report of “A process for obtaining proofs in
Turnbull’s blue … by Professor Schwarz, of Breslau”,206 who thought that
calico printers might avail themselves of it as a dyestuff, evidently
unaware that Mercer had already done so two decades before. This
brought forth a letter from Herschel himself in the following issue of The
Photographic News, which the editor entitled “Sir J F W Herschel’s
Cyanotype Process. - A Reclamation”. Herschel firmly pointed out that
Schwarz’s process was essentially identical with the one described in his
1842 paper. Notwithstanding such sporadic revivals, and the endeavours
of a few experimentalists such as Mercer and Carey Lea (see §4.4), there
followed a period of disuse so prolonged that the readers of Herschel’s
obituary in the British Journal of Photography of 19 May 1871 had to be
reminded of these photographic achievements:
“In the course of his experiments he discovered several elegant
processes possessing much interest, which, although lying dormant
to a great extent at present, may yet be resuscitated...” 207
And resuscitated they were - but in a new guise: out of the rich legacy
of unpatented processes left by Herschel, the cyanotype was seized upon
commercially by Marion and Company of Paris. This company launched
the so-called ‘Marion’s Ferro-prussiate Paper’ on the market in 1872.208
No doubt for proprietory business reasons they chose to re-brand the
cyanotype, and the publicity heralding the appearance of this product
made no hint of an acknowledgement to Herschel, nor intimated that
there might have been any ‘prior art’.209 Unsurprisingly, no patent was
granted, nor apparently even applied for; instead, the formula was kept a
secret which appears never to have been disclosed, so it is still
impossible to tell if “the ferro-prussiate method of M. Marion” actually
contributed anything by way of an improvement on Herschel’s original
formula. This commercial usurpation of an invention of thirty years
standing, one that had originally been given freely to the world by its
creator, excited indignation at the time in at least one commentator, the
editor of The Photographic News. His remarks in the issue of 24 August
1877 merit quotation as a polemic which is still full of relevance today for
those who practice ‘re-inventing’ the processes due to others:
“The Cyanotype Process.
It is not unusual, as every photographer knows, for a process to
64
remain perdu for some years, and then to get made known again
with some mystery and circumstance. We see old friends revived in
this way from time to time, and, like somebody whom we meet for
the first time in a quaintly shaped hat or a different-coloured coat
than usual, we fail to recognise him at the first moment of our
rencontre. But it is only for an instant that the deception holds
good, and the well known features are as apparent as ever as soon
as we get on close terms again. The cyanotype process, as Sir John
Herschel used to call it, is the latest attempt at “fraudulent
enlistment,” and the circumstance that we have here to do with a
genuine article, and not a spurious one, does not make the
deception the less barefaced. The cyanotype process as practiced
by Sir John Herschel is a very simple affair, and one that is found
extremely useful for copying outline sketches or multiplying
tracings, instead of using tracing-paper. Sir John Herschel
employed it, as does Professor Herschel, of Newcastle, to the
present day, for making copies of their calculations of astronomical
tables, the characters or figures being produced in white upon a
blue ground. The paper, ready prepared for use by draughtsmen or
others, is to be purchased in Paris; and also, we believe, in this
country, of Messrs. Marion, of Soho Square; but those who desire
to make it for themselves can easily do so. Good, smooth paper is
treated with
Citrate of iron (or ammonio-citrate) … 140 grains
Red prussiate of potash …………………120 grains
Dissolved together in two fluid ounces of water.” 210
Having thus exposed Marion’s ‘secret’, the article then went on to

describe its use. This little polemic is actually of great value in providing
us with one of the earliest specifications of Herschel’s formula, other than
those in his own Memoranda, and its implications will be discussed in
§4.1 on the varieties of the process.
The advent of commercial blueprint paper, marketed by Marion and
others, enabled the widespread use of the process for expedient copying
of engineering drawings. In the USA, for instance, students of mechanical
engineering at Massachusetts Institute of Technology were instructed in
the use of the process as early as 1875, and they caused its introduction
into many drawing offices. The first commercial blueprint machine, made
by a Swiss company, was introduced into the USA at the Philadelphia
Centennial Exposition of 1876. The American teachers of the process
only acknowledged that it had been ‘originally introduced from France by
Mr A L Holley’.211 The name of Herschel, by then, had been totally
eclipsed.
65
Fig 2.9 Marion’s advertisement of 1885 for Ferro-prussiate paper212
66
3 The Molecular Basis of Blueprinting

“Stone walls do not a prison make,
Nor iron bars a cage”
Richard Lovelace (1618-1657)
The processes of cyanotype show a surprising range of behaviour that

cannot be understood without some appreciation of the underlying
chemistry. This need not make deeply technical demands on the reader
nor require a formal scientific background - the sterner approach is
confined to Appendices II and III, which are intended only for those
readers who are comfortably fluent in the language and symbols of
chemistry. In this short chapter, I shall try to provide an account that is
accessible to the majority of nonspecialist readers, who may have only a
slight acquaintance with chemical science, but are prepared to endure a
little of its jargon - in return for one or two anecdotes.
3.1 Chemistry of Prussian blue

The chemical identity of Prussian blue is ferric ferrocyanide.213 Among
the many compounds of iron known to chemistry, Prussian blue is quite
unusual in containing the element combined in two different states of
oxidation: both as ‘ferric’ iron and ‘ferrous’ iron.214 Intense colour is a
quintessential property of this mixed state of oxidation, and it is seen in
a number of other such iron compounds.215 The deep-blue colour is
directly attributable to the absorption of energy, in the form of red light,
by electrons ‘hopping’ easily from the ferrous to the ferric iron.216 The
removal of red from the spectrum of white light leaves us with a
sensation of the complementary colour, blue, generally called ‘cyan’ in
modern colour language.
The simple experiment of preparing Prussian blue in a test-tube is
visually astonishing, and has long been the basis for tricks of parlour
magic and ‘sympathetic inks’ for secret writing. Two dilute water
solutions are required, each of them unremarkable in appearance: one, of
a ferric salt (it matters not whether it is the sulphate, nitrate, or chloride)
is pale yellow; the other, of potassium ferrocyanide, is colourless. On
mixing the two solutions, an inky liquid of the deepest blue is
instantantly formed, from which a dark blue precipitate slowly separates:
ferric cation + ferrocyanide anion ® ferric ferrocyanide
67
Performance of this dramatic reaction was part of the stock-in-trade

for the public lecture-demonstrations of celebrated nineteenth-century
chemists such as Michael Faraday. The wonder of carrying out this
experiment has been beautifully captured in a fine, but anonymous
nineteenth-century painting in the collection of Alfred Bader,217 who
identifies the demonstrator of the reaction as William Thomas Brande
(1788-1866), while a youthful Michael Faraday looks on. Bader points out
that, in the copy of Brande’s seminal Manual of Chemistry once owned by
Faraday, there are extensive annotations in Faraday’s hand on the subject
of Prussian blue, which evidently engaged his attention for some years.218
Fig 3.1 Thomas Phillips, Preparation of Prussian blue

(Collection of Alfred Bader)
The compounds of the Prussian blue family all have a similar structure
in which there are two different types of environment for the iron atoms;
one of them is distinguished by cyanide groups strongly-bound through
their carbon atoms. Since there are also two possible oxidation states for
the iron in each type of environment, we therefore have a total of four
possibilities when combining ferric or ferrous salts with a ferricyanide or
a ferrocyanide, and the properties of the resulting substances are
68
summarised in Table 3.1.219 Note how the other three compounds are all
readily transformed into Prussian blue, which is by far the most stable
product of the four.
Reactant Ferricyanides Ferrocyanides

Ferric Ferric ferricyanide Ferric ferrocyanide
salts Prussian yellow (or Berlin Prussian blue (or Berlin blue)
brown) Soluble; a powerful Highly insoluble; most
oxidant, easily oxidises water, intensely coloured, and most
and paper etc., being reduced stable of all four possible
via green intermediates (Berlin products in this table, to
green) to Prussian blue which the others revert
Ferrous Ferrous ferricyanide Ferrous ferrocyanide
salts A highly unstable substance Prussian white (also Everitt’s
that instantly reverts by salt220 or Williamson’s salt221
electron transfer to ferric or Berlin white) Insoluble and
ferrocyanide, Prussian blue colourless, but readily
which, formed this way, is oxidised by air (and other
often called Turnbull’s blue oxidants) to Prussian blue
after its manufacturer
Table 3.1. Varieties of complex iron cyanides
Because negative-working cyanotypes are usually formed by the

reaction of a ferrous salt (resulting from the decomposition of ferric salts
by light) with a ferricyanide, a common error in all older published
accounts of the cyanotype process is to describe the resulting pigment,
Turnbull’s blue, as ferrous ferricyanide, and this misconception even
persists among some antagonists today.222 Modern chemical research of
the last 30 years has conclusively shown this to be wrong (see Appendix
II.4); when formed, ferrous ferricyanide rearranges ‘instantly’ by
transferring an electron between its two types of iron to give ferric
ferrocyanide, Prussian blue:
ferrous cation + ferricyanide anion ® [ferrous ferricyanide -unstable]
® ferric ferrocyanide (Prussian blue)
However, the early experimentalists may be forgiven for thinking that
Turnbull’s blue was different from Prussian blue, because the appearance
of its hue often suggests so, for reasons to be described below.
69
The colourless substance, ferrous ferrocyanide, or Prussian white, will

be important to our considerations later, because it is formed directly in
one variant of commercial blueprint paper. Prussian white is also the
product of the fading of Prussian blue by light; its formation accounts for
the tonal reversal that the cyanotype image undergoes upon extended
exposure, where the paling of the shadow areas is called ‘solarisation’ by
photographers. However, it should be understood that this fading
transformation is not brought about solely by light; the chemical
reduction of Prussian blue to Prussian white must be accompanied by a
corresponding oxidation of something else:
light + Prussian blue + reductant ® Prussian white + oxidation products
but what reductant exactly is oxidised is an open, and very important
question, which will be addressed later. Prussian white is unstable in air
and is slowly oxidized by oxygen back to Prussian blue:
Prussian white + oxygen + water ® Prussian blue + hydroxide ions
If this were all that there is to say about the chemistry of Prussian
blue, my task would have been simple. But nature at the molecular level
is seldom unsophisticated, and sometimes ‘imperfect’. In fact, there exist
several ‘Prussian blues’, all of them in essence ferric ferrocyanide, but
each having an individual composition depending on its precise method
of preparation.223 The analytical complexity and intractability of this
‘substance’ has engaged - and frustrated - chemists for no less than two
centuries,224 but only in the late 20th century did a clear understanding
begin to emerge, thanks to the powerful instrumental techniques of
modern chemical research. The structural details are set out in Appendix
II, but for a general discussion can be summarised as follows.
The formulae for Prussian Blue are usually stated to range from an
‘insoluble’ Prussian blue,225 to the so-called ‘soluble’ Prussian blue,226
which also contains potassium ions.227 The latter is a serious misnomer:
in fact, all forms of Prussian blue are highly insoluble in water. The
apparent ‘solubility’ in the latter case is an illusion created by the ease
with which this solid may be dispersed in water as minute particles,
called nanoparticles, which form a blue suspension giving the appearance
of a true solution. Chemists call this process peptization,228 and the
resulting suspension of the nanoparticles or colloid is known as a sol
which is fine enough to pass through a filter paper.229 The ease of
70
peptizing ‘soluble’ Prussian blue is responsible for some of the problems

that beset the making and conservation of cyanotypes.
The Prussian blue pigment of commerce, which is manufactured
globally on a scale of 50,000 tonnes per year, is now prepared via
Prussian white by a method which ensures that a substance is obtained in
which ammonium ions replace potassium.230 This variety has a brighter
colour and is more resistant to peptization and destruction by alkalies.231
3.2 Molecular structure of Prussian blue

Many of the remarkable properties and present-day uses of Prussian blue
can be understood in terms of the three-dimensional structure of the
solid at the atomic level – its “molecular architecture”. This structure is so
simple and (ideally) regular, that visualizing it should present no difficulty
even to non-chemists. Imagine a cubic scaffolding: iron atoms are
located at the joints, alternately ferric and ferrous, the cyanide groups (–
C≡N) form the rods that connect them, with their carbon atoms adjacent
to the ferrous ions, and nitrogen atoms to the ferric ions. In one
direction, the structure runs like this:
| | | | | |
–N≡C–Fe–C≡N–Fe–N≡C–Fe–C≡N–Fe–N≡C–Fe–C≡N–Fe–
| | | | | |
and likewise in two other mutually perpendicular directions. For a
representation of the three-dimensional structure, see Figure 3.1. Note
how spacious this arrangement is: not close-packed as are many simple
crystalline solids.Throughout this open lattice we can see tunnels running
in three perpendicular directions, and a large void within each cube,
which can accommodate molecules of water, for example, or metal ions,
and which allow Prussian blue to act as a kind of chemical host. In the
case of ‘soluble’ Prussian blue, half of these spaces are occupied by
potassium ions. The incorporation of varying amounts of water and other
ions into the voids in this lattice accounts for some of the variations
observed in the formulae, and the puzzlement of early analytical
chemists.
The other cause of variability in the composition of Prussian blue is
the presence of defects in the lattice; that is, departures from
symmetrical regularity due to the absence of a proportion of the
ferrocyanide groups. The formula of the ‘insoluble’ variety is explained
by a lattice that has one quarter of its ferrocyanide groups missing, and
71
water molecules in their place. In the language of the crystallographer,

the lattice is said to have ‘systematic vacancies’.
Fig. 3.2 The crystal lattice of Prussian blue
The formula recently found for Turnbull’s blue implies that in its
lattice one third of the ferrocyanide groups are missing, and replaced by
water molecules and anions. This is nominally, at least, the form of
Prussian blue formed in the cyanotype process; it will be seen later that
the possibility of a defect lattice which can act as host to anions is of
great significance to its light-stability.
3.3 Prussian blue for ion exchange

The open structure of the Prussian blue lattice gives it a porous quality
that explains why the substance is a very useful ‘chemical sponge’ for
soaking up and trapping cations similar to potassium, such as those of
the metals caesium and thallium.232 Prussian blue itself is non-toxic,
unreactive and tasteless, and can be safely eaten in gram quantities
without harmful effects,233 and its insolubility ensures it is excreted
72
easily, making it a valuable antidote to poisoning by these heavy metals,

from which it has been used successfully to detoxify both animals and
humans.
As an example of the ability of the microporous Prussian blue lattice
to act as host for relatively large amounts of impurity ions, it was put to
good use for ‘locking up’ some of the radioactive fallout, in the form of
the radioisotope caesium-137, that was deposited on the uplands of
North Wales, Cumbria and Scotland following the disastrous Chernobyl
reactor explosion of April 1986.234
Tests were conducted to see if this contamination could be countered

by spreading Prussian blue on the contaminated soil of the pastures, thus
inhibiting the uptake of caesium by growing grass which, in turn, would
prevent its ingestion by the animals grazing the slopes, but it was found
more effective to feed the Prussian blue directly to the affected sheep,
and thus avoid turning the green hills of Britain to navy blue!235 The
effects of Chernobyl lingered on. In 1992, it was discovered that migrant
herds of reindeer in Norway had also absorbed caesium-137 from this
disaster, because they graze on lichens and are particularly partial to
mushrooms, which both concentrate the caesium. The livelihood of their
Sámi herders was in jeopardy,236 so Prussian blue was added to the ‘salt
licks’ that these nomadic animals are so fond of, with beneficial effects
for the migrant herds.237 Prussian blue continues to be used to detoxify
73
the livestock in the Chernobyl–contaminated regions of Belarus, Russia

and Ukraine, and has saved their agricultural industries of milk and meat
production from enormous losses.
In Germany, appropriately, the benefits of feeding Prussian blue to
radioactively-contaminated pigs has also been amply demonstrated in a
recent scientific study.238 Caesium-137 is also used in radiotherapy to
treat certain cancers, so there are substantial quantities of this
radioisotope in specialist medical facilities worldwide. Acknowledging the
risk that these stocks could be abused by terrorists to make a
radiological “dirty bomb”, the Federal Drug Administration in the USA has
recently licensed the use of Prussian blue as a safe antidote to Cs137
contamination in humans.239
Poisoning of humans by salts of the scarce heavy-metal, thallium
(which resembles lead), is fortunately not frequent, but In the mid-20th
Century, it enjoyed a spell of notoriety as a subtle toxin: its unusual
symptoms could be mistaken for diseases such as encephalitis or
epilepsy, but incidentally the victim’s hair dropped out, so sudden
alopecia offered a clue.240 (Its use as a depilatory, however, declined in
the 1930s.) This highly toxic metal is a favourite of poisoners who
happen to be both chemically erudite and homicidally inclined; their hope
is that its unfamiliar toxicology may help them evade detection.
Thallium(I) sulphate is colourless, odourless, tasteless, and sufficiently
water-soluble to be administered easily to an unsuspecting victim. The
symptoms do not appear for several days, but a dose of a gram or so is
usually fatal, because the body cannot get rid of the element, which
mimics potassium but blocks its vital enzymic action, especially in the
brain.
Thallium was used quite extensively by the secret police of the late
Saddam Hussein’s regime in Iraq for disposing of senior political
dissidents. A useful trick was to administer the dose just before the
victim was allowed to leave the country, so he died abroad. The
substance was readily available in the Middle East, where it was
commonly employed to destroy rats and cockroaches, so accidental
poisoning was also common. Fortunately – if the cause of the poisoning
is correctly diagnosed – there is an antidote which has saved many lives:
Prussian blue.241 The thallium(I) cation can enter the Prussian blue lattice
in place of potassium. There it becomes trapped, and can be safely
excreted by the body. Prussian blue was therefore put forward by German
74
pharmacologist, Horst Heydlauf of Karlsruhe, as the preferred antidote

for thallium poisoning, and its value has since been amply confirmed.242
These two instances of the ability of Prussian blue to act as a ‘prison’
for chemical poisons are not just irrelevant digressions to the present
work. Later in §8, we shall consider methods of toning the cyanotype
image, among which are processes using salts of lead and thallium. Lead
resembles thallium quite closely, and its ability to cause a shift in the
colour of Prussian blue may probably be attributed to its incorporation
within the lattice, as just described. As we shall see in §9 on the
vulnerability of cyanotypes and in Appendix III, the fading of Prussian
blue in the light may also be greatly promoted by the presence of
impurities trapped within it.
3.4 Prussian blue for electrochromic displays

Over the last decade, Prussian blue has attracted an enormous amount of
research interest from electrochemists because it can be commercially
exploited as an electrochromic display; that is, a solid-state device whose
colour can be changed by varying the electric potential applied to it. As
the voltage across a Prussian white electrode is raised, and negatively-
charged electrons are abstracted from it, the colour passes through a
spectacular, reversible sequence: white, blue, green, and finally yellow:
ferrous ferrocyanide → ferric ferrocyanide → ferric ferricyanide
white blue (green) yellow
a beautiful visual demonstration of the increasing oxidation states of the
two types of iron atom that are summarised in Table 3.1.243
3.5 Chemical history of siderotype

The cyanotype process is just one member of a whole class of alternative
photographic printing processes which all share a common type of light-
sensitive component: the ferric salts of certain organic substances, which
were appropriately called ‘vegetable acids’ by Herschel, owing to their
widespread natural occurrence within the plant kingdom. However, the
discovery of this light-sensitivity in organic iron salts considerably pre-
dates its photographic uses, as we have seen in §2.2.
The action of light on ferric salts of some organic acids produces
ferrous salts by promoting an internal oxidation-reduction reaction.244 All
the iron imaging systems have this same basis; the key compound is
usually a ferric salt of citric, tartaric, or oxalic acid.245 For simplicity in
75
our brief excursion into the chemical formulae, we will take the last-
3+
named as our example: ferric oxalate, which contains iron(III), Fe , and
2-
oxalate, C2O4 , chemically bound together. Under the influence of ultra-
violet light, this salt undergoes the reaction first observed by Döbereiner
in 1831, as described in §2.2. Iron(III) reacts with oxalate ions to form
iron(II) and carbon dioxide gas - a process which is expressed
symbolically by the following chemical equation (the only equation, I
promise, that readers will encounter, unless they venture into the depths
of Appendices II and III):
UV light + 2Fe3+ + C2O42- ® 2Fe2+ + 2CO2
UV light + ferric ions + oxalate ions ® ferrous ions + carbon dioxide
Note how two negative charges (which are carried by electrons) have
been transferred from oxalate to two ferric ions, reducing them to ferrous
ions.
This light-induced change occurs in the dry, solid ferric oxalate, and
in related compounds such as ammonium ferric oxalate and ammonium
ferric citrate; but it is only accompanied by a slight colour change from
pale yellow-green to pale yellow-brown; moreover this change is
temporary, because the iron(II) is slowly re-oxidised back to iron(III) by
oxygen in the air. In order to make a permanent photograph, the iron(II)
which is formed by the action of light must be immediately reacted to
produce a more visible and enduring substance for the image.
Chemically, iron(II) is a reducing agent, because it readily gives up an
electron and reverts to iron(III); so one option is to use it to reduce the
compounds of a ‘noble’ metal, such as platinum, palladium, silver or
gold, to the metallic state:
ferrous ions + platinum salts ® ferric ions + platinum black
Reactions of this type provide the basis for the well-known
platinotype, palladiotype, argyrotype and chrysotype processes.246
3.6 Photochemical principles

Photochemical theory had its beginnings in 1818 with the fairly self-
evident Grotthuss-Draper Law: only light absorbed by a substance
can induce a chemical change within it. Light is energy. Across the
visible spectrum, going from red to blue, the intrinsic energy of light
nearly doubles. When a molecule absorbs visible light, which is one of the
causes of colour, its internal energy is increased, and it is said to be lifted
76
into an ‘excited’ state. Sooner or later, this excess energy is transformed

and lost, so that the molecule returns to its normal or ‘ground’ state - a
process that is appropriately called ‘relaxation’. The manner of the
transformation of the absorbed energy, however, can be various, and it
depends on the lifetime of the excited state and other factors. The least
interesting, and most common process is ‘thermal relaxation’, that is the
dissipation of the excess energy as heat - embodied in more agitated
motion of the molecules. In some spectacular instances, however, part of
the excess energy may be re-emitted as light, giving us the attractive and
useful luminescent phenomena, called fluorescence and
phosphorescence, depending on whether the excited state is short- or
long-lived, respectively. Most importantly for our purposes, the excess
energy may be directed to bring about a chemical reaction if the
molecular circumstances are favourable. It should come as no surprise
that a molecule in an excited state is capable of more vigorous chemical
reactions than when it is in its ground state. Ultra-violet light is often
required to supply sufficient energy for such chemical transformations.
There is an equivalence between the quantity of light energy absorbed
and the amount of substance that is chemically transformed, both of
which are measurable, and may be related by an elementary principle of
quantum theory, as it applies to photochemical changes.247 This
principle, which constitutes the second law of photochemistry, dates from
ca 1912 and is usually referred to as the Stark-Einstein Law.248 It is easily
appreciated without invoking much jargon, and the non-technical reader
should readily see that it amounts to nothing more than a statement of
‘good book-keeping’ at the molecular level. The Stark-Einstein Law
effectively requires that: for each molecule of the substance to be
transformed, one photon must be absorbed.249 (A photon is the
fundamental, indivisible particle or ‘quantum’ of light.)
From this basis, a calculation may be performed assuming the most
favourable circumstances (see Appendix III), which yields an estimate for
the maximum light sensitivity of any ‘proto-photographic’ material - no
matter what its detailed photochemistry.250 This limiting sensitivity is a
maximum in the near ultraviolet and blue regions of the spectrum, over a
wavelength range of about 300-400 nm (known as UVA). Unsurprisingly,
the sensitivity turns out to be very low, requiring a radiant energy density
of about 34 J/m2 (Joules per square meter) to form a just-perceptible
image. On a scale familiar to photographers, the ‘speed’ of such a
material can be rated approximately at 10-5 ISO (arithmetic);251 that is,
77
one ten-millionth of the speed of an ordinary camera film of 100 ISO

(arithmetic).
According to the Bunsen-Roscoe Law of Reciprocity – the third law
of photochemistry - photographic exposure is defined as the product
of light intensity and time. We may write our minimum ‘proto-
photographic’ exposure as 34 W s/m2 (Watt-seconds per square
meter),252 in order to calculate the approximate duration of exposure to
sunlight. The tropical sun at its zenith delivers an energy flux at the
Earth’s surface with an intensity of about 900 W/m2 (Watts per square
meter), but only about 7% of this total is ‘actinic’ light – UVA and blue -
able to bring about photochemical change, amounting to ~63 W/m2,
maximum. Taking account of the average angle of elevation of the sun,
the average UVA irradiance will be about 30 to 40 W/m2, comparable with
an exposure for one second to form a just-perceptible image. Since the
exposure scale for a proto-photographic process such as cyanotype is 7
to 8 ‘stops’ between ‘just perceptible’ and maximum density (a factor of
128 to 256), we see that the exposure needed to fully darken the
unobscured sensitized paper should be in the order of 2 to 4 minutes, for
a ‘proto-photographic’ process in average sunlight. This theoretically
predicted exposure is in agreement with all experience.
3.7 Chemistry of blueprinting

In the negative-working cyanotype process, the iron(II) photo-product is
reacted with ferricyanide to yield Prussian (alias Turnbull’s) blue:
ferrous ions + ferricyanide ions ® Prussian blue
This reaction entails a switch in the oxidation states of the two types
of iron, as indicated in Table 3.1: the intermediate ferrous ferricyanide is
highly unstable and transforms very rapidly to ferric ferrocyanide.
If the light-sensitive iron(III) salt is present in excess, and the
exposure is prolonged sufficiently, the photochemistry proceeds one
stage further: some of the Prussian blue (ferric ferrocyanide) is further
reduced to Prussian white (ferrous ferrocyanide), and the deep blue
image begins to pale in its regions of greatest exposure.
light + Prussian blue + ferrous ions ® Prussian white + ferric ions
This tonal reversal is the phenomenon of ‘solarisation’ first observed
and named by Herschel. It is actually desirable for making a print-out
image, because it can ultimately yield a higher maximum density: the
78
self-masking action by the Prussian blue in the shadow regions of a

printed-out image is diminished by the reversal to white, which allows
the ingress of more light and formation of more (white) product. After the
exposure, the Prussian white so formed is oxidised back to Prussian blue,
either slowly by the oxygen of the air, or more rapidly by including a bath
of an oxidising agent, such as hydrogen peroxide or a dichromate, in the
wet-processing sequence:
Prussian white + hydrogen peroxide ® Prussian blue + hydroxide ions
When pure Prussian blue absorbs light, the excited state quickly loses
its excess energy by thermal relaxation only, and returns to the ground
state unaltered (apart from a slight warming of the substance). However,
if there is an oxidisable impurity in intimate contact with the Prussian
blue - for instance, trapped in its lattice - then the excited state may
have time to accept an electron and undergo chemical reduction to
Prussian white. This photoreduction underlies Herschel’s observation of
bleaching Prussian blue (§4.1.6) and is the reason why cyanotypes tend
to fade in the light; the phenomenon is central to the problems of
conserving cyanotypes, and will be fully discussed in §9.1 and Appendix
III.9.
79
4 Historical Variations on Cyanotype

"The Artists go on boldly,
and are not afraid to be Chemists,
the Chemists gain courage
and long to be Artists."
The Athenæum 1858 253
This chapter reviews some of the many variations on a cyanotype theme

which have been used at various points in photographic history to arrive
at images in Prussian blue, including the employment of glass, plastic
and fabrics as alternatives to paper for the substrate of the cyanotype.
The retrospective application of image-toning chemistry to convert the
Prussian blue into differently coloured substances will be largely deferred
until §8, which is devoted to the various methods of ‘toning’ cyanotypes.
4.1 Herschel’s formulae

From Herschel’s observations in his 1842 paper together with the notes
in his Memoranda, we can distinguish at least fifteen variations on his
photochemical theme for making images in Prussian blue, but only the
most significant will be considered here. As we shall see, subsequent
experimenters have re-invented several of these, or played further
variations upon them.
4.1.1 Proto-cyanotype
This name is newly coined here to distinguish the method of cyanotype
that has emerged from the historical evidence presented in §2.6 as the
very first method discovered by Herschel to provide prints in Prussian
blue. This was simply the exposure of a coating of potassium ferricyanide
alone. The concentration of the solution to be spread on the paper
should be lower than saturation (which is 33% w/v at room temperature),
say 10% to 16%, otherwise undesirable crystallization may occur on the
surface before the solution is absorbed by the fibres. This is the simplest
cyanotype method of all, and is capable of producing a well-graduated,
clean, negative-working print of a somewhat greenish-blue (23C7 to
23E7), but of rather low maximum density, because it does not undergo
tonal reversal (‘solarisation’), and is therefore self-masking during
exposure. The chief disadvantage of this simple method is the very long
exposure required - in the order of one or two hours to bright sunlight -
because potassium ferricyanide is intrinsically not very sensitive to light.
80
Experimenters are also advised that perceptible amounts of deadly

poisonous hydrogen cyanide and cyanogen gases are released during the
exposure of this substance, which should not, therefore, be carried out in
a confined space.254 (The same warning does not apply to the much
faster processes of cyanotype using ammonium ferric citrate or oxalate.)
Herschel noted that greater ‘speed’ and higher densities can be
obtained by developing the image in a bath of ferric chloride or nitrate
solution, but this invariably causes blue fog in the highlights.255 These
are probably the reasons why Herschel appears to have abandoned the
method soon after he discovered it, in favour of the much faster
sensitizer which also incorporated ammonium ferric citrate provided by
Alfred Smee. However, the use of potassium ferricyanide by itself
continued to be recommended for several years as ‘one of the best
methods’, by authorities such as Robert Hunt256 and Golding Bird:
“This easily prepared paper affords a ready and excellent means for
copying ferns, sea-weeds, lace-work, engravings, &c., but is not
sufficiently delicate for the camera.” 257
Bird’s popular text, The Elements of Natural Philosophy, in its second

and later editions, is thought to have provided an accessible instruction
manual for many early practitioners of the cyanotype process.258 There is
at least one important album containing cyanotypes and their original
botanical specimens, dating from 1843, which appears to have been
printed by this simple means, as will be described in §5.6.
A similar proto-cyanotype procedure was devised in 1864 by
Professor Gintl of Prague, who sensitized paper with potassium
ferricyanide in conjunction with gelatin and exposed it to achieve
selective photohardening of the colloid (see §4.10), the soluble portions
of which were washed away by water before the image was developed in
Prussian blue by means of a bath of ferric chloride.259
The ‘proto-cyanotype’ method was still being published in the 1867
edition of the influential Dictionary of Photography,260 but by the time of
the first ‘revival’ of cyanotype towards the end of the nineteenth century,
the method using potassium ferricyanide alone had disappeared entirely
from the manuals and encyclopaedias. Very recently the colloid
photohardening capability of ferricyanide has been revived by Jim
Patterson as a means of making pigented gelatin prints by the so-called
'Carbon' process, without having to employ the carcinogenic, and widely
banned dichromate substances, which many are now seeking to avoid.261
81
4.1.2 Negative-working cyanotype

In Herschel’s ‘standard’ negative-working cyanotype process, iron(II) is
generated photochemically from ammonium iron(III) citrate, and reacts
with the potassium ferricyanide in the sensitizer to form Prussian blue in
proportion to the light that has fallen; i.e. it is a Turnbull’s blue process
(§3.1) and the potassium ferricyanide is usually in excess over the iron(II),
so the product may be of the so-called ‘soluble’ variety. A recipe for this
negative-working cyanotype process can be found in every photographic
text that touches on alternative processes, and many of these claim to be
quoting Herschel’s original formula, but they never support this by citing
any reference. The reason for their omission is simple: nowhere in
Herschel’s published work does he specify the concentrations and relative
proportions of the solutions of potassium ferricyanide and ammonium
ferric citrate that he used for this cyanotype process. This information
can only be found in the Memoranda, as is clear in the following entry:
“Photography
FerroSesquiCyan. P Cyanogen Iron
Aug 13/42. Mixed sol of F3/2CP and Ammo Citr of Iron as in Paper
780 ie. equal parts of satd sol of F3/2CP and ACI 1/11 (1 salt + 10
water). The properties of this paper are singularly complex &
interesting.” 262 [underlining in original]
Herschel’s use of a saturated solution of potassium ferricyanide (ca.

33% w/v) and the rather dilute solution of ammonium ferric citrate (ca. 9%
w/v) are surprising. He goes on to describe in detail his observations of
the effects of exposing this sensitizer to light; in particular, his attention
was caught by the tonal reversal of the image as the exposure progressed
- a phenomenon for which he promptly coined the term ‘solarising’:263
“A. It first darkens to a dull bluish grey the original hue of the
paper being green-yellow. The darkening reaches a maximum in
about 30 or 40 seconds in good sun (the paper is very sensitive), -
It then whitens again - to a pale greenish-grey. Hence a picture
withdrawn at various stages of exposure presents several aspects.
1st If not long exposed the margin is dark and the picture negative
in a bluish or lead grey.
2d If very long exposed the picture is positive the parts where the
light has acted being actually paler than the original formed tint.
This I call over-sunning or solarising.
3d If the exposure be not so prolonged as to “solarise” it the
picture is very remarkable. It is positive, but in 3 tints the ground
being the original greenish-yellow. - the full lights being the
82
aforesaid solarised white - a very pale greenish-grey. - The half

lights are darker than the ground but contrast sufficiently with it in
colour to relieve the lights on the one hand and to stand off from
the ground on the other.
B. When the picture so produced is thrown into whater [sic] it
gradually changes to a fine Prussian Blue Negative Photograph, &
the process of the change is highly curious as to Pictorial effect.
Especially if a solarised Photogr be used.- The white faces take by
degrees a leaden hue & the change is like the effect of the
“dissolving views”.” 264 [Underlining in original]
It is clear from his 1842 paper that Herschel had a correct chemical
understanding of the cause of his ‘solarisation’; a footnote to §207 reads:
“The whitening is very obviously due to the deoxidation of the
precipitated Prussian blue and the formation of the proto-
ferrocyanuret of iron; the resumption of colour in the shade, to the
re-oxidisement of the compound, which is well known to absorb
oxygen from the air with avidity. Simple Prussian blue, however, is
not whitened by the violet rays. Its state must be peculiar.” 265
On 16 August 1842, Herschel modified the proportions of mixing his

solutions of ammonium ferric citrate to potassium ferricyanide from 1 : 1
to 5 : 3, respectively, but there is no record of his making further
changes after that. It is puzzling that the proportions and concentrations
he had then arrived at still differ considerably from those that were
subsequently recalled and publicised by his son, Professor Alexander
Stewart Herschel (1836-1907), in the 1870s as representing his father’s
working recipe.266 The concentrations resulting in the final working
sensitizer from these various formulations are summarised in Table 4.1.
Formulation Ammonium Ferric Potassium

Citrate % w/v Ferricyanide % w/v
Herschel ‘proto’ None ca 16
Herschel original 4.7 16.6
Herschel modified 6.6 12.4
Alexander Herschel 14.7 12.8
Later C19th average ca 10 (brown salt) ca 8
Usual C20th average ca 13 (green salt) ca 6
Table 4.1 Early cyanotype formulations (final concentrations)
The processing of an exposed negative cyanotype could not be

simpler: ‘fixation’ of the image is most easily accomplished by washing in
water, because the excess sensitizer and other photoproducts are entirely
83
soluble. This is one of the strongest recommendations for the cyanotype,

since it meets an ideal that Herschel had always sought in his processes;
it offers a considerable advantage over the practice necessary in silver
photography, where the insoluble silver halide requires a fixing agent,
sodium thiosulphate, for its removal, and residues of thiosulphate can
lead to deterioration of the silver image, as many early photographers,
including Talbot, found to their cost.267
4.1.3 Hydrargyro-cyanotype
This cumbersome name appears only in Herschel’s Memoranda, but the
1842 Paper carries an account of the phenomenon in Article 223, which
was republished by Hunt and others,268 and goes as follows. A cyanotype
is made and processed in the usual way from ammonium ferric citrate
and potassium ferricyanide by Herschel’s ‘Method (a)’. This is then coated
with a solution of mercurous nitrate;269 within a period of a few hours the
blue image disappears and the blank paper is washed and dried. On
heating strongly with a flat-iron - but not so much as to cause scorching
of the paper - a brown image becomes visible. This tends to fade slowly
in the dark over a matter of days, but may be regenerated again by
further heating. We can only conjecture what chemistry may be involved
here, for the observations deserve further scientific investigation; the
reversibility of the phenomenon may be a consequence of the mercury(I)
being trapped in the Prussian blue lattice in the same way as thallium(I) is
known to be, and then reducing the substance to Prussian white. The
effect of heat may be to cause decomposition of the mercury ions to
mercury metal, which in a finely divided state would appear brown. I have
repeated and confirmed this curious process, but cannot recommend it
generally owing to the high toxic hazard presented by the mercury salts,
especially when heated.
4.1.4 Positive-working cyanotype
In the positive-working cyanotype process, the Prussian blue image is
meant to be formed in inverse proportion to the illumination, by the ferric
ions derived from the unchanged ammonium ferric citrate sensitizer
reacting with potassium ferrocyanide, which is applied in solution after
the exposure. This procedure introduces an inherent and rather
intractable problem: the lighter regions of the image, which have received
the most exposure, contain much ferrous iron that also reacts with the
applied ferrocyanide to give the insoluble product, ferrous ferrocyanide
or Prussian white. Now this substance does not remain white for very
84
long, being readily oxidised by the air to Prussian blue, so it must be

removed in order to fix the positive image. Any treatment intended to
achieve this is likely also to remove the Prussian blue image itself, which
explains why Herschel - and everyone who has followed him - found the
positive-working version of the process so much more difficult to carry
out satisfactorily, and virtually impossible to perfect. Herschel’s first
recourse was to include gum Arabic in the developer to increase its
viscosity, but this did not fully answer the problem. Had he included it in
the sensitizer, as Pellet subsequently did (see below) he would have met
with more success. His second attempted solution was to include
mercury(II) chloride, ‘corrosive sublimate’270 in the sensitizer in order that
the image, after development with a coating of gum and potassium
ferrocyanide, would be fixed without any washing at all:
“All the care that is required is, that the film of liquid be very thinly,
evenly, and above all, quickly spread. Being then allowed to dry in
the dark, it rarely fails to produce a good picture. And what is very
remarkable, it is ipso facto fixed as soon as dry, so at least as not
to be injured by exposure to common day-light, immediately; and
after a few days’ keeping it becomes entirely so, and will bear
strong lights uninjured.” 271
In spite of Herschel’s claim for the success of this procedure it cannot be

recommended, in view of the very high toxicity (and expense) of
mercury(II) chloride. Even today this chemistry is somewhat obscure.
Herschel offered no rationalization, but it is known that mercury(II) has
an enormous affinity for cyanide, forming a very stable complex; its likely
effect is to destroy the Prussian white in the highlight areas, which cannot
then revert to blue.
4.1.5 Anachronistic ‘Herschelotype’
A curious inversion of the foregoing process was proposed in 2007 as a
practical alternative method for negative-working cyanotype: to expose
ammonium ferric citrate alone, then develop it in potassium ferrocyanide
solution to form an invisible negative image in Prussian white, which then
slowly re-oxidizes in air to Prussian blue, so providing a negative-
working process overall. This procedure may sound plausible but, as we
have just seen in §4.1.4, it has an obvious problem: a strong positive-
working image in Prussian blue is also formed immediately by the
ferrocyanide reacting with residual iron(III) in the unreduced sensitizer,
and this initial blue image must be completely and quickly dispersed – by
scrubbing it away, according to the author - before the Prussian white
85
negative image becomes oxidized by air to Prussian blue. It seems very

improbable that this procedure could be employed successfully for
making prints of any size and quality; no specimens hae been offered for
inspection. The author goes on to announce, rather stridently:
"This new process, as the inventor, I'm attributing to Herschel and
call it the "Herschelotype"." 272
Retrospectively to enlist the name of Sir John Herschel for this

unworkable and falsely-claimed "new process” is a clumsy
misappropriation that insults his memory. It is presumptuous to assume
this 'right of naming' because Herschel never attempted to use this
process, nor did he employ potassium ferrocyanide in the negative-
working/reoxidation manner. He did, of course, attempt to use
ferrocyanide development in the contrary sense, as a positive-working
process, but rarely with much success (§4.1.4). Moreover, the process
described is in no way new, and any modern claim to be its inventor is
empty. The inclusion of potassium ferrocyanide in negative-working
cyanotype was, in fact, widely patented in the USA and UK in the 1930s,
and employed commercially for the production of very fast blueprint
papers of the ‘Blue Layer’ type as described in §4.11. The only difference
was that the potassium ferrocyanide was more conveniently and
economically incorporated in the sensitized coating, rather than used as a
developer bath.
4.1.6 Light-bleaching of Prussian blue
Herschel discovered that there is yet another way of obtaining a positive-
working image in Prussian blue; namely, to cause the image-forming
light to bleach a uniform ground of prepared Prussian blue, by promoting
its reduction to Prussian white in the presence of certain oxidisable
substances (the function of which he may not have been aware).
Unfortunately, such images cannot be fixed because there is no good way
of removing the Prussian white without damaging the Prussian blue, as in
the case above, so the picture quite rapidly reverts to its original uniform
deep-blue ground under the effect of aerial reoxidation. Herschel was
quite taken with this process, nonetheless, to the extent that he
described it in a letter to the readers of The Athenæum, in preference to
his negative-working process, as follows:
“Mix together equal parts of the solution of the above-named salt
[ammonio-citrate of iron], and of a saturated solution of, not the
ferrosesquicyanate of potash (as stated in the Athenæum), but the
common yellow ferrocyanate, or as it is called, prussiate of potash.
86
The result is a very black ink, which, washed over paper, gives it a
deep violet-purple colour, and is remarkably sensitive to light -
whitening rapidly, and giving positive pictures - the only defect of
which (and it is a fatal one for use) is their want of durability, as
they fade with darkness in a few hours. And what is very singular,
the same paper is again and again susceptible of receiving another
and another picture, which die away in like manner without any
possibility, so far as I have yet discovered, of arresting them.” 273
(italics in original)
The reported reversibility of this phenomenon is most striking, and it

is given further consideration in Appendix III.9 and §9.1 on experiments
on the fading of cyanotypes by light. This formulation is, in fact, a very
early precursor of the ‘blue layer’ type of blueprint paper that was
developed as late as the 1940s (see §4.8) using potassium ferrocyanide
instead of ferricyanide. But the fact that Herschel was using brown
ammonium ferric citrate meant that free ferric iron caused the formation
of much Prussian blue immediately, which could not be removed and
enable the use of the paper in a negative-working mode, through the
intermediate formation of Prussian white, which reoxidised to Prussian
blue. Herschel’s description implies that his chemicals must have been
rather impure; when this combination is made with modern reagents, one
does not get ‘a very black ink’, only a greenish coloration. Had Herschel’s
chemical been of a higher purity (the green variety of ammonium ferric
citrate rather than the brown), the benefits of using potassium
ferrocyanide might have been discovered much earlier.
4.2 Pellet’s process

Thirty-five years had to elapse following Herschel’s experiments on the
positive-working cyanotype process before it was commercially exploited
and gained some acceptance in practice. It was promoted by the
carbohydrate chemist Henri Pellet, who obtained a British patent in 1877
for a process that bears his name.274 It was also patented in France in
1877,275 and in the USA in 1881,276 under the names of Cyanofer and
Papier Gommoferrique. The new feature that Pellet introduced to solve
the problem of fixing the positive-working process was based on earlier
observations by Alphonse Poitevin in 1863, that ferric salts cause gum
and similar colloids to harden and become insoluble in water, whereas
ferrous salts do not.
Pellet included ferric chloride and gum Arabic in the sensitizer layer, in
order to harden it and protect the Prussian blue image formed in the
87
unexposed regions upon development. In the exposed regions, there

were two consequences of the light-induced reduction of iron(III) to
iron(II): the gum became re-solubilised and, on development in
potassium ferrocyanide, insoluble Prussian white was formed, which
could be readily wiped off, thus fixing the image without much disturbing
the Prussian blue. Pellet’s successful positive cyanotypes were exhibited
in the 1878 International Exhibition held in Paris, and apparently excited
the admiration of J M Eder.277 Pellet’s patents, however, are framed in
such vague ‘catch-all’ language that it is not possible to extract from
them a clear working method - doubtless, just as he intended. Solutions
of many types and compositions are cited as examples. Lietze’s Manual
sets out an abridged version of Pellet’s US Patent of 1881; the following
brief quotation will convey a taste of its wholly ambiguous style:
“After exposure, it is passed through a bath of yellow prussiate
(ferrocyanide of potassium), the bath being concentrated, or dilute,
neutral, slightly acid, or slightly alkaline, hot, or cold.” 278
Given information of this opacity, it comes as no surprise that, at the

end of his transcription, Lietze is moved to observe:
“The author has tried Pellet’s process, several times ... without
succeeding in producing good prints. The background was always
more or less colored.” 279
Pellet is said to have received much benefit and distinction from his
process at the time,280 but his commercial paper met with varying
approbation because the processing was quite complicated and not a
little uncertain in its outcome. In addition to Lietze, other contemporary
authors have reported on its unsatisfactory performance.281 By 1943, the
process was deemed to be defunct.282
In contrast to Pellet’s obfuscated patent specification, two clearly-
formulated, but slightly different, versions of Herschel’s positive-working
process were subsequently published in many manuals.283 It should be
noted that these recipes are given according to the ‘old fashioned’
method for making up solutions (see Appendix III), but are here
converted into metric units.
The first was published by Giuseppe Pizzighelli and Ludwig von
Itterheim in 1881. Three stock solutions are required:
A. Gum Arabic: 20 g in 100 cc of water
B. Ammonium ferric citrate: 50 g in 100 cc of water
C. Ferric chloride: 50 g in 100 cc of water.
88
These three solutions are mixed in order, in the volume ratios:

A : B : C = 20 : 8 : 5
Regarding the coating, the published accounts differ significantly in
their advice: the mixture becomes very viscous after a few minutes, in
which condition, one source says, it is suitable for coating the paper;
however, some accounts assert the contrary and recommend immediate
coating before it thickens. Yet another description states that the thick
cloudy mixture eventually becomes clear and liquid again, and that this is
the time when it is ready for use.
After exposure (which is markedly shorter than for the conventional
cyanotype) the image is developed by a solution of potassium
ferrocyanide: 20 g in 100 cc of water, which may be brushed on. It is next
washed in water and the soluble gum is removed by a soft brush. Finally
it is cleared in a solution of dilute hydrochloric acid, washed again and
dried.284
The second process is often attributed to Colonel James
Waterhouse,285 but he, in turn, acknowledges that the method was
recommended by Fisch.286 According to Duchochois,287 the formula is
substantially that devised by Poitevin, with the addition of gum Arabic,
according to Pellet’s suggestion. Three stock solutions are again
combined at the time of coating (the quantities below have been scaled to
make them directly comparable with the method described above):
A. Gum Arabic: 26 g in 100 cc of water
B. Tartaric acid: 27 g in 100 cc of water
C. Ferric chloride: 100 g in 100 cc of water288
These three solutions are mixed in order, in the volume ratios:
A : B : C = 13 : 3 :2
The mixture is set aside in the dark for 24 hours then diluted with
distilled water to a hydrometer reading of specific gravity 1.100, when it
is ready for coating. The processing procedure is identical to that given
above.
Under the direction of Waterhouse, both these procedures were tested
and compared by a Mr Pope in the Survey of India Photographic Office,
Calcutta, in 1887-8. In the climate of that country, they found that paper
sensitized by the second method had the better keeping properties. The
process was quite extensively used in Waterhouse’s office for duplicating
maps in the the Survey of India.289 The process was much less common
than blueprint, and since it called for greater skill, it tended to be
confined to the hands of professional blueprinters.290
89
4.3 Survey of negative-working formulae

As was seen in Table 4.1, Herschel’s two recipes for negative cyanotype
in his Memoranda differ from the proportions that his son Alexander
recalled in the 1870s, and these differ again from the proportions that
are commonly claimed to be Herschel’s by the principal manuals on the
subject. In the absence of further information to resolve this disparity, I
felt it necessary to investigate all four formulations in the experimental
investigations described in §9.1.
The only significant chemical advance in the entire history of the
cyanotype process was brought about by the preparation of an alternative
form of ammonium ferric citrate in 1897 by Eduard Valenta (1857-
1937).291 This new substance is green and somewhat acidic, in contrast
to the brown, and basic form of ammonium ferric citrate which had been
used hitherto; it was claimed that the green form provided a sensitizer
eight times faster than the brown.292 However, the apparent speed
depends on the criterion for its assessment; there is less speed difference
in the mid-tones than in the shadows where solarisation has an effect. A
comparison of the two substances is summarized in Table 4.2.
Property Brown Salt Green Salt

First preparation ca. 1840 Valenta 1897
Iron content 19% - 28% 14% - 18%
Nature of salt Basic Acidic
Speed of printing Slow ca 1 stop faster in midtones
Reversal or ‘solarization’ Relatively slight Stronger effect: higher Dmax
Edge etch effect Distinctive Absent
Colour of image Dull blue Brighter blue
Light fading of image Fairly resistant More susceptible
Table 4.2 Comparison of green and brown ammonium ferric citrate

as cyanotype sensitizers
The sources of the green salt were limited at first to a few chemical
manufacturers in Germany,293 but it soon became available in the USA,
and after 1900 the majority of cyanotypes were probably made by its
means. The chemical implications of this discovery are examined in more
detail in Appendix III.5. Most 20th century formulations of cyanotype
recommend the green form of the salt in preference to the brown.294
90
I have compared the formulae in about sixty published recipes for the
cyanotype process,295 by converting the various archaic expressions for
solution strengths into a single system of units: the concentrations of
potassium ferricyanide and ammonium ferric citrate are expressed as
‘weight per cent volume’ (% w/v - see Appendix IV) in the final, mixed
sensitizer solution, thereby avoiding the problem of comparing recipes
where the two stock solutions are not mixed in equal amounts.296 The
results of this survey take account of the introduction of green
ammonium ferric citrate in 1897, and are conveniently displayed in
graphical form in the scatter chart, Figure 4.1, which conveys some
impression of the range of different proportions that have been
recommended overCyanotype
150 years. Formulae -- Brown Green
K3Fe(CN)6 % w/v
18.0
14.0
10.0
6.0
4.0 8.0 12.0 16.0 20.0
Ammonium ferricAmfecit
citrate(brown)
Amfecit % w/v %✳ w/v
(green) green salt; • brown salt
Fig 4.1 Composition of cyanotype sensitizers
91
Apart from three wild ‘outliers’, the recipes using brown ammonium
ferric citrate are grouped around average final concentrations of 10% for
the brown ammonium ferric citrate and 8% for the potassium ferricyanide.
The recipes using the green salt are clustered around final averages of
13% for the green ammonium ferric citrate and 6% for the potassium
ferricyanide.
As can be seen from the range of concentrations in the chart, the
process appears to work satisfactorily almost regardless of the
proportions chosen; presumably there is little difference to the visible
outcome - at least so far as these many practitioners were able to judge.
This is not particularly surprising when it is realised what a small
proportion of the sensitizer originally applied to the paper actually
remains as Prussian blue in the final image (see Appendix III).
The flexibility in the formulae using the brown ammonium ferric
citrate is born out by the report of Channing Whitaker in 1883, who
tested sensitizer solutions with a wide range of compositions, having the
concentration ratio of ammonium ferric citrate to potassium ferricyanide
varying from 2% : 20% to 20% : 2%, respectively. On the basis of his
results, Whitaker settled on an optimum ratio of 15% : 10%. He remarks
that ‘the same blue results with any good proportions of the chemicals
named.’297 According to Whitaker, excess of the citrate speeds printing
but shortens the storage life of the coated paper; excess of the
ferricyanide provides a finer quality but greatly lengthens the exposure
time (due, as we now know, to the ‘internal filter effect’).
Anyone wishing to make use of this method today will find it
convenient to mix equal volumes of the two stock solutions; the
recommended average concentrations of these should therefore be made
up approximately to the following:
20% w/v for brown ammonium ferric citrate, and
16% w/v for potassium ferricyanide;
or
26% w/v for green ammonium ferric citrate, and
12% w/v for potassium ferricyanide.
Further practical details are given in §7.1. The differing ratios of the
two components reflect the different iron contents of brown and green
forms of ammonium ferric citrate, which are further discussed in
Appendix III.4.
92
4.4 Modifications of the cyanotype sensitizer

4.4.1 Robert Hunt’s ‘chromo-cyanotype’
In 1844, Robert Hunt (1807-1887) published a paper entitled ‘Chromo-
Cyanotype, a new Photographic Process’.298 Hunt had been studying the
light sensitivity of potassium dichromate, and had already devised several
photographic processes based upon it, such as the chromatype, in which
it was mixed with copper sulphate. The chromo-cyanotype sensitizer was
a mixture of potassium dichromate and potassium ferrocyanide;
depending on the developing agent, which could be either ferrous
sulphate or ferric sulphate, a positive- or a negative-working image,
respectively, was obtained, the latter being of rather stronger quality.
Hunt found this difficult to explain at the time, and it may be significant
that he made no further mention of this process in his subsequent
Manual of 1853.299 It is now known that potassium dichromate rapidly
oxidises potassium ferrocyanide to the ferricyanide;300 if the effect of
light is to cause the photoreduction of this back to the ferrocyanide,
possibly assisted by a chromium(IV) or (V) species which would be
present, the development phenomena reported by Hunt become
comprehensible. A number of cyanotypes feature among Hunt’s early
photographic specimens which were re-examined in 1890, when his
widow sent a parcel of them to John Spiller.301
4.4.2 John Mercer’s ‘chromatic photographs’
John Mercer (1791-1866) was a self-taught colour chemist and calico-
printer of Manchester, and is remembered today for inventing in 1850
the process to which his name has been given – that of mercerization, the
strengthening and shrinking of cotton by treatment with caustic alkali. I
have already noted in §2.2 that he has the distinction of being the
earliest observer (in 1828) of the photochemical formation of Prussian
blue; but he did not follow up this discovery until 1847, when he evolved
the following method for obtaining negative cyanotypes:
“I found that if paper or cotton cloth is smeared with a solution of
pernitrate or persulphate of iron with certain quantities of oxalic
and tartaric acids, dried in the dark, then exposed to the light
(solar), and immediately dipped in solution of red prussiate
containing a little free sulphuric acid (preferably free from yellow
prussiate), where the light has caused deoxidation, the blue is
fixed, and where the peroxide remains unchanged there is no
colour. The cloth or paper must be washed immediately in water
containing a little sulphuric acid, and afterwards in pure water. The
picture is a reverse blue print.” 302
93
Mercer’s business commitments allowed him little time for progressing

these experiments until 1854, when he began to produce striking, richly-
coloured prints on paper, and cotton fabrics especially, which were
exhibited at the annual meeting of the British Association for the
Advancement of Science, in Leeds in 1858.303 The short note in the
Transactions describing this lecture seems to be the entire extent of
Mercer’s publication of his work,304 but fortunately Parnell’s biography of
Mercer contains details of his formulae,305 taken from Mercer’s own
notebooks and memoranda.
Fig. 4.2 John Mercer’s chromatic photographs on cloth

(Mercer Collection of Lancashire Archive)
His preferred process was based on a cyanotype formulation of a

singular kind, that seems not to have been previously employed by
anyone else.306 His sensitizer was a solution of ferric oxalate, prepared by
94
the oxidation of freshly precipitated ferrous oxalate by nitric acid in the

presence of excess oxalic acid. He does not explain his choice of this
troublesome preparation, rather than using the universally available
ammonium ferric citrate, but it must have been due to his particular
interest in printing on textiles rather than on paper. He had already
shown that ferric oxalate is a good mordant for dyeing wool and cotton
and had introduced its use into his own dyeworks; whereas the citrate is
ineffective in not being absorbed by the cellulose fibres.
Following the light exposure, he would sometimes treat the product
directly with potassium ferricyanide solution to obtain a cyanotype as
described above; but Mercer also adopted another unique procedure:
fixation in a bath of sodium arsenate. This disagreeably toxic substance
does not appear to have caused him any problems (it was then commonly
employed by calico-printers) but it can hardly be recommended for
general use today. Its effect is to convert both iron(II) and iron(III)
oxalates into the corresponding arsenates, which are insoluble in water
and photoinactive, but soluble in acids; thus the image in iron(II) is
‘fixed’, and then it may finally be ‘brought up’ by treatment with an
acidified solution of potassium ferricyanide, which dissolves the iron(III)
arsenate and converts the iron(II) arsenate into Prussian blue. If this
procedure seems convoluted, it cannot be denied that it yielded fine
permanent cyanotype images on calico and mercerised cotton, which
survive well from 1857 to the present in the Mercer archive of Lancashire
County.307
4.4.3 Charles Burnett’s ‘cobalt cyanotype’
Charles John Burnett (1820-1907) is an equivocal figure in the history of
photographic processes. His publications attest an eclectic interest in
photoactive substances of all kinds, but it is sometimes hard to
discriminate in his writings between the results of genuine experiment
and the speculations of an inventive mind. As far as contributing to the
cyanotype process goes, there is clear evidence that he exhibited a
‘Photograph of Sea-weed’ in Glasgow in September, 1855; the catalogue
carries this description, originally contained in a letter from Burnett to
William Church:308
“Printed in natural colour, by process discovered in March last;
paper prepared by ferridcyanide[309] of potassium, development by
nitrate of cobalt solution.” 310
95
Burnett describes his experiments in more detail in Photographic

Notes of 1857,311 but is obliged there to admit that the ‘cobalt process’
prints, although remarkably beautiful initially, suffered a ‘very marked
deterioration’ within a few months.
Another innovation suggested by Burnett in this article is the use of
sodium nitroprusside instead of potassium ferricyanide for forming the
pigment. Despite the differing names these are very similar compounds
as may be seen from the formulae.312 Burnett states that this salt “by
itself gives a blue and white picture” and when mixed with ammonium
ferric oxalate the resulting sensitizer yields a “dull green” image. The
present writer has confirmed these claims but my investigation found a
number of disadvantages – such as crystallization and very slow printing
– which do not recommend these as viable printing procedures, especially
in view of the toxicity and expense of sodium nitroprusside. Apparently
Burnett exhibited prints made by these methods in 1859 at the London
Photographic Society.313 However in 1899 a US patent claimed that
soluble nitroprussides could be combined with ammonium ferric citrate
to give sensitizers yielding brown or black images – a surprising
colour.314 This is still an area with potential for some experimentation.
4.4.4 Other iron(III) sensitizers
Several alternatives to ammonium ferric citrate as the light-sensitive
component of the cyanotype have been suggested over the years.
Ammonium ferric tartrate was first tried by Herschel, and the
corresponding potassium salt was re-introduced in 1892 by an author
who admitted that he had not read Herschel’s work.315 No particular
benefit has been claimed to result from their use, and they are rather
uncommon chemicals.
In 1900 the brothers Lumiére found that ferric sucrate, made from
cane sugar, is light-sensitive and has good keeping properties; a paper
coated with this salt and exposed in the usual manner could be
developed as a negative or positive cyanotype in potassium ferricyanide
or ferrocyanide solutions.316 This process does not seem to have been
followed up.
Valenta, in 1917, reported the preparation of ammonium ferric
glycolate, and found that paper sensitized with a mixture of this and
potassium ferricyanide yielded prints of a
“...fine intense indigo-blue colour, such as are not yielded with
ferric ammonium citrate. But unfortunately the sensitized paper has
exceedingly poor keeping quality, fogging in the course of a day or
96
two in the dark ... It was not found possible to avoid this drawback
by addition of oxidising substances to the sensitizer.” 317
The likelihood is that the quality of paper at that time may well have
been inferior, and impurities may have been responsible for the reported
rapid degradation. In view of this possibility, this novel photosensitive
salt, and the cyanotypes that it furnishes, would seem to deserve re-
examination.
The use of ferric oxalate as a cyanotype sensitizer has already been
mentioned in §4.4.2, and will arise again in §4.4.5, but this salt is highly
variable in its properties and composition, and is troublesome to obtain;
a much better alternative is the well-characterised substance, ammonium
ferric oxalate, which can be widely purchased and does offer some real
advantages over the traditional citrate, as will be described separately in
§7.5. The use of ammonium ferric oxalate for ‘pictorial’ quality cyanotype
sensitizers has been proposed in the past (§2.9), but when mixed with
potassium ferricyanide, the sparingly soluble salt, potassium ferric
oxalate, tends to crystallise out, ruining the coat of sensitizer. The only
procedure suggested to counter this was the wasteful one in which the
‘potential’ image was formed in ammonium ferric oxalate by itself, and
then after exposure developed in a separate bath of potassium
ferricyanide solution. Hermann Halleur in The Art of Photography of 1854
described this negative-working cyanotype, and recommended exposing
“oxalate of sesqui-oxide of iron and ammonia” (i.e. ammonium ferric
oxalate) alone, then developing it by brushing over a solution of
potassium ferricyanide.318 Matthew Carey Lea used this method of
making cyanotypes to avoid the crystallization problem, and described it
to the Annual Meeting of the British Association in 1863.319 Although this
method was also advocated with enthusiasm in an annual editorial review
by Traill-Taylor in 1889,320 it never passed into general practice,
probably because the development bath rapidly becomes contaminated in
use with excess sensitizer and precipitated Prussian blue, which will stain
the print - this is clearly uneconomic in view of the price of potassium
ferricyanide. In 1921 the claims of a German patent that ammonium ferric
oxalate could be employed in the sensitizer in place of the citrate were
tested experimentally by Valenta, who came to the conclusion that the
claims were unjustified and it was inferior because of the crystallization
problem.321 However, it will be shown in §7.3 how ammonium ferric
oxalate can be used in a modern improved cyanotype process.
97
A related variant to this is the employment of the unusual cation in

guanidinium ferrioxalate for the cyanotype sensitizer.322 A US patent of
1941 claimed that this sensitizer had a better shelf life,323 but it does not
seem to have been taken up for cyanotype, although it has been
employed recently for an iron-silver process.324
From time to time a number of additives have been suggested to
improve the performance of the traditional sensitizer, but these
proposals are rarely acccompanied by direct experimental comparisons to
demonstrate their putative benefits. However, in 1955 Hnatek published
a detailed investigation of the enhancement of the sensitivity of brown
ammonium ferric citrate by the addition of carboxylic acids.325 He found
that oxalic or tartaric acids increased the sensitivity and printing range
considerably, but that citric or acetic acids had no significant effect. Most
acids will assist clearing because a low pH prevents hydrolysis of the
iron(III) to its insoluble yellow hydrated oxide. Moreover the light
sensitivity of ammonium ferric citrate has been found to pass through a
maximum at pH 3 to 4. The presence of a small amount of dichromate
assists sensitizer stability and shelf-life by preventing premature
formation of Prussian blue. Many of these findings probably stem from
endeavours to improve the characteristics of commercial blueprint papers
which will be considered later in §4.8.
4.4.5 Terry King’s ‘cyanotype rex’
In 2004 the then Chairman of the Historical Group of the Royal
Photographic Society, the late Terry King, announced his 'Cyanotype Rex'
process which claimed "modern improvements" both on the traditional
cyanotype and the New Cyanotype of 1995 described in §7.326 The first of
these improvements was the use of ferric oxalate, rather than ammonium
ferric citrate or ammonium ferric oxalate, as the sensitizer. The author
seemed unaware that ferric oxalate had been employed by Mercer in the
1850s (§4.4.2) and that its use for negative-working cyanotype was also
clearly described in the British Journal of Photography of 1874;327
moreover, ferric oxalate eventually became a common ingredient in the
many patented blueprint formulations of the 1930s (§4.11). However, the
availability of ferric oxalate today is very limited: its preparation is beset
with uncertainty over the nature and purity of the product and, owing to
this variability of the substance's composition, few fine chemicals houses
will stock it at an acceptable price. Ferric oxalate is a polymorphic and
notoriously 'ill-characterised' compound, reluctant to dissolve in water,
98
unstable when persuaded to, and it gives a solution that is limited in

shelf-life and unreliable in practice.328 Its present-day use for alternative
printing processes is further disadvantaged by its monopolistic price,
which is many times that of pure ammonium ferric oxalate, the preferred,
commercially-available photosensitive salt.
King’s second improvement was to adopt the misapprehended
procedure attributed to Herschel, discussed above in §2.9, of exposing
the sensitizer coating alone, then applying the potassium ferricyanide in
a separate development bath.329 The serious print-maker will find this
method to be both expensive and wasteful, as mentioned in §4.4.4,
because the ferricyanide developer bath soon accumulates a staining
impurity of Prussian blue, which causes a rapid fall-off in the quality of
successive prints put through it, and it must therefore be frequently
replaced. A litre of 10% potassium ferricyanide solution may cost over £5.
The third “improvement” offered by the ‘Cyanotype Rex’ process is
stated thus:
“Where traditional cyanotypes can prove difficult to tone, the
Cyanotype Rex tones beautifully in alkaline solutions of varying
strength…” 330
It has long been recognised that the destruction of Prussian blue by

alkalies is far from “difficult”, so there seems to be little novelty in this
“improvement”. The well-known historical toning procedures for
cyanotypes are described in detail in §8.
King’s policy of ‘separate development’ can, however, justly claim one
advantage: an increase in printing speed. In a conventional mixed
cyanotype sensitizer the ferricyanide ion inevitably absorbs some of the
actinic blue and UVA light non-productively, owing to its strong
absorption band peaking at 420 nm, which thereby lowers the sensitivity
(Appendix III.2). By placing the ferricyanide entirely in a development
bath, this ‘internal filter effect’, as it is known, is eliminated, with a
consequent restoration of speed amounting to about two or three stops
(4-8x) at most. This was hailed as a “huge improvement in speed” for the
Cyanotype Rex process, which was said to provide “viable camera-speed
photography” - a claim that calls for careful examination. Where Maunder
and King refer in their articles to:
“…exposure times well down to a minute.” 331
“…exposures cut down to a few seconds or minutes.” 332
99
©Mike Ware 2017 Cyanomicon_II100
it transpires that they are being economical with the truth. They do not
make it clear that these times refer to contact exposures, which require
roughly a hundredth of the in-camera exposure time – see §4.6. Contact
exposures of a minute agree with the theoretical calculations of §3.6, and
with the experience of anyone who has made a new cyanotype (§7), which
takes about one or two minute’s exposure to the sun for a photogram (in
contrast to the traditional sensitizer formula, which may take 10-20
minutes). In-camera exposure of the new cyanotype sensitizer is found to
take a few hours at f/3.5 to produce a negative having a realistic density
range, according to the tests shown below in §4.6.333
King’s publication in the BJP does not disclose the camera exposure of
his 10x8 inch test negative, but it is believed to have been in the order of
an hour at maximum lens aperture, which suggests that there is actually
little point in trying to produce cyanotype negatives in the camera. An
hour is hardly a useful exposure duration for portraiture, living people, or
subjects in any state of motion, so the claim that these processes are
“viable in-camera technologies” that will “revolutionise photography” is
an absurd exaggeration. The only evidence adduced to support
Maunder’s optimistic assertion that:
“In camera exposures of around a minute appear achievable.” 334
is his deceptive camera negative consisting of a pale blue sky, with

silhouetted plants and windowframe, exposed for 5 minutes at f/2.8.
Careful inspection suggests that this is about 4 or 5 stops underexposed,
so the proper exposure for this negative should be 2-3 hours, consistent
with the calculations above.
Finally we note King’s announcement that he applied the philosophical
working method that he called “retro-invention” to discover the
advantages of using ammonium ferric oxalate in the process he named
‘Cyanotype Rex (2)’.335 As we saw in §4.4.4, this was first published by
Hermann Halleur just 160 years ago: “retro-invented” indeed!
4.4.6 Cyanotype using uranium and vanadium salts
Among the many variations suggested for the cyanotype process, one
which has receded into obsolescence, probably for safety reasons, is the
addition of uranium(VI) salts to the sensitizer.336 This was claimed to
enhance sensitivity, and to speed up printing by catalysing the
photoreduction of iron(III) to iron(II). Tests by this author have shown that
addition of uranyl nitrate at the 1% to 2% level to both ferrioxalate and
ferricitrate sensitizers does indeed increase speed of printing by about
100
one stop (a factor of 2x) but further increases in the uranium

concentration bring little additional advantage, so its use is not worth the
toxic risk.337 Another cyanotype additive for which a 4x speed increase
was claimed is vanadium pentoxide, as described in a US patent of
1924.338
A form of so-called ‘uranium cyanotype’ (a serious misnomer) or,
more accurately, ‘uranotype’ should also be mentioned here, in which the
sensitizer consists entirely of a uranium(VI) salt, which is transformed by
light to uranium(IV). Treatment after exposure in a development bath of
potassium ferricyanide yields an image in red uranyl ferrocyanide, a
process which was originated by Burnett.339 Examples of contemporary
users in the 21st century include alternative process artists Robert W.
Schramm,340 Bob Kiss,341 and the late Blake Ferris (Fig. 4.3).342
Fig. 4.3 Blake Ferris, Uranotype
The rich red colour and relative stability of uranyl ferrocyanide have
also suggested the use of this substance as a toner of silver-gelatin
materials,343 and of platinotypes, where the action is catalytic.344 By
101
treating an uncleared platinum print, which still contains iron(III), with this
toner, the two pigments, uranyl ferrocyanide and Prussian blue, one red
and one blue, are formed simultaneously, and the mixtures of them can
therefore produce a whole gamut of unattractive false colours
superimposed upon the long-suffering platinum image.345
4.5 Cyanotype diapositives on glass and film

From the end of the 19th century, when the ‘magic lantern’ enjoyed great
popularity, it is not surprising to find published instructions for the
making of cyanotypes on glass slides. According to one account, the
procedure is simply to coat a glass plate with gelatin and, when dry,
immerse it in a conventional cyanotype sensitizer.346 A more detailed
procedure starts with old, unused (and probably fogged) silver-gelatin
plates and describes how the silver salts are first removed by a
thiosulphate fixer bath, and the gelatin layer then hardened by chrome
alum. If the plates already have a silver image, they can be cleaned by a
ferricyanide/thiosulphate bath (‘Farmer’s Reducer’). After washing, the
hardened gelatin-coated plate is soaked in a conventional cyanotype
sensitizer for several hours, then carefully wiped and dried. Exposure and
processing is similar to that of a conventional paper cyanotype. The
transparencies are said “to show remarkably fine detail when compared
with an ordinary blueprint.” 347 George Brown, in his treatise on Ferric and
Heliographic Processes, also gives detailed instructions for making
cyanotype transparencies on glass and translucent paper. He prefers
iodine cyanide to Farmer’s reducer for removing old silver images from
spoiled lantern plates. He also provides a recipe for making
transparencies using collodion, instead of gelatin, as the binder vehicle
for the iron sensitizer.348 The problem of coating glass to make it
receptive to aqueous sensitizers has recently been addressed by Bill
Winkler using some modern chemistry – an amino-alkoxysilane to bind to
the glass surface, and mediate a coating of PVA and organic hydrophilic
colloid such as starch which can absorb the sensitizer.349 Winkler uses
the method for platinum-palladium printing, but it should be adaptable
to cyanotype, provided it is non-alkaline.
If a flexible substrate (plastic film) is acceptable, rather than a rigid
one (glass), then the transparency films intended for making digital
negatives (e.g. those from Pictorico or PermaJet) will be found to take
aqueous iron-based sensitizer solutions very well, and the zeolyte
coating holds the image nanoparticles without 'bleeding'. Diapositive
102
contact prints have been made by the author on Pictorico OHP

transparency film coated with new cyanotype sensitizer.
Tonalities are very good, the only problem is in laying down the coat
evenly: the usual glass rod coater is unsuitable, suffering intolerable
friction on the ceramic-treated surface of the film. A squeegee or foam
brush is better, and a straight Teflon (PTFE) rod makes an acceptable
coating instrument for Pictorico - smooth and clean with no friction,
although its hydrophobic nature is a disadvantage. 350
Fig. 4.4 Mike Ware cyanotype diapositive coated on Pictorico film
There is one consequence of using plastic film as a substrate for

alternative printing whose possible importance remains to be tested: the
plastic film is presumably much more dimensionally stable than cellulose
paper under wet processing - so it might prove a useful substrate for
those who favour multiple coating and developing for their image-
making, as in 3-colour processes. If it works for new cyanotype, there is
no reason why it should not also work for platinum, palladium and other
siderotype processes. It also has the advantage that this substrate can be
used 'out of the box' with no pretreatment at all, because it does not
contain any alkaline buffer.
103
4.6 Cyanotype negatives made in camera

Of all the incident light scattered by a photographic subject, only a small
fraction is captured by a camera lens and then projected onto the
photographically sensitized surface. It can be calculated by geometrical
optics that this fraction of captured light produces an intensity at the
focal plane in the order of one hundredth to one thousandth that of the
light incident upon the subject, depending on the lens aperture; it then
follows, assuming that the law of reciprocity holds good, that an in-
camera exposure will take about one hundred to one thousand times
longer than the direct contact exposure to the sun’s rays needed for the
same sensitized surface, see Appendix III.2.
The necessary in-camera exposure time is proportional to the square
of the lens 'aperture' (f/stop number) - this will not surprise anyone who
understands the physical principles of photography. The light intensity
for in-camera exposure is given by the Standard Camera Image
Illumination Equation, also known as the Jones-Condit Equation, which
connects subject luminance with image forming illumination at the
camera focal plane, details of which can be found in standard texts.351 An
estimate of the ratio of exposure times, in-camera : contact exposure,
can be calculated by rearranging the Jones-Condit equation and entering
plausible numerical values for various camera and lens parameters, to get
the constant of proportionality. The desired ratio is then found to be
given approximately by 5 x the square of the 'aperture' (or f/stop).
Here is a table for this ratio:
f/stop (A) 2.8 4 5.6 8 11 16 etc
Ratio (5A2) 40 80 160 320 640 1280 etc.
(This is for a lens focussed at infinity. If focussed for 1:1 reproduction,
multiply the ratios by 4. NB this is only 'order-of-magnitude' calculation
- maybe good to a factor of two or three either way.)
In-camera cyanotype photographs as described by King in §4.4.5,
therefore require exposures in the order of hours, and are consequently
more difficult to make and more limited in application than photograms
or prints by direct contact.352 Moreover, cyanotype negatives can only
produce very dingy positive images when printed with the alternative
printing processes, which are only sensitive to UVA: the lack of contrast is
due to the fact that the negative is blue, and therefore has low optical
density in the operative region of the spectrum. The tonal rendering is
also unbalanced, since the spectral response of the original sensitizer is
104
almost entirely confined to the ultraviolet and blue wavelengths, so

objects like green foliage do not register in the negative, see Fig. 4.5.
An alternative to trying to print such cyanotype negatives was devised
in 2002 by Dr. John Beaver who used the digital manipulation of scanning
the cyanotype negative and then inverting it tonally in Photoshop, to
produce the ‘mysterious, primitive’ images that can be seen on his web
pages.353 This procedure was later adopted by J.E. Piper, with the results
that can be seen on the Alternative Photography website.354
Fig. 4.5 Mike Ware, New Cyanotype negative 6x9 cm on Pictorico

made with an in-camera exposure of 5 hours at f/3.5
The question is sometimes raised, especially by newcomers to

alternative processes, whether it is possible to make cyanotype negatives
in a pinhole camera. The feasibility of this is easily assessed by
estimating a rough figure for the duration of the exposure, assuming the
most favourable conditions. The aperture, A, or "f/stop number" of a
pinhole camera is the ratio of the “focal length”, f, (the distance from
pinhole to film) to the diameter, d, of the pinhole: A = f/d. For most
pinhole cameras, values of A will fall in the region of 100 to 1000; a value
of 200 will be assumed here, for convenience in the arithmetic.
It is further assumed that the New Cyanotype sensitizer (§7.4) would
be used because it is 4 to 8 times faster than the Traditional formula. It
has been found that a camera image can just be obtained on a New
Cyanotype sensitizer at a lens aperture of A = 4 ("f/4") with a sunny
exposure of an hour as minimum – whereas the traditional formula takes
a day. Exposure is proportional to A2, so, for an aperture of A = 200 and
assuming the Bunsen-Roscoe law of reciprocity holds (§3.6), the
exposure duration would have to be at least (200/4)2 = 2500 hours.
Given about 8 hours of sun per day in an ideal sunny climate, this
105
amounts to an exposure of about one year. But that is to ignore the

phenomenon of reciprocity failure, which would, in all likelihood, ensure
that no image at all is ever formed. It also ignores the probability of the
sensitizer becoming chemically fogged. Even assuming the most
favourable scenario then, the conclusion is clear: it won’t work! This
conclusion has recently received some experimental support – of a
negative kind.355 The only image that might be obtained in this way
would be of the sun’s disc itself, which is sufficiently intense to produce,
in a day, a streak across the sensitized paper.
4.7 Printing cyanotypes by projection

A related optical question which is commonly raised is the feasibility of
projection printing onto cyanotype paper using a photographic enlarger.
Obviously conventional enlargers will not do because the light source
contains little or no UV, and even if the tungsten bulb is replaced with a
UVA light source, the thick condenser lenses and enlarger lens will
normally attenuate this considerably. It is possible that cyanotypes may
be made by projection using the kind of ‘solar enlarger’ once employed
commercially for making enlarged Platinotypes, because the two
processes have comparable sensitivity, but this technology is
cumbersome – requiring a heliostat to keep tracking the sun - and the
exposures will be long – in the order of hours. Enlargement onto the
platinotype process – which is comparable in speed to cyanotype - has
been investigated in modern times by Greta Glaser, who found that
exposures of two hours provided only moderate image densities.
Recently it has been demonstrated that a powerful and efficient LED
light source can enable A4 sized enlargements from 35 mm film
negatives to be printed directly onto the more sensitive 'blue layer'
cyanotype paper (§4.12) with surprisingly short exposures of only 20
minutes. Since there can be little or no UV light emitted by the light
source, it seems likely that the absence of an 'internal filter effect' in the
sensitizer due to the use of ferrocyanide, rather than the usual
ferricyanide which absorbs at 420 nm (Appendix III.2), provides a greater
sensitivity for this type of paper to the blue region of the spectrum.356
4.8 Cyanotypes on ceramics

The use of cyanotype on ceramic substrates evokes reminiscences of the
attractive and celebrated blue Delftware. Regrettably this procedure is
limited by the fact that if Prussian blue is fired under a glaze it will
106
decompose thermally around 200°C, with the evolution of poisonous

cyanogen and hydrogen cyanide gases, destroying the blue and leaving
only weak brown 'iron earth' coloured images (iron oxides). However, the
sensitizer is easily imbibed into a bisque-fired ceramic surface, such as
possessed by unglazed tile bodies. This is still very absorbent and needs
its porosity to be restricted with a sizing agent, such as gelatin.
Dutch designer and artist Sarah van Gameren of Studio Glithero in
North London has succeeded in making plant photograms in cyanotype
on tiles and effectively clearing and sealing the image surface for
permanence, possibly with an acrylic varnish.357 The Studio has devised
means to shape their botanical specimens onto the curved surfaces of
earthenware, so they can also offer uniquely decorated blueware vases
and lampshades.358
Fig. 4.6 Blueware vases by Studio Glithero
It is worth warning that cyanotype does not work on plaster surfaces

treated with 'gesso' which contains chalk - calcium carbonate - an
alkaline substance very hostile to cyanotypes, which are destroyed by
alkali (§9.2). The inappropriateness of this substrate for cyanotype is also
the reason why Prussian blue pigment was not used by artists for fresco
painting.
107
4.9 Cyanotypes on fabrics and textiles

One of the earliest commercial uses found for Prussian blue itself was for
dyeing all kinds of traditional fabrics - cotton, linen, silk and wool - in a
range of shades of blue, controlled by the details of the process. In the
industrial practice of the nineteenth-century, two baths were generally
used: the first was a solution of a ferric salt, usually the nitrate or
chloride, which mordanted strongly to the fibres of the fabric, then, after
rinsing, it was followed by an acidic bath of potassium ferrocyanide,
which formed the Prussian blue pigment in situ.359
Of all the possible methods of printing photographic images, the iron-
based processes would seem to be the best-suited for delicate textiles,
because they do not employ a sizing agent or binder medium such as
gelatin or gum, that would stiffen the flexibility and spoil the ‘drape’ of
the fabric. I have already described (§4.4.2) how John Mercer discovered
at an early stage that textiles can be impregnated with cyanotype
sensitizers to receive decorative photographs or photograms. The
process was popular,360 and continues to be used successfully by many
present-day practitioners, such as the British textile artists Wendy Wilson
Barbara Jones and Ruth Brown, the quilter Cathy Corbishley Michel, and
the American artist and designer, Jan Arnow.361 Wilson has found that the
mordanting of sensitizer to cotton and silk fibres appears to be better
with ferrioxalate, rather than with ferricitrate sensitizers.362
In the past, a bath of potassium dichromate was commonly
recommended as part of the wet-processing of fabric cyanotypes, both
for development (reoxidation), and reputedly for improving the light-
fastness of the image. Dichromate could act either by destroying those
potential reducing agents that would assist fading, or possibly by
incorporating itself into the Prussian blue. However, the toxic, allergenic
and carcinogenic hazards now recognised in dichromates make this
practice particularly inadvisable for prints on fabrics, which may come
into contact with human skin. There is at least one well-documented
case-history of an artist working with cyanotype on textiles who suffered
considerable ill-health as a direct consequence of using dichromate
baths.363
The process of making cyanotypes on fabrics to provide photographic
decoration for quilt-covers and cushions became quite popular in North
America in the period 1880-1920, and even built up to a minor industry,
as is evidenced by at least one commercial catalogue, from the Nashua
108
Pillow Company of New Hampshire, which offered for sale several cushion
covers illustrated with cyanotype scenes of natural beauty or national and
patriotic significance. The catalogue proudly declares that:
“There is no happier combination of ornamentation and usefulness
for the Home Beautiful, than a Blue Print Sofa Pillow Top.” 364
Fig. 4.7 Cyanotype quilt 1902.

(Collection of Alstead Historical Society, New Hampshire, USA)
109
The Nashua Pillow Company also offered a service for transferring the
customer’s own negatives into cyanotype prints on prepared fabric. The
assertion is made in this catalogue that, in terms of brilliance and
permanence, the cyanotype prints on cloth were superior to those on
paper, although it is hard to understand the basis of this claim. A fine
specimen of this genre, which is probably one of the Nashua Pillow
Company’s products, can be seen in the collection of the National Gallery
of Canada: a cushion cover bearing twelve cyanotypes of unidentified
coastal landscapes.365 The rediscovery of other specimens of the Nashua
Pillow Company’s products has been recently reported. The company also
offered its sensitized blueprint cloth for sale in 1900, under the
registered trade mark of ‘Silkdown’, at 75 cents per square yard.
The tradition of fabric printing in cyanotype was continued in the USA
by Barbara Hewitt, a specialist supplier of sensitized textiles who founded
with her husband, John Basye, the Californian company of Blueprints-
Printables in 1983, and published an attractive craft handbook on the
subject of Blueprints on Fabric.366 On their retirement, the company was
carried on by Linda Stemer of Vashon Island WA, as Blueprints on Fabric
and it continues to supply this market in the USA today.367 In the UK,
Ruth Brown was also inspired by Barbara Hewitt’s artistic enterprise with
textiles, and founded her own cyanotype fabric printing business, Stone
Creek Textiles, at Sunk Island, Yorkshire,368 publishing her own useful
contemporary handbook on the subject.369 The history of the application
of cyanotype to fabrics has recently been reviewed in the Journal of the
Surface Design Association.370 The modern use of cyanotype for
illustrated quilt-making has been reviewed by Dr. Cathy Corbishley
Michel at a recent seminar of the British Quilt Study Group.371 In addition
to historic examples, and her own work on depicting the historical record
photographs of polar exploration, Michel highlights the work of several
contemporay quilters and textile artists using the medium, including Tafi
Brown and Barbara Hewitt in the USA, and Ruth Brown, Pauline Burbidge,
Linda Barlow, and Dorothy Stapleton in the UK.
4.10 Gel lithography: ‘true-to-scale’ process

The invention of this photomechanical process for offset duplication of
plans and drawings has been attributed to A. Tellkampf and A. Traube,
who obtained a German patent in 1905.372 However, Nadeau in his
Encyclopedia cites evidence that the idea was conceived by Félix J. Dorel
of England in 1900 who licensed the secret process in 1904,373 and later
110
re-patented it in the USA.374 The explanation for its alternative name ‘the
true-to-scale process’ is that no wet-baths were involved in processing
the reprographic image, so there was no dimensional change in the paper
substrate, and copies of exactly the original size were obtained - a
characteristic of great importance for the accurate reproduction of scale
drawings, without the necessity to annotate them with dimensions.
Shrinkage of prints during wet processing had previously been one of the
chief difficulties besetting the photo-printer.
Gel lithography was also variously known in its hey-day by several
proprietory names: Lithoprint, Ferro-gelatine, Fotol, Ordoverax,
Velograph or Fulgur printing. The method took an over-exposed, but
unprocessed negative blueprint image as its source, which was lightly
squeegeed into contact with a matrix of moist gelatine, known as a
“graph”, containing a ferrous salt. Diffusion of excess potassium
ferricyanide out of the lightly-exposed and unexposed regions of the
cyanotype (the image shadows) formed Prussian blue in the gelatin
matrix, with the effect of hardening it locally (see §4.1.1), where it
became receptive to a greasy lithographers’ ink, which was still repelled
by the moisture in the unhardened regions of the gelatin. After a minute
or so, the cyanotype was peeled off and the jelly surface inked with a
roller. About 25 positive copies of the original image could be ‘pulled’,
using little pressure, from the jelly, which was re-inked between each.375
There are a number of fine points of detail to observe in formulating the
gelatine mass and the cyanotype paper if best results are to be
obtained.376 This first successful ‘dry copying’ process became the
preferred method for reproducing engineering drawings and it was a
major reprographic process until the 1950s,377 but then rendered
obsolete by the advent of diazo and electrostatic reprographic methods.
As it was originally formulated, the process is only suitable for line
drawings and cannot adequately reproduce the continuous tonal
gradation of a photograph, in the way the collotype process, for example,
can. However, a number of modifications were introduced, using a half-
tone screen to enable pictorial matter to be reproduced by either
positive- or negative-working processes,378 and most recently Roger
Farnham of Glasgow has created very effective halftone gel-lithographic
prints by adding a grain screen to the image in Photoshop.379
A similar principle of colloid hardening by the diffusion of ferric salts
underlies the ‘Iron-Ozobrome Process’, which also takes a cyanotype as
its starting point.380 This is first hydrolysed by a dilute sodium hydroxide
111
bath to ferric hydroxide, then squeegeed into contact with a carbon

tissue which has been immersed in a bath of dilute sulphuric acid and
formaldehyde. Over a period of about an hour the ferric ions in the
cyanotype are rendered soluble by the acid and diffuse into and harden
the gelatin of the carbon tissue in the shadow regions of the image, so
providing a positive-working process. After separation, the carbon tissue
can be developed in the usual way. Another variation is to coat the ferric
hydroxide image with pigmented gelatin, and when dry, to immerse in
the acid/formaldehyde bath, then develop with warm water.
4.11 Commercial blueprint papers

The great majority of commercial blueprint papers were made for plan-
copying - i.e. ‘line’ work - which naturally calls for contrasty
characteristics rather different from those desirable in papers intended
for continuous-tone pictorial purposes. The reader is referred to §5.9 for
an account of the high quality cyanotype papers that were used for
illustrative and documentary picture-making ca, 1894, notably from the
A.S. Aloe Company. Commercial blueprint papers intended for plan-
copying will now be considered only in outline, but insofar as they did
find some pictorial use, it will be appropriate to summarise the history of
their development.381 Actual samples of the original “Reliance” brand of
‘blue process papers’ manufactured and marketed by the Schindler
Brothers of New York, dating from 1891, have recently been rediscovered
by Adrienne Lundgren and Katherine Mintie in the archives of the Library
of Congress, providing information on paper weights and formulae.382
112
The paramount considerations for blueprint copying were expediency

in exposure and processing, high contrast, and good shelf-life. The
properties of an ideal copying paper were described as follows:383
1. It should be as sensitive as possible, being fast enough to print by
electric arc light, or mercury vapour lamp, within a few seconds.
2. The contrast should be high, ‘line’ images with an intense blue,
and whites that clear rapidly in the wet processing, which was
usually carried out in a continuously-operating machine.
3. The blue colour should retain its quality even when over-exposed.
4. There must be no ‘bleeding’ or running of colour from the exposed
portions into the unexposed white lines, even when the paper is
over-exposed.
5. The paper should remain stable on long storage, for two or three
months, without the unexposed portions developing blue pigment.
In practice, using the customary formulations, these requirements
proved to be mutually conflicting. The ferricyanide component absorbs
blue and ultra-violet light strongly, but contributes little itself to the
light-sensitivity, and therefore acts as an ‘internal filter’, attenuating the
printing light to no good effect. In order to achieve high speed (criterion
1) the ferricyanide concentration must be kept low, with the result that
the maximum intensity of blue (criterion 2) is limited. Necessarily, it
follows that the light-sensitive component, ferricitrate or ferrioxalate, will
be present in great excess over the ferricyanide; with the undesirable
consequence that even slight over-exposure will generate excess ferrous
iron, which washes over into the unexposed portions causing ‘bleeding’
(criterion 4), and in the exposed regions it adversely affects the quality of
the colour, giving it a greyish cast, called ‘burning out’ (criterion 3). In
order to improve the keeping and washing properties of the paper
(criterion 5), substances such as alkali metal oxalates, citrates and
phosphates were added to the sensitizer mixture, but their presence
diminished the sensitivity (criterion 1) and degraded the intensity and
quality of the colour (criteria 2 and 3).
As a result, the sensitizers employed for early copying papers
embodied a compromise between these conflicting requirements; they
were only capable of relatively low maximum density - just a mid-blue
sufficing for plan blueprints. In view of this lack of density and the high
contrast, photography texts often warned the amateur against the use of
commercial copying paper for pictorial purposes.
113
The formulations used by early manufacturers of blueprint paper were

generally withheld as commercial secrets, and even today they are
reluctant to release explicit recipes. Fortunately, there are at least two
published accounts of early formulae that are readily accessible: by
Robert E. Kwech, a chemist of the Huey Blue Printing Company of
Chicago,384 and by K Willcock.385
The paper employed for coating had a high proportion of rag stock,
usually between 50% and 100%, and was surface-sized with ‘animal glue’
or gelatin, but not highly calendered. Kwech describes manufacturers of
the time (1924) adopting four varieties of blue-printing formulae,
depending on the speed of the final product, which he designates as
‘Regular’, ‘X’, ‘XX’, and ‘XXX’, requiring relative exposures of 4 minutes,
2 minutes, 45 seconds and 15 seconds to bright sunlight. The sensitizers
for these are summarised in Table 4.3 which gives both the proportions
of the various stock solutions that were taken, and the resulting final
concentrations (as % w/v) of the ingredients in the mixed sensitizer.
Parts by volume (final % w/v) Paper Grade

Solute concentration Regular X XX XXX
Ammonium ferric citrate 26% 4 parts 4 parts - 3 parts
(16) (11.6) (10.4)
Ammonium ferric citrate 16% - - 4 parts -
(6.4)
Sodium ferric oxalate 10% - 3 parts 4 parts 1 part
(3.3) (4) (1.3)
Ammonium ferric oxalate 10% - - - 1 part
(1.3)
Ferric chloride 7% - - 0.5part 1 part
(0.35) (0.93)
Citric acid 10% - 0.5part - -
(0.6)
Oxalic acid 10% - - 0.5part 1 part
(0.5) (1.3)
Potassium ferricyanide 16% 2.5parts 1.5part 1 part 0.5part
(6.2) (2.7) (1.6) (1.1)
Exposure to bright sunlight 4 mins 2 mins 45 secs 15 secs
Table 4.3 Commercial blueprint sensitizers - Kwech’s formulae
114
The ‘Regular’ solution was claimed to be Herschel’s, and deemed most

satisfactory for photographic purposes. The faster papers were intended
for continuous printing in arc-light machines, and were suitable only for
line subjects. The recipes for these tend to be more dilute, (the
concentration of ferricyanide has to be very low - about 1 to 2% - to
minimise the internal filter effect) and to contain some ferric ammonium
oxalate (which is more light-sensitive than the citrate), and even some
ferric chloride.
The complexity of these recipes is almost Byzantine, and it is hard to
credit that the formulations have actually been optimized by testing all
the variables. The use of citric acid as a ‘clearing agent’ will be noted; the
implications of this for image fading will be discussed later. In 1935, K
Willcock published three recipes for blueprint sensitizer solutions suitable
for a commercial coating machine: ‘Herschel’s’, a ‘medium-rapid’, and an
‘ultra-rapid’ formula; these were reprinted by Neblette in the 4th edition
of his treatise in 1942.386 They differ in some respects from Kwech’s
formulae, in containing gum Arabic, and may be compared in Table 4.4
where the concentrations have been converted to the percentages of the
components present in the final mixed solutions.
The faster the blueprint paper, the worse were its keeping properties.
There is an extensive patent literature, claiming a variety of
improvements to blueprint papers, which has been summarised by Kosar,
whose monograph should be consulted for a full list of references.387 The
criteria for such improvements were usually dictated by commercial
factors, such as the shelf-life of the paper, speed of exposure, or
expediency of wet-processing, rather than by consideration of its
pictorial qualities.
Component Herschel Rapid Ultra-rapid

Ammonium ferric citrate 10 7.5 4.4
Ammonium ferric oxalate - - 2.3
Sodium ferric oxalate - 2.5 3.8
Ferric oxalate - 0.16 0.6
Gum Arabic 1.9 1.9 1.9
Potassium ferricyanide 5 2.5 2.3
Table 4.4 Commercial blueprint sensitizers - Willcock’s formulae
(Figures are % w/v in final mixed sensitizer)
115
Processing of commercial papers usually entailed ‘development’ in a

dilute (ca. 7%) dichromate bath rather than just pure water,388 to re-
oxidise any solarized Prussian white rapidly back to Prussian blue, and
possibly to improve the light-fastness of the image (see Chapter 7). A
preferable oxidant, which was marketed under a proprietory name as a
blueprint ‘developer’, was urea peroxide;389 it has none of the toxic
disadvantages of dichromate, and in solid form is more convenient to
distribute than hydrogen peroxide solution, which may also be used, but
there are serious safety reservations about the storage of such organic
peroxides, which present an explosion hazard.
4.12 Type B blueprint papers-Blue Layer

By 1930, the patent specifications then current imply that ammonium
ferric citrate had been displaced almost entirely from the commercial
formulae by ammonium ferric oxalate, in the interests of faster printing.
Around the mid-1930s, however, the manufacture of blueprint paper saw
the introduction of a significantly different way of formulating the
sensitizers, involving the addition of potassium ferrocyanide. This reacts
with the ferrous iron generated by light to yield Prussian white, which can
subsequently be readily oxidized to Prussian blue by the ‘developer’ in
the wet processing sequence. For further discussion of the chemistry of
this process see Appendix III.
It was noted earlier (§2.9, §4.1.4) that potassium ferrocyanide had
been first employed by Herschel (with ammonium ferric citrate), in the
hope that it would furnish a positive-working process. Superficially this
appeared to be the case, because Herschel’s chemicals were of low
purity, and mixing them resulted in the formation of a significant amount
of Prussian blue, which was at first bleached by light to give a positive
image. However, the major portion of the reactants still constituted a
negative-working sensitizer producing Prussian white which oxidised to
Prussian blue, so the exposure overall presented the rather puzzling
observation of an image, initially positive, but gradually fading and being
replaced by a more permanent negative one.
In modern times, the use of potassium ferrocyanide was first
suggested by Kwech in 1931 as a ‘pre-development’ bath which followed
the exposure, in order to eliminate ‘bleeding’.390 Later it was found that
potassium ferrocyanide could be successfully incorporated into the
sensitizer itself.391 The replacement of some or all of the potassium
ferricyanide by ferrocyanide in the customary formulation brought
116
benefits in printing speed and the suppression of ‘bleeding’, for reasons

that are explained in Appendix III, but it also imparted a blue colour to
the coating, which tended to leave a stain of Prussian blue in the
unexposed highlight regions after processing, unless suitable agents
were added to improve the ‘wash-out’.392 One such agent, whose use was
patented by Jahoda in 1950, is sodium hexametaphosphate, and the
following is one of his formulations which made use of it.393
In 120 cc of water at room temperature, dissolve, in order:
2.5 g sodium hexametaphosphate; 4.5 g potassium oxalate;
20 g ammonium ferric oxalate; 5 g potassium ferrocyanide.
This provides a ‘B-Type’ layer of fast printing speed. Although the
sensitizer solution as made appears dark blue, the concentration of
Prussian blue is very low, and when spread on paper it leaves only a sky-
blue layer. However, the presence of potassium ions in this sensitizer
tends to cause crystallization in the coating even at these relatively low
concentrations, and substitution of ammonium salts is recommended.
So, by the 1940s, the formulae for commercial blueprint papers had
diverged into two main families, called ‘yellow layer’ and ‘blue layer’
papers, whose characteristics are summarized in Table 4.5.394
Property Yellow Layer paper Blue layer papers
Components Ferrioxalate and Ferrioxalate + Ferrocyanide B
Ferricyanide Ferrooxalate + Ferricyanide C
Types A B, C, mixed AB, AC, BC
Colour Yellow Blue (B, C, BC)
(unexposed) Green (AB, AC)
Speed Variable to High Very High (B, C)
Variable (AB, AC)
Contrast Very High Medium to Low
No latitude Wider latitude
Max Density Low to medium High
Bleeding Susceptible Absent
Burnout Susceptible Absent
Washout Clear easily Problematic
Good whites Tend to blue highlights
Purpose Line work Mixed subjects, Pictorial
Printout Negative, reversing Positive, changing to negative
shadow exposures in processing
Table 4.5 Characteristics of two types of commercial blueprint paper
117
The ‘blue layer’ types of paper were favoured for copying pencil and
ink originals, or thick and thin lines, to register which a lower contrast
was desirable. A paper of this type, the Huey Company’s No. 347-6, was
adopted by Heron as a medium suitable for printing continuous tone
photographic negatives.395 It provided an excellent rich blue with a very
high Dmax and a well-separated tonal scale similar to a Grade 1 silver-
gelatin paper.396
Commercial manufacture of blueprint paper ceased in the UK in the
1960s, but production continued in the USA, where there had been much
more widespread investment in commercial processing machines. In
1964, the Andrews Paper and Chemical Company, the principal suppliers
of the raw materials and chemicals to the coating industry, listed 30 US
companies as manufacturing blueprint paper.397 But by the early 1970s
the annual output figures had been showing a steady decline for several
years, and 1972 was the last year that blueprint paper featured separately
in the annual statistics of the reprographic industry products in the USA,
at which time there were still 18 firms coating and marketing the
product.398 Surprisingly, a relatively small amount continues to be
manufactured in the USA today,399 and there is at least one retail supplier
of ready-made blueprint paper and sensitized fabrics for recreational and
educational purposes.400
4.13 Type AB blueprint paper

The type AB sensitizer recipe has been recently reviewed by
contemporary practitioner David Knierim:401 his criteria for a desirable
cyanotype paper were speed and good storage characteristics without any
toxic dichromate added, so making it suitable for popular use by
children, e.g. on summer camp activities. For this type AB sensitizer
Knierim used three stock solutions, which are described below with his
permission. He believes that at least the first two could probably be
combined and stored, and possibly all three; shelf life of the coated paper
has not yet been fully tested experimentally, but is at least two weeks.
Solution 1 (contains everything except the iron salts):

In 91 ml of warm water, dissolve in this order:
3.0 g sodium hexametaphosphate
2.4 g tetrasodium EDTA (0.06M)
2.8 g ammonium oxalate monohydrate (0.20M)
0.5 g oxalic acid dihydrate (0.04M)
118
Solution 2:
In 34 ml water dissolve 26 g ammonium ferric oxalate trihydrate
Solution 3:
In 48 ml water disssolve:
10.1 g potassium ferrocyanide trihydrate (0.4M)
2.1 g potassium ferricyanide (0.1M)
The solutions are mixed in this order:

6.8 ml solution 1
1.2 ml solution 2
2.0 ml solution 3
10 ml of the final sensitizer will coat about 8 sheets of 10”x8” paper
using a glass rod or tube as spreader.
119
5 Studies in Blue 1842-2013

“Quiénes gritaron de alegria
Cuando nació el color azul?”
(“Who shouted with glee
when the colour blue was born?”)
Pablo Neruda (1904-1973)
Book of Quotations
The first cyanotypes in 1842 were not derived from camera negatives in
any way – they were Herschel’s cyanotype copies, sun-printed from paper
prints of steel engravings. The next cyanotype oeuvre historically was the
archive of botanical photograms by Anna Atkins. It was not until the
1870s that cyanotype began to find employment for printing images that
were fully photographic in origin, in the sense that they were made from
silver negatives that had been obtained in a camera. This chapter will
trace an outline history of the documentary and artistic uses of the
process, from its beginnings up to the present day.
5.1 Herschel’s use of engravings

Sir John Herschel produced very few photographs compared with his
colleague in photographic innovation, Henry Talbot, and the images that
Herschel did make were obtained by contact-printing engravings rather
than silver-based camera negatives.402 The latter were still quite
troublesome to prepare in 1842, notwithstanding the great advance in
speed represented by Talbot’s recently patented Calotype process.403
Most (and possibly all) of the printed pictures that Herschel pressed into
service, commencing in January 1839, were small, high-quality steel
engraving prints supplied by one of his brothers-in-law, Peter Stewart,
who was a printer.404 The craft of engraving on hardened steel plates was
of relatively recent invention, having been patented in 1819. It rapidly
gained commercial success because the long life of the steel plate meant
that large editions of 10,000 or more could be pulled from it, unlike
engravings on copper, which were limited to a hundred or two only,
before the plate became too worn. Some artists responded with
indignation at this extent of the proliferation of their work, feeling that it
was cheapened thereby.405
From the very beginnings of his photographic researches, Herschel
had always displayed a preference for employing these ready-made
120
positive images on paper; as early as 30 January 1839, his notebook

records his first accomplishment of a ‘transfer from a print’ by a silver
halide process.406 But he immediately recognised the two disadvantages
of this procedure: it gives rise to images that are always spatially
reversed, because the image transfer has to be formed face-to-face; the
picture is usually tonally reversed too, because the majority of useful
photographic processes are intrinsically negative-working.407 Herschel
was thus obliged to carry out a second derivative printing, which he
called a ‘re-transfer’ or ‘re-reversal’, in order to restore his reversed
negative to a correct positive image.408 It might be noted, in passing, that
Herschel was the first to apply the words negative and positive in the
photographic context, presumably because his mathematically-trained
mind immediately saw that their phenomenology parallels the behaviour
of the integers -1 and +1, respectively, under the operation of
multiplication.409
Herschel expressed a particular concern to achieve a high degree of
definition and translucency in his ‘first transfers’ (which today we would
call inter-negatives) so that positive prints made from them should not
be too deficient in detail. In his unpublished 1839 paper he observed:
“To operate a second transfer, or, by a double inversion to
reproduce the original effect, is a matter of infinitely greater
difficulty, and one in which I have only recently ascertained the
cause of former failures and the remedy to be applied. In
consequence, the specimens produced will be found somewhat
inferior in points of sharpness, and are by no means to be taken as
favourable instances.” 410
In his 1840 paper, Herschel set out the criteria that must be observed
for achieving a good result in this respect; foremost among them was the
application of sufficient pressure to ensure extremely close contact of the
sensitized paper with the original, but the quality of the paper itself was
also important, and he found that a double negative, carefully aligned,
resulted in a better print.411
The alternative to this procedure of ‘re-transfer’ was to devise direct
positive-working processes, but these had the disadvantage of producing
a spatially reversed, positive image, unless the chirality were corrected by
printing through the verso of the original. From a paper diaphane this
results in such a severe loss of definition that it was generally
unacceptable. The use of engravings as source images may also have
imposed some limitations on Herschel’s technical assessment of his
121
process’s quality, which he did not fully appreciate in those early days of
photography.412 It is true that ‘line’ originals provide a good test of
spatial resolution but, having effectively only two densities, of ‘black’ and
‘white’, they are not capable of disclosing the ability of a process to
reproduce continuous tonal gradation, because they lack the subtly-
modulated optical densities of a photographic negative. A well-printed
steel engraving provides a suitable source for making striking and
attractive photographic copies, even via processes of very high contrast
with no capability for rendering intermediate tones. This may be one
reason why the cyanotype process, as developed by Herschel, later
acquired a reputation for rather poor tonal separation, and was not highly
regarded by most photographers as a satisfactory pictorial medium.
5.2 Identifying the sources of engravings

Herschel only selected engravings with no printing on the verso as
diaphanes, and he often rendered them more translucent by treatment
with wax, oil, or varnish. Included in the Memoranda is a document
headed List of Engravings, & how rendered Transparent consisting of 24
numbered titles, often abbreviated, of the original pictures; a summary of
this list, together with some additional titles mentioned by Herschel
elsewhere in his Memoranda, is given in Table 5.1. The Herschel archive
at the Museum of the History of Science in Oxford holds most (53) of the
actual engravings that Herschel used as diaphanes; some of these show
clear evidence of having been treated with substances to render them
semi-transparent.
It is regrettable that, when preparing his diaphanes, Herschel
invariably trimmed off the title together with the names of the engraver
and the original artist, which were usually printed below the picture.413
He did not annotate the resulting test prints with their appropriate titles,
which to him would have seemed superfluous, but he did usually inscribe
his ‘prepared paper number’ on the verso. At several places in his
personal Memoranda he refers to pictorial test prints by their subject
titles alone. Clearly, the purposes of photohistorical research would
benefit from the identification of the pictures to which the titles refer; the
assignment of an image to each title would provide a correlation between
the actual specimen prints which have come down to us, and the
Memoranda in which Herschel described their preparation and
subsequent distribution to friends, colleagues and institutions.
122
This assignment is obscured by the rather non-specific, romantic

nature of most of the titles quoted in Herschel’s ‘List of Engravings’, such
as Still in my Teens, Youth & Beauty, and The Countess; none of which,
per se, can be attributed with certainty to any particular image. The key
to discovering the original source, in which many of these engravings
were first published, was provided by just two of Herschel’s titles, which
refer to depictions of identifiable ladies of the time, namely, Mrs. L.
Stanhope and Mrs. Sheridan.
Fig. 5.1 Sir John Herschel, Cyanotype (1842) "Mrs Stanhope"

from an engraving by Charles Rolls (1836): The Honourable Mrs
Leicester Stanhope, published in Friendship's Offering 1836.
(Collection of the Harry Ransom Center, University of Texas, Austin)
123
I am indebted to Dr Sara Stevenson, Chief Curator of Photographs at

the Scottish National Portrait Gallery, for her invaluable help in searching
for and identifying these two engravings. An engraving entitled The
Honourable Mrs. Leicester Stanhope (Fig. 5.1) is listed in the Art Libraries
Association Portrait Index414 as having been first published in the 1836
issue of Friendship’s Offering,415 which was one of the many Victorian
presentation annuals, or illustrated ‘gift books’: anthologies of verses,
stories, and essays.416 The particular raison d’être of these publications
was to provide a vehicle for the new craft of fine steel engraving, and
they became very popular as Christmas presents with the genteel classes
during the second quarter of the nineteenth century.
Mrs. Sheridan proved harder to identify uniquely, because engravings
of several notable ladies of that name are on record. But in searching
among these, Dr Sara Stevenson found that the British Museum Catalogue
of Engraved British Portraits417 lists an engraving of Lady Louisa Henrietta
Wyatt (neé Sheridan) ‘standing, playing the harp’, (Fig. 5.2) and this too
had been published as a plate in Friendship’s Offering for 1838.418
A standing lady harpist happens to be one of Herschel’s best-known,
but hitherto unidentified, test images; he reproduced it extensively in
several processes, including cyanotype.419 The dimensions quoted for the
images agreed to within a millimeter with those of Herschel’s test prints
in the Oxford archive, and, on consulting copies of the annual
Friendship’s Offering for 1836 and 1838,420 I found the original
engravings and Herschel’s cyanotypes to be perfect positive-negative
pairs.
Friendship’s Offering having been identified as the source of two of
Herschel’s diaphanes, the List of Embellishments in every issue of this
annual between 1824 and 1844 was examined for other titles that might
coincide with the titles in the List of Engravings. Of the 30 titles used by
Herschel, 19 were found to correspond with engravings from this
publication, although in a few cases it transpired that Herschel had
invented his own titles, quite different to those published for the image;
this only became apparent, of course, when the images themselves were
identified as described below.
124
Fig. 5.2 Sir John Herschel, Cyanotype (1842) "Miss Sheridan"

from an engraving by H. Cook (1838): Miss Louisa H. Sheridan
(Collection of the Museum of the History of Science, Oxford)
Table 5.1 summarises the titles and sources of these engravings.

Herschel’s numbered List of Engravings and how rendered Transparent
from his Memoranda in HRHRC constitute numbers 1 to 24 in the
following Table, but his details of rendering them transparent have been
omitted. The seven remaining unnumbered titles occur elsewhere in
Herschel’s Memoranda, especially in the ‘Lists of Specimens’ sent to
various people.Titles in parentheses are alternative versions used by
Herschel. Tentative identifications are prefaced with ‘?’
The references refer to the year, plate number, and opposite page of
Friendship’s Offering where the engraving was originally published.
125
When the original engraving diaphane also exists in the Oxford

Herschel archive it is denoted by *, which assists the attribution in spite
of Herschel’s changed titles. It is also significant that the publishers of
Friendship’s Offering, Smith, Elder and Company, of 65 Cornhill, London,
had the same address as Peter Stewart, Herschel’s brother-in-law printer,
who had furnished him with a number of engravings in 1839. Stewart’s
working address is revealed in a note left by Herschel giving directions
for the safe return of his prints via this person.421 Since Peter Stewart
would have been able to supply loose specimens of the engravings, there
is no need for us to suppose that Herschel had descended to mutilating
bound copies of these ‘friendship books’!
All the engravings in Friendships’ Offering from 1824 to 1844 were
also compared for matches to the 53 cropped engravings in the Herschel
archive at Oxford. Of these, 28 were so identified.
Some of the remaining engravings were also used as diaphanes by
Herschel, but were never explicitly referred to by him in the Memoranda;
two have been identified by Simcock422 as originating from a part-work
of 1831 depicting Edinburgh: Modern Athens! Displayed in a series of
Views.423 These were used by Herschel for some of his earliest test prints
that are still extant; one, an engraving of the Edinburgh Royal Exchange
by W. Watkins, is the diaphane for what is believed to be Herschel’s
oldest surviving dated photograph, a hypo-fixed silver print dated 7
February 1839.
It is possible that the remaining Herschel diaphanes may be found in
other gift books published by Smith, Elder and Company, who employed
Peter Stewart’s services. Of the remaining engravings in the Oxford
archive, there are no specimen prints by Herschel known as yet.
126
No Herschel’s Title Published Title Year Pl. P.

1 Wise Men. Rembrandt Halt of the Caravan* ? 1831 11 305
2 Mrs L. Stanhope The Honourable Mrs 1836 3 36
Leicester Stanhope
3 Bridal Morn The Bridal Morn 1837 7 181
4 Still in my teens Still in my Teens 1838 8 217
5 Early Morning Early Morning* 1837 8 217
6 Orpheus unidentified
7 Infant Lyra Lyra* 1830 2
8 Lady with flowers Mine Own* ? 1830 8 181
9 The Countess The Countess* 1836 7 181
10 Royal Prisoner The Royal Prisoner* 1839 9 325
(Royal Captive)
11 Old barns &c. unidentified
12 Mrs. Sheridan Miss Louisa H. Sheridan 1838 4 73
13 Opening the Chest The Co-heiresses 1840 5 157
14 Scene in Italy A Scene in Italy* 1839 8 289
(Italian Landscape)
15 Torcello Torcello 1838 5 121
16 Burghersh Family The Children of Lady 1840 7 252
Burghersh
17 Frescatura The Festival* ? 1836 4 73
18 Youth & Beauty Youth and Beauty 1838 7 181
19 Amazons Basso relievo unidentified
20 Hercules & Lion Basso R unidentified
21 Schooner unidentified
22 Roman Lady [may be ‘Donna Elena’] 1837 3 28
23 Poetry Poesie* ? 1831 13 373
24 Boy & Lyre unidentified
Rosolia Rosolia 1837 8 217
Helena Donna Elena* 1837 3 28
Lady & Lute The Lady of the Palace* 1830
Letter of Bad News Reading the News 1830 57
(Reading the Letter)
Biancha Bianca Vanezzi 1834
Italian Lady The Gondola* ? 1834 11 325
Square-rigged ship unidentified
Table 5.1 Herschel’s titles for diaphanes, and when published
127
5.3 Herschel’s cyanotype tests

Many cyanotypes still survive that were made by Herschel according to
the procedures described in his 1842 Paper and the Memoranda at the
HRHRC. The majority of these specimens is divided between the
collections of the Museum of the History of Science, in Oxford,424 and the
National Museum of Photography, Film & Television, (now the National
Media Museum) in Bradford; two are in the Library of the Royal Society,
London; eleven are in the collection of HRHRC at Texas,425 and isolated
specimens exist in a number of other collections in the USA: the J Paul
Getty Museum, the University of Princeton, and the Smithsonian Institute.
In 1965, the Oxford archive was given prominence by R S Schultze in a
publication setting out the broad structure of the collection and making
some specific points of interest in relation to the several processes that
Herschel had tested in the early 1840s.426 In his paper, Schultze raised
some unanswered questions relating to the cyanotype processes
designated (a) (b) and (c) by Herschel, which can now be resolved by
reference to the Memoranda, which were not available to Schultze at the
time of his research. In 1989, the year which was taken to mark the
sesquicentenary of the announcement of the invention of photography, to
celebrate Herschel’s contribution to this discovery, an exhibition at
Oxford was curated by A V Simcock, whose accompanying monograph
nicely illustrates and elaborates the relationships between the processes
that may be found in this archive.427
Among the HRHRC Memoranda there are several lists of titles of test
specimens, with other details of process, which Herschel had sent to
various interested friends and colleagues, including one addressed to
Julia Margaret Cameron, which contained a number of cyanotypes as may
be seen below in the transcript of Table 5.2. Herschel first met Mrs.
Cameron in 1836 at the Cape of Good Hope, during his stay from 1834 to
1838 for the purposes of astronomical research. It was in a sense
prophetic that he should subsequently send her specimens of the new
photography in 1842, for this was a pursuit that she was destined to take
up herself twenty-one years later, and then with most memorable effect.
She first photographed her mentor Herschel in 1867. In her unfinished
manuscript Annals from my Glass House, Mrs Cameron refers to him as
follows:
“...I photographed my illustrious and revered as well as beloved
friend, Sir John Herschel. He was to me as a Teacher and High
Priest. From my earliest girlhood I had loved and honoured him,
128
and it was after a friendship of 31 years’ duration that the high task
of giving his portrait to the nation was allotted to me. He had
corresponded with me when the art was in its first infancy in the
days of Talbot-type and autotype. I was then residing in Calcutta,
and scientific discoveries sent to that then benighted land were
water to the parched lips of the starved, to say nothing of the
blessing of friendship so faithfully evinced.” 428
Mrs. Cameron’s reference to the ‘autotype’ is an error; the process

was not Herschel’s, nor was it introduced until 1864. She may possibly
have meant amphitype, or even cyanotype.
List of Photographs sent to Mrs. Cameron. Aug 22/42

1 Burghersh Fam + Red Stock 2
2 Royal Prisoner + Do. 1
3 Pompeii + Do. 1
4 Mrs. L. Stanhope + Do. 2
5 In my teens + Papaver Hybridana 2
No. 1216
6 Helena + No. 1182 2
7 Do. + No. 1202. Violet 2
8 Do. Pos + Sparaxis Alk. 1153 3
9 Pompeii Neg – Cyanotype (a) F3/2 2
10 M rs. Sheridan – Do. (a) 1
11 Rosolia + Do. Deep blue vary (b) 1
12 Ships – Do. (a) light blue 2
13 In my teens – Do. (a) dark blue 1
14 Rosolia + Do. (c) before fixing 2
15 In my teens + Do. (c) purple half fixed 2
16 Ships + Guiacum green 2
17 Helena – Chrysotype (thin) 1
18 Burghersh fam – Do. Bwpon 1
19 Mrs. L. Stanhope 725. Mercury 2
20 My Teens 750. Do. 1
21 Scene..... 721. fixed 1/2 by Chrom 2
22 Specimens of Mr Collen’s Talbot’s Calotype
23 portraits
24 Calotype.
Table 5.2. Specimens sent by Herschel to Julia Margaret Cameron 1842
129
The following Table 5.3 is a transcript of a list of specimens, sent by

Herschel to Robert Hunt in 1842, which also contains references to the
cyanotype processes designated as (a), (b), and (c). All three of Herschel’s
cyanotype processes are exemplified here in specimens numbered 7, 8,
9, and 10, and it is interesting to use this information to identify their
nature and thus answer the question raised thirty years ago by
Schultze.429 The fact that one specimen of type (a), no. 7, is listed as a
negative image, is strong evidence that (a) can therefore be identified as
Herschel’s negative-working cyanotype process. If (a) had been a
positive-working process, this image would have to have been made
from a negative of the original engraving - a singularly pointless
exercise.
Photography
Specimens Distributed.
List of Photographs sent to Mr. Hunt Oct 2/42
1 Orpheus + Senecio Splendens.
2 Royal Prisoner + Papaver hybridium
3 Opening the Chest + Red Stock.
4 Rosolia – )
5 Italian Landsc – > Chrysotype
6 Mrs. L. Stanhope – )
7 Mrs. Sheridan – Chrysotype Cyanotype (a)
8 Lady & Lute + Do. (a)
9 Rosolia + Cyanotype (b) deep blue
10 Do. + Cyanotype (c) purple
11 Do. – No. 725. Aug 1
12 Mrs. L. Stanhope – No. 750 Aug 2
13 In my teens + Thermocyanotype
14 Rosolia – 758. Oct 2
on blue wove Post NB
15 Specimen of + Cyanotype paper
Table 5.3 Specimens sent by Herschel to Hunt 1842
The fact that no. 8, which is also of type (a), is listed as a positive,
presents no problem, however, because the latter could well have been
printed from a negative of the original image - what Herschel referred to
as a ‘re-transfer’. In support of this hypothesis, in the Oxford archive
130
there is a high-quality brown reversed negative print of ‘Lady and Lute’

(presumably made by a silver process) which has been rendered highly
translucent for just this purpose by waxing.430 In his exhibition, Simcock
displayed one of these negatives.
Most of Herschel’s specimens are inscribed on the verso, in his own
hand, with his prepared paper number, so the original light-sensitive
coating may be identified with reference to the list of prepared papers in
the Memoranda at HRHRC. It is more difficult, however, to discover how
they were processed after exposure, because one sheet of prepared
paper might suffice to provide ten or twenty smaller ‘squares’ which
could be used for different tests.
For one of the listed specimens sent to Hunt, no. 10, it is possible to
go further and make a unique identification with an existing positive
cyanotype of ‘Rosolia’ held in the collection of the NMPFT.431 This print is
reproduced in Schaaf’s book;432 it bears the number 10 at the top left
hand corner of the recto, which is further annotated at the bottom in ink
in Herschel’s hand: ‘Positive Cyanotype. Aug 17 1842. JFWH.’ On the
verso, is pencilled the prepared paper number ‘724’, which we find from
the Memoranda corresponds to a coating of ammonium ferric citrate. The
chirality of ‘Rosolia’ in specimen number 10 is reversed, i.e. it is the
mirror image of the original engraving,433 so we may infer that it was
derived directly by one printing from the original, and that type (c) was
therefore one of Herschel’s positive-working cyanotype processes. Even
the paper is identified by Herschel’s annotation ‘yell. w. post’ as yellow
wove post. The image colour, however, is not purple, as stated in the list,
but blue. This does not present a serious objection because Herschel
states subsequently that he observed the colour of type (c) to change
from purple to greenish-blue. This does not resolve the difference
between positive-working methods (b) and (c) however. A clue to this
comes from another list, of specimens sent to the Royal Society in
1842,434 of which the following Table 5.4 is a partial transcript. Although
no (a), (b), (c) notation is used here, Herschel does distinguish between
three cyanotype processes: negative working; positive-working ‘half-
fixed, without corrosive sublimate’, and positive-working fixed with
corrosive sublimate.
These specimens still exist intact: among the items in the Herschel
archive at HRHRC are seven sheets of Herschel test prints, six containing
six mounted specimens each, and one seven. Of these 43 varied
experimental tests, some are now completely blank (especially those
131
derived from mercury), but others are obviously surviving cyanotype and
chrysotype images. The specimens are numbered and can be correlated
exactly with Herschel’s lists of specimens sent to the Royal Society in
1842, of which the contents of the fourth, fifth, and sixth sheets are
transcribed in Table 5.4. Examples may be identified among them of
negative-working cyanotypes (Nos 22, 24, 26) and positive working
cyanotypes (Nos 20, 25, 27, 28, 30) as listed in the HRHRC photography
collections database.435
Photography Lists of Specimens distributed

List of 4 additional sheets exhibited to RS. Nov/42
Nos.
19 Mrs Stanhope. –
21 Rosolia – Chrysotype
23 Italian Landscape –
29 Lute Playing Lady +
20 Royal Captive + Cyanotype + process
28 Helena + Let go on and half fixed
- without Corr. Subl.
22 Still in my teens – Negative Cyano-
24 Mrs Stanhope – (type. ACI + F3/2CP.)
25 Rosolia + Cyanotype + process
27 Mrs Stanhope + with corrosive subl. Nov. 1842.
26 Lady & Lute + Retransfer. Negative Cyanotype.
30 Rosolia. + Cyanotype, no corr. subl.
arrested by water (b.w.p.)
31 Thermocyanotype – (Countess) 3rd revival
32 Ital.n Landscape – 758 one of the finest but 3/4 faded.
33 ACI + Pict...[??] + NS. ACI on both sides of paper.-
Mrs Stanhope. very intense.
34 Rosolia. (ACI + N[Hg]) + Pict...[??] + water
made fine [?] cut out.+ O
35 No 758. In my teens best specimen Iridescent - greatly faded
36 Italian Landscape. + Cyanotype very intense,
Water + Perchlor Iron to fix yellow
border.
Table 5.4. Specimens sent by Herschel to the Royal Society in 1842
132
I have attempted to estimate from the data in the Memoranda how

many cyanotypes, in all, may have been made by Herschel. In the case of
the standard negative-working cyanotype there are only three prepared
papers to consider, and this is how Herschel described them in his list:
“780 Aug 10 10 Ammo Citr I + 10F3/2CP no precip greenish paper
20 grs [?] covers 3 squares
784 Aug 11 F3CP20 + AmoCitrI30 + Tinct. Turmeric 10 on yell
wove post. the turmeric proved of no value.
793 Aug 16 (30F3/2CP + 50ACI) exceedingly sensitive when wet
browns very rapidly in O this quantity did 18 squares 90 grs did 20
squares = 4½ per square”
The most likely interpretation of a ‘square’ is that it was of a size

suited to printing Herschel’s smaller diaphanes, and is therefore well
represented by the specimens that are most commonly found in the
archives. These are of fairly consistent dimensions, on average about 120
x 95 mm, bearing images about 100 x 80 mm.
The numerals in Herschel’s notes, transcribed above, must define the
relative quantities of the ingredients - but they raise the question: what
units was he using for his measurements? The only plausible
interpretation is that these numbers denote drops of liquid (or minims in
apothecaries’ measure - see Appendix IV) taken from his stock solutions,
whose concentrations are specified elsewhere. This assumption is
confirmed by the Memoranda entry dated 29 October 1842, where
Herschel refers explicitly to his “method of drops”. With this
interpretation and the knowledge that one minim of water weighs
approximately one grain, or 0.059 grams, we can calculate the ‘coating
weight’ of Herschel’s sensitizers and see if the result is consistent with
experience.436 4½ grains, or minims, per ‘square’ of 120 x 95 mm
corresponds to a specific coating volume of 23 cc/m2, which is
comparable with the specific coating volume used by the author, and
others, for contemporary coatings of plain papers.437 If the three coated
papers cited above are the only ones that Herschel prepared with this
formulation - and there is no indication in his memoranda of any others -
it follows that the total number of negative cyanotype specimens that he
could have made was about 34. His positive-working cyanotype tests may
well have been more numerous than this, but the variability of the
process and its frequent failure to yield a satisfactory image leave much
uncertainty concerning the number of such specimens that we can expect
133
to find. The Herschel archive at the National Media Museum has 16 from
one sheet, which all bear the annotation of ‘prepared paper number’ 724,
which consisted of a coating of ammonium ferric citrate alone. After
exposure, these must have been treated with potassium ferrocyanide
solution, and many of them bear Herschel’s rather cryptic annotation
‘yop’ or possibly ‘yoss’, which could refer to this substance - ‘yellow
prussiate of potash’.
5.4 Cyanotypes of British algae by Anna Atkins

The first application of Herschel’s cyanotype process for an illustrative
purpose was begun in 1843 by a lady botanist called Anna Atkins (1799 -
1871), Fig. 5. The full significance of her contribution to early
photography and the history of book illustration was not brought to light
until 1988, as a result of the research of the American photohistorian Dr
Larry Schaaf, while he was Carnegie Trust Fellow in the Department of Art
History at St. Andrews University, Scotland. Dr. Schaaf was able to curate
an Exhibition in Glasgow, St Andrews, and Edinburgh, following which he
unfolded the history of Anna Atkins in his beautiful book Sun Gardens:
Victorian Photograms by Anna Atkins.438
134
Fig. 5.3 Catalogue of exhibition curated by Dr Larry J. Schaaf
Little has since emerged to add to this comprehensive and beguiling

account of a most important, but previously uncelebrated, woman in the
history of early photographic endeavour. For the benefit of readers who
may not be acquainted with this work, the facts will be briefly
summarised here, but with the proviso that this terse account is no
substitute for Schaaf’s deeply-researched and beautifully illustrated
book.
135
Fig. 5.4 Anna Atkins
Anna Children was born in 1799 in Tonbridge, Kent, the only child of
Hester Anne and John George Children (1777-1852), a distinguished
scientist who was elected to Fellowship of the Royal Society in 1807 and
later became its Secretary. Anna’s mother died within a year of the
childbirth, and perhaps partly to compensate for her loss, the bond
between father and daughter was always close. Their relationship also
encompassed a publishing collaboration when Anna’s considerable talent
for draughtsmanship enabled her to undertake the illustration of her
father’s translation of a seminal work on the taxonomy of seashells.439
Thus Anna was, as Schaaf describes:
“...brought into touch with most of the tight-knit British scientific
community of the early part of the nineteenth century.” 440
In 1825, Anna married John Pelly Atkins, a man of property, who

fostered her interests in botanical collecting and illustration. In 1830 the
couple, who remained childless, moved to Halstead, near Sevenoaks in
Kent, where they were joined in 1840 by Anna’s father, following his
retirement from a lifetime’s career as Keeper of Zoology at the British
Museum. His proximity undoubtedly continued to provide an important
channel of communication between Anna and the scientific world, which
at that time seldom welcomed women wholeheartedly into its ranks.
Children, as a member of the Royal Society’s Committee on Publications,
had been at the centre of things during the announcement of
136
photography, and he had been in the chair at the meeting of the Society
in February 1839 to which Talbot disclosed his method for making
photogenic drawings. From the outset, Anna showed great interest in this
innovation - to the extent that in 1841 she was presented by her father
with an early Ross camera, although no camera photographs by her are
presently known. Anna’s husband was also a friend of Talbot, sharing a
common interest in transport engineering, so she could not have failed to
be aware of the early use of Talbot’s photogenic drawing process for
making botanical photograms, such as that illustrated in facsimile in the
popular periodical The Mirror of 1839.441
In 1840 the Herschel family also moved residence to Kent, taking a
house in Hawkhurst, only about 30 miles away from the Atkins; the
Herschel daughters were friendly towards Anna, whom they regarded as
an aunt. When Sir John’s 1842 paper was published, containing details of
his cyanotype processes, a copy was sent directly to John Children, but it
is likely that Anna, as a frequent visitor to the Herschel household, also
acquired experience of the process by more direct and less formal
means. Anna must have quickly realised that cyanotype offered a method
for botanical illustration by photography which was altogether more
expeditious and inexpensive than Talbot’s silver halide processes; she
was not then to know that it would also prove more permanent.
No evidence has yet come to light to tell us precisely the method or
formula that Anna adopted to make her cyanotypes, but from their
appearance (the colour is usually in the range 21E7-22D7) and the
productivity of her work (as we shall see, her edition of British Algae
alone demanded the hand-printing of over 5000 cyanotypes), it seems
likely that she used Herschel’s rapid, standard negative-working Method
(a). The means to carry out this process would not, for her, have been
hard to come by: her father had constructed several powerful galvanic
batteries for the purposes of electrochemical research, with which he
could have prepared the necessary potassium ferricyanide from readily
available potassium ferrocyanide, using Smee’s electrolytic method.
Ammonium ferric citrate was growing in popularity as a pharmaceutical
preparation by then, and was available at most druggist’s. A suitable fine
paper, Whatman’s Turkey Mill, which had already been used extensively
by Talbot for silver images, was widely marketed as a superior writing
paper; and, of course, the processing of the cyanotypes needed nothing
more than pure water, which the well at Halstead could supply in
abundance.
137
By 1843, Anna had set to work on the production of her monumental

hand-printed publication, British Algae: Cyanotype Impressions, which
she intended as a supplement to the treatise on seaweeds by Hervey,
which was not illustrated.442
Fig. 5.5 Anna Atkins Photographs of British Algae
Anna’s work consists of negative cyanotypes made directly from

diaphanes of the seaweeds themselves, dried and pressed, but usually
unsupported. Each is labelled with its Latin name written in Anna’s hand
on a translucent slip, and the illustrations are preceded by a title page, a
dedication of the work to her father, a list of contents, and a handwritten
prefatory note, which with modest anonymity, is signed ‘A. A.’ Herein she
explained her intention:
“The difficulty of making accurate drawings of objects as minute as

many of the Algae and Confervae, has induced me to avail myself
of Sir John Herschel’s beautiful process of Cyanotype, to obtain
impressions of the plants themselves, which I have much pleasure
in offering to my botanical friends.” 443
138
Anna issued the work as a privately-distributed part-book over the

decade 1843-53. Each of the soft-bound sewn fascicles contained about
12 plates at first, but this number increased considerably with the later
parts, and there were twelve parts issued in all, making an approximate
total of more than 400 plates for the final work, although this number
varies in the known copies, partly due to Anna’s retrospective issues of
‘improved’ prints of certain specimens. It was her intention that the entire
work should be bound in three volumes. The size of the edition is
uncertain, but seems to have been at least 13 in number. By 1985, Schaaf
had identified a probable 12 known copies, in varying states of
completion, and he listed their original provenance, and their current
whereabouts.444
Notable among the recipients of Anna’s part-work was Robert Hunt,
who reported favourably upon this use of the cyanotype process in his
early (1848) review of progress in the new photography:
“They are so exceedingly simple, the results are so certain, the
delineations so perfect, and the general character so interesting,
that they recommend themselves particularly to ladies, and to
those travellers who, although not able to bestow much attention
or time on the subject, desire to obtain accurate representations of
the botany of a district. We have seen specimens of the British
Algae executed by a lady, by the Cyanotype process, that are
remarkable for the extreme fidelity with which even the most
attenuated tendrils of the marine plants are copied.” 445
Hunt’s compliment is double-edged in being tainted with a

patronising implication that this was a process for women and therefore,
dilettantes. Henry Fox Talbot was also on Anna Atkins’ distribution list,
and much later in 1864, he commented approvingly on the permanence
of the process, having by then suffered grievously from the
impermanence of many of his own silver images:
“Moreover, we know that some photographic processes are
permanent; for instance, the cyanotype of Sir J. Herschel, by which
process a lady, some years ago, photographed the entire series of
British sea-weeds, and most kindly and liberally distributed the
copies to persons interested in botany and photography. The whole
of these prints remain unaltered after the lapse of several years.”446
These words, published in The British Journal of Photography, were

read two decades later by the Scottish photohistorian and book collector,
William Lang, who was then President of Glasgow Photographic
Association. Lang was a compulsive book collector – he described it as an
139
"incurable disease" - and was spurred on to seek a copy of this

remarkable botanical work, although he was not even aware of the
identity of its author. After a search lasting several years, he eventually
acquired, in 1888, the complete set bound in two volumes that Anna
Atkins had originally given to Robert Hunt.447 Lang proudly displayed and
lectured upon his acquisition on several occasions,448 although he only
knew the work was by "A.A.", but a curator at the British Museum
identified the "lady" for him as Anna Atkins. It is indeed fortunate that we
have his clear description of it, because this copy disappeared in 1910,
following the auction of Lang’s library, and only re-emerged for re-sale
at Sotheby’s on 10 May 1990, realising £229,250. Lang even made his
own cyanotypes as frontispieces to the Journal of the Philosophical
Society of Glasgow where he published his account.449
Fig. 5.6 Cyanotype frontispiece by William Lang
There remain seven substantially complete copies of British Algae in

British institutions, including the Royal Society, Lacock Abbey, the British
Library, the Linnean Society, the British Museum, the Royal Botanic
Gardens, Edinburgh, and the Art Gallery and Museum, Kelvingrove,
Glasgow. Three other holdings are seriously depleted. Two copies are in
the USA, including Sir John Herschel’s personal copy, which was sold at
140
auction by Sotheby’s of London in 1985 for £48,400, and is now in the

collection of the New York Public Library. The entire set of images in this
item may now be browsed and viewed online, courtesy of the Digital
Collections of the NYPL.450
The exciting discovery of a thirteenth intact copy was made in 1995 by
David Park of Bonham’s auction house, London, who recognised the work
among the uncatalogued contents of the small library of the Natural
History Museum of Eton College.451 This fine copy, having 411 plates,
bound in three volumes, was sold at auction by Bonhams for £130,000 in
1996 and was granted a licence to be exported to the collection of Tel
Aviv Museum, where it now represents one of the great achievements of
early British photography. The attention of the popular media was briefly
attracted by this significant sale of a hitherto unrecorded treasure, so
Anna Atkins’ achievement was finally celebrated in the popular press, 144
years after its completion.452
Anna Atkins distributed more than a dozen copies of the first fascicle
of her British Algae in October of 1843, whereas the first part of Talbot’s
Pencil of Nature, illustrated with silver salted-paper prints, was not
published until June 1844. British Algae therefore takes precedence in the
claim to be the first photographically illustrated book,453 albeit a
privately-published one, and earns Anna Atkins a permanent place of
honour among the pioneers of publishing. Anna’s ‘herbarium’, the
collection of dried, pressed botanical specimens, or exsiccati, that she
had been amassing and using since 1830, containing many of her
diaphanes, was finally presented to the British Museum in 1865.
5.5 Albums by Anna Atkins and Anne Dixon

Upon the death of Anna Atkins’ father in 1852, the bereaved daughter
was much comforted by the companionship of her lifelong friend and
‘almost sister’ Anne Dixon.454 At John Atkins’ request, she came to stay at
Halstead for the summer, in order to offer solace to Anna and help her
sort out her father’s papers for the substantial biography that Anna was
preparing for private publication in early 1853.455 It is likely that, finding
action a remedy for despair, Anne Dixon also assisted Anna in
progressing her printing for British Algae, which was completed later in
1853, just ten years after its inception.
The two women may well have continued cyanotyping together, and
their collaboration seems to have broadened their horizons regarding
subject matter, which first encompassed ferns, then flowers, leaves,
141
feathers, and lace as diaphanes. One of the fruits of their joint endeavour
was an album, nominally authored by Anna, entitled Cyanotypes of British
and Foreign Flowering Plants and Ferns which she presented to Anne
Dixon in 1854 as a token of gratitude for her dear friend’s support
during her mourning the loss of her father. The binding was simply
stamped ‘A.A. to A.D. 1854’. This unique and beautiful work survived
until 1981 when, immediately following its sale at auction, it was broken
by dealers in an act of collaborative philistinism.
In 1853, probably with Anna’s help, Anne Dixon herself completed
two very similar presentation albums entitled Cyanotypes of British and
Foreign Ferns. One was dedicated ‘A.D. to H.D.’; the recipient is known to
have been Henry Dixon, her young nephew, who became a noted
photographer of India. This album is preserved in the collection of the J
Paul Getty Museum, Los Angeles. The other album by Anne Dixon was
previously unrecorded until its existence, in private hands, was first
brought to light in 1988 by John Wilson.456
It carries the dedication ‘A.D. to C.S.A.’ and it is now in the collection
of the National Media Museum (previously the National Museum of
Photography, Film & Television) at Bradford, England. The researches of
Wilson into the history of cyanotype have included a careful comparison
of these two albums.457 He found that both were printed from the same
set of 98 botanical diaphanes, some of which were also used by Anna
Atkins, but in a substantial number of the images the specimens are
reversed or have other differences in positioning between the two
albums. The identity of the recipient bearing the initials ‘C.S.A.’ remains
something of a mystery. Wilson has reviewed the rather slender evidence,
in particular the suggestion put forward by Fred Woodward of the name
of Miss Catherine Allardyce (1802-1886) of Cromarty, a known collector
of algae. Although Wilson has found no reference to her in the
contemporary botanical literature, I have come across an obscure
acknowledgement to her in Johnstone and Croall’s Nature-Printed British
Seaweeds, as one of the contributors of specimens.458
142
Fig. 5.7 Anna Atkins and Anne Dixon, Cyanotypes of British and
Foreign Ferns (1853)
The first volume of this elegant four volume set of nature prints was
reviewed in The Athenaeum in 1859. The very name of Miss C. Allardyce
evidently fired the imagination, if not the heart, of the reviewer, whose
fantasy waxed quite rhapsodical:
143
“The only relief we experience in poring over these densely

technical pages is not a sea breeze, but an occasional name of a
fair spinster who has contributed some weed to the author. We
cannot restrain our imagination from picturing “Miss C. Alardyce,”
who has found Bonnemaisonia asparagoides at Moray Frith [sic] (p.
77), ... O, for a weed-walk and a weed-talk this very day with
Caroline Alardyce, or any and all of the spinster phycologists! ... But
why not immortalize the spinsters, even though they should
change their names, by affixing them to the weeds? ... how much
more euphonious would be Melobesia Alardycii, or, sweeter still,
Alardycia Virginea!” 459
This whimsical eulogy by the anonymous reviewer (discovered to be
John R. Leifchild)460 drew forth a reply from Johnstone and Croall in the
preface to their fourth volume. It pays tribute, if rather patronisingly, to
the important role that many gentlewomen played at the time in
collecting and cataloguing botanical specimens – it was one of the few
scientific pursuits available to them that was not deemed unladylike. A ‘C.
Allardice’ [sic] is acknowledged again in the second volume of Johnstone
and Croall’s work.
In 1861 Anne Dixon created a third album of cyanotypes which she
again dedicated to her nephew, now Captain, Henry Dixon.461 This
beautiful compilation is anonymous and simply titled Cyanotypes in gold
on the spine. A flyleaf bears the handwritten inscription in ink: ‘Capt
Henry Dixon with his Aunt Henry’s Kind love - Ferring Vicarage June 21
1861’. The album contains 74 cyanotypes of a most varied kind: ferns,
grasses, algae, mosses, leaves, feathers, fabrics and needlework, chosen
rather for their aesthetic qualities, than for any systematic purpose. The
handwriting of both Anne Dixon and Anna Atkins is apparent on the
labels, and some of the specimens have been used in other albums by the
two ladies; the paper used, where watermarks can be discerned, is J.
Whatman 1845.
Anne Dixon died on 20 March 1864, and Anna Atkins died at Halstead
Place on 9 June 1871. After the death of John Pelly Atkins in 1872, the
house later became a school, but this was demolished in 1954 to make
way for a new building. In 1997 Anna Atkins was finally honoured for her
pioneering contribution to photography and book illustration by the
unveiling at Halstead Place of an appropriately blue commemorative
plaque, one of ten in the Photographic Heritage Programme promoted by
the Royal Photographic Society and Olympus Cameras.462
144
Anna Atkins achievement has recently been evaluated in Jennifer

Howard’s thesis on A Victorian woman’s experiments in botanical
illustration:
“This dissertation discusses how Atkins‟ work fits into the mid-
nineteenth century context of women, botany, and illustration, and
provides justification for a historical bibliographical study of
Photographs of British Algae. A descriptive bibliographical
approach is taken in the second half of the dissertation, which uses
a sample analysis of twenty-four plates taken from six copies of
Photographs of British Algae to discuss how the cyanotype process
functions as a printing medium. It also discusses how Atkins used
the reprographic capabilities of the process, and how the book
functions as a botanical reference book.” 463
Concerning this fruitful episode in early photography, it is fitting that

the last word should come from Schaaf:
“Anna Atkins and her circle had been the primary users of
Herschel’s cyanotype process. With her own hands she crafted
thousands of lovely and significant photograms of natural objects,
and in an imaginative and brilliant undertaking, she had exploited
the most practical technology of her day to produce the first
photographic book. Anna Atkins’ legacy embodies all that was
promising about the invention of photography. Her cyanotypes not
only provide documentation of scientific interest, but in an
intriguing duality, they also remain some of the most creative and
beautiful artefacts from the earliest days of photography.” 464
5.6 An anonymous cyanotype album of 1843

In 1993 the National Museum of Photography, Film & Television,
Bradford, (now the National Media Museum) acquired a very early and
unusual album entitled Dried Ferns, containing both cyanotypes and the
botanical specimens (exsiccati) that they had been printed from. The
author of this work is unknown. It was bought at auction, and no
information was available as to its provenance. This is a red morocco-
bound, quarto album of mounted specimens of exsiccati (pressed and
dried ferns) which, as is evident from their labels, have been gathered
from extremely diverse locations: Borneo, East Indies, Guernsey,
Philippine Islands, Britain, Java, Tropical America, Chile, New Zealand, etc.
On the pages following, the volume also contains 16 cyanotypes of some
of these dried ferns, of dimensions 9 x 7.5 inches (one is 9 x 15 inches,
and folded) made on chain-laid paper, watermarked T. Edmonds 1821.
145
Fig. 5.8 1843 Album of pressed ferns and their cyanotypes

(collection of the National Media Museum, Bradford)
The versos of all the cyanotypes carry rather faint pencil annotations,
most of them legibly inscribed as ‘Herschel formula cyanotype
impression’ and variously dated during March and April 1843. Many of
the annotations also specify the exposure, which was generally between
one and two hours, and two notes refer to ‘very hot sun’ and ‘bright sun’.
Such references to the weather in the months of March and April are
hardly consistent with the meteorological record for typical locations in
the British Isles (although, on occasion the record described conditions as
‘fine’),465 which suggests that these cyanotypes may not have been
printed in Europe - or even in the Northern Hemisphere. It may also be
significant that an analytical investigation of the album by X-ray
fluorescence spectrometry has revealed that its pages are heavily
contaminated with arsenic,466 which was possibly applied in the form of
arsenious oxide, as an insecticide used especially for the protection of
organic specimens in tropical environments.
The handwriting on the versos of these cyanotypes is not that of Anna
Atkins or Anne Dixon; it even appears to differ from the handwriting on
the labels of the botanical specimens in the same album.467 The colour of
the images is a consistent, slightly greenish-blue (around 23D8 -
‘sapphire blue’), unlike the cyanotypes of Anna Atkins (which are typically
21D7, and which correspond with that of cyanotypes made by Herschel’s
standard method (a)). The greenish-blue colour 23D8 is obtained when
146
potassium ferricyanide alone is used as the sensitizer in the ‘proto-

cyanotype’ (see §4.1.1), moreover the relatively long exposures of one or
two hours to ‘bright sun’ are consistent with the use of this much less
light-sensitive substance alone. The infrared spectrum of the blue
pigment is close to that of the proto-cyanotype process (see Appendix
II.7) rather than the usual Prussian blue of Herschel’s standard sensitizer
using ammonium ferric citrate and potassium ferricyanide. Exposure of
the standard sensitizer for hours would also have resulted in massive
over-exposure, to the detriment of the image quality. The highlights of
the cyanotypes in this album are clear white, again characteristic of the
simple ferricyanide method, and unlike the somewhat blue-stained
highlights of Atkins’ and Dixon’s cyanotypes. The versos of the prints are
also clean white, unlike the more or less blued versos of many of Atkins’
and Dixon’s prints, suggesting a different coating technique, of brushing
rather than immersion.
Without further historical and graphological research, the author(s) of
this album will remain unidentified, but it can be suggested with some
confidence that it did not issue from the hands of Anna Atkins or Anne
Dixon. It is interesting that there should have been another botanical
cyanotypist active within a few months of the publication of Herschel’s
invention; particularly in view of the circumstantial indications that the
prints may not have been made in Northern Europe: the fine weather
conditions prevailing at the season of exposure, and the printer’s
apparent lack of (or possibly, ignorance of) ammonium ferric citrate as a
means of greatly accelerating the exposure.468 The exotic origins of some
of the specimens might also suggest a widely-travelled author, although
exchange of specimens between collectors was a common activity. Until
further evidence comes to light, the identity of this ‘anonymous
autobotanographer’ must remain a rather tantalising mystery.469
5.7 Later botanical cyanotypes

Following Anna Atkins’ pioneering endeavours, the cyanotype process
continued to be used sporadically for botanical photograms; but over the
years the purpose in this can be seen to shift away from simply recording
plant taxonomy towards the making of works of art. Connoisseurs have
always maintained that the vision and sensibility displayed in Anna
Atkins’ and Anne Dixon’s botanical cyanotypes transcend mere scientific
illustration, and elevate their images to the status of art-works, although
there is very little to suggest that they originally intended them as such. It
147
is interesting to observe how the recognition of creative botanical

cyanotype as art has slowly gained ground over the intervening years, to
the point where such works are now to be found on the gallery walls,
rather than the herbarium shelves. A recent celebration of botanical
drawings and photograms as “exhilarating artistic experimentation…in
drawing with light” may be found in the sumptuously illustrated book
Ocean Flowers.470
A later echo of Anna’s work may be found in the albums prepared by
Eric Craig (1829?-1923), who was a fern mounter and curio dealer in
Auckland, New Zealand. He had a shop near Auckland Museum in 1892-
3, from which he sold books of pressed ferns. He published two editions,
in 1888 and 1892, of 100 cyanotype illustrations of 172 varieties of fern,
following the layout and some of the typography originally employed by
Herbert Boucher Dobbie in his New Zealand Ferns published in 1880; it is
also apparent that he used some of Dobbie’s labelling but his specimens
are different.471 Examples of Craig’s work exist in the National Library of
New Zealand and in the National Museum and Gallery of Wales, in Cardiff.
The traditions of the botanical cyanotype were also carried on in the
USA by Bertha Evelyn Jaques (née Clausen, 1863-1941) of Iowa, who used
the process around 1905 to record images of numerous plant species,
some of them endangered. Attractive examples of her work have recently
commanded substantial prices at auction. The modelling of the light
permeating her diaphanes confers a delicacy of gradation on the
photogram which gives it the illusory appearance of a photographic
image of a three-dimensional original (see §6.2). Some of her work from
the collection of Hans P. Kraus, Jr., is illustrated in DiNoto and Winter’s
The Pressed Plant: The Art of Botanical Specimens, Nature Prints and Sun
Pictures.472
More recently, in 1996 a touring exhibition of ‘Floral Blueprints’ by
Alexander Hamilton, created as part of the Scottish ‘Fotofeis’
photography festival and sponsored by the Scottish Arts Council, brought
botanical cyanotypes to a wide contemporary public in several exhibition
venues. Hamilton’s technique differs from the customary use of exsiccati
for botanical photograms: his flower specimens, which he grows himself,
are fresh, and the act of pressing the flower into the surface of the
sensitized paper leaves an impression of its fleshy centre, and traces of
the natural pigments and pollen stain the image, adding sepia and other
tones to contrast with the blue.473 Although cyanotype is still the
principal medium, there is also a distant echo here of Herschel’s early
148
anthotype or phytotype process of 1840, which used the natural

pigments from flowers to make photographic images. Hamilton’s
preoccupation with light is manifest in the delicate translucency of his
flower heads and petal arangements, which are formally and
symmetrically set against a large ground of very dark blue, shading
almost into black; their mandala-like quality invites deep contemplation.
One reviewer has also likened them to astronomical photographs of
distant galaxies.
5.8 Julia Herschel’s lacemaking handbooks

To return to the use of cyanotype for photograms, a rare example
originates from Herschel’s own family: in 1869, his seventh daughter,
Julia Herschel (1845-1933, the 10th of his 12 children by Margaret),
prepared A handbook for Greek and Roman lace making illustrated with
specimens printed using her father’s cyanotype process, fig. 5.9.474
Figure 5.9 Cyanotype photogram of Greek lace by Julia Herschel
149
The work was published by R. Barrett and Sons of London, but the
author chose, like Anna Atkins, to remain anonymous; in 1870 Julia’s
second edition carried the shorter title Handbook of Greek Lace
Making.475
5.9 Early pictorial studies in cyanotype

While platinotype was still being developed in the 1870s and 80s, a few
photographic artists, concerned for the endurance of their work, had
begun to reconsider the humble cyanotype, whose simple processing in
water avoided any use of the fixing agents that tended to later degrade
the silver image. The French artist, Henri Le Secq (1818-1882), became
one of the most significant users of the process for this reason. His entire
oeuvre is reviewed in the Catalogue Raisonné, compiled by Janis and
Sartre,476 and a brief biography can also be found in The Art of the
French Calotype by Jammes and Janis.477 Although Le Secq virtually gave
up making new photographs after 1856, he continued to be concerned
about the longevity of his work. Of his architectural studies, those
mounted with gum into albums and presented to the Bibliothèque
Nationale in 1853-4 now contain faded silver prints of a pale yellow-
brown, which contrast starkly with the rich, unmounted specimens that
were deposited at the Bibliothèque des Arts Décoratifs. Having observed
that his silver prints were prone to fade, Le Secq, quite late in his life,
used the cyanotype process to reprint many of his negatives.478
It is not known exactly when, around the 1870s, Le Secq made his
important cyanotypes; his Fantasies Photographiques in still-life are
engaging and mysterious works, but studies of sculpture and architecture
predominate, and there are also some portraits, including Le Secq’s
father and mother, which suggest that he did not have any particular
aesthetic difficulty with using the process for this genre.
These cyanotypes constitute his enduring legacy: not a single still-life
print by him is known to have survived on silver salted paper.479 Among
the cyanotypes by Le Secq in the collection of the George Eastman House,
Rochester NY, is his famous image Farmyard Scene, and several of his
architectural studies, most of which are in excellent condition, except one
which shows some evidence of fading in comparison with its edges,
which were presumably matted at some time when the picture was
mounted and ‘exposed’. Jammes and Janis attribute the cyanotype
process that Le Secq used, and possibly the source of his sensitized
paper, to Marion and Company, or to Testud de Beauregard, but offer no
150
information as to whether this method differed from that originally

invented by Herschel.
Fig. 5.10 Henri LeSecq, Fantaisies
A more recent example of the ‘artistic rehabilitation’ of the cyanotype

can be seen in the treatment of the photographs of Paul Burty Haviland
(1880-1950), who was one of the lesser-known members of Stieglitz’s
Photo-secessionist circle in the first decade of this century. An exhibition
of Haviland’s work was recently mounted in Paris, including a substantial
number of beautiful cyanotype studies in portraiture and the female
nude.480 As a curatorial footnote in the catalogue, Heilbrun and Bajac
observe that these were most likely intended by Haviland himself only as
trial proofs of his negatives, which he subsequently preferred to print in
platinum or gum bichromate for exhibition or sale:
“Quelques tirages au platine, destiné probablement à l’exposition
voire à la vente, portent un monogramme et, plus rarement, une
date, à la diference des cyanotypes, tirés sur des papiers plus fins,
sur lesquels ne figure aucune mention et qui furent peut-être
151
utilisés par Haviland pour ‘vérifier’ ses négatifs, comme on utilise

aujourd’hui une planche contact” 481
Nonetheless these cyanotype proofs now stand as finished artworks in

their own right, and the authors of the catalogue also suggest that
Haviland may have had the issue of permanence in mind when he
originally made them:
“Ces images se caractérisent, en effet, par un manque d’afféterie,
un charactère direct mais gracieux, une poésie simple et raffinée
auquel le bleu du cyanotype - dont l’utilisation au départ n’était
peut-être pas une recherche esthétique mais un simple désir de
conservation, le cyanotype étant un procédé permanent - ajoute
encore, du moins pour notre regard contemporain, charme,
mystère et bien sûr couleur.” 482
The renowned American artist and teacher, Arthur Wesley Dow (1857-
1922), used cyanotype around 1899 to make a poetic study of his
hometown of Ipswich in New England; he compiled 41 prints into a
personal album dedicated to his friend, the poet Everett Stanley Hubbard.
Fig. 5.11 Arthur Wesley Dow, “Little Venice” from Ipswich Days, 1899
(Collection of John T. and Susan H. Moran)
152
This privately owned and previously undocumented album of his

vernacular cyanotypes was published in 2007 by the Addison Gallery of
American Art as Ipswich Days.483 Being a celebrated painter and man of
means, Dow had no need of cheap expedients for his materials, so we
may presume that the cyanotype – including its colour - was his medium
of choice for this affectionate and wistful portrayal of his beloved
environs. As its curator, Trevor Fairbrother, says of this choice:
“He loved the blue monochrome of cyanotype prints, delighting in
their capacity to give reality a strange or otherworldly cast.” 484
5.10 Documentary and topographical cyanotypes

Compared with the rather limited extent of its pictorial use, cyanotype
has found considerable application for documentary purposes, including
scientific, archaeological, ethnographic, topographical and engineering
studies. A noted British user of cyanotype during the 1880s was the civil
engineer, Washington Teasdale (1831-1903), from Leeds, who may have
been the subject of Frank Meadow Sutcliff’s complimentary observation
already recorded in §1.2. Teasdale was a prominent member of several
Victorian Scientific Societies – Astronomical, Microscopical, Geological,
Photographic, etc., - and he used cyanotype effectively to illustrate his
multifarious interests, as can be seen from an archive of his prints
preserved in the Museum of the History of Science, Oxford,485 which also
provides a commentary.486 It was said of him in a biographical note:
“…the doyen of local scientists, Mr. Teasdale was well known and
welcomed wherever men of science most do congregate.” 487
In the field of archaeology, an unusual illustration of the capability of

the process to render atmospheric studies of sculptural artefacts is
provided by a set of 30 cyanotypes produced around 1895 by Henri
Ranoux. These are described in Ken Jacobson’s beautiful catalogue,
Études d’après Nature,488 and they illustrate quite dramatically the Roman
finds from an archaeological excavation in North Africa, which includes a
self-portrait of the photographer himself working on the site. This album
also contains three collodion or silver-gelatin prints, which look rather
sadly washed-out and dull in comparison with the vibrant cyanotypes.
The Rijksmuseum in Amsterdam has substantial holdings of
cyanotypes within its photography collection, including albums of
documentary cyanotypes dating from the early 1900s by Hendrik Dooyer,
153
showing the previous Dutch colonies such as Suriname, documenting life

on the large colonial plantations of the time. 489
The Canadian National Archive has substantial holdings of cyanotypes,
including many full-plate architectural records of dwelling houses. The
simplicity of the process in not calling for elaborate darkroom facilities
may also be the reason for its employment to illustrate several personal
diaries - accounts both of everyday life and adventure - which date from
the 1890s. These include an album of about 100 cyanotype prints from
half-plate photographs of North American domestic life by W. Braybrook
Bayley taken in the period 1889-91. In view of the apparent well-to-do
character of his family, the economy of his chosen process is a little
surprising and the prints are actually mounted in a ruled cashbook! A
similar set of 17 cyanotypes of local Canadian landscape and family by
Able Silver Bowers date from ca. 1895. An anonymous diary of ‘An
Account of a Trip to Hunters’ Island’ dating from 1899 is illustrated with
24 cyanotypes, and on another adventurous theme, there are 18
cyanotypes by Walter Strand of gold-rush scenes in the Klondyke,
northern British Columbia, in 1899.
It is apparent from a substantial collection of photographic postcards
in the archive of the George Eastman House, in Rochester, NY, that
postcards were commercially available in the USA in the period 1900-
1914 precoated with cyanotype sensitizer; amateur photographers could
easily sun-print their personal pictures onto the cards, with minimal
resources and processing, before adding their greetings, and posting
them to friends and relatives.
The Agricultural College of Utah (now Utah State University) holds an
extraordinary collection of 2,608 cyanotypes, dating from 1896 to 1916,
documenting many aspects of the history of the College, including
images of students, faculty, class-room scenes, buildings and grounds,
events, experiment stations, athletics, student projects, and agricultural
products. An alumnus of the college, George Melvin Turpin (1882-1943),
took many of the photographs. He was appointed Assistant Station
Chemist in 1906, then Professor of Poultry Husbandry from 1909 to
1913. These are the oldest photographic records of the College; the full
2,608 images are listed in a comprehensive inventory,490 of which 364
may be viewed in an accessible online archive.491
In the area of ethnographic studies, the Canadian National Gallery, in
Ottawa, has a set of fine cyanotype prints of photographs of Native North
American Indians made by Edward S. Curtis around 1910. Several tribes
154
are represented, Nootka, Piegan and Arapaho; the hunting scenes are
particularly striking and the medium seems sympathetically well-suited
to the subject-matter, moreover, cyanotype enabled him to proof some
of his negatives in the field – which numbered 40,000 in total!
The Photographic History Collection of the Smithsonian Institution has
an extensive archive of the chronophotographic work by Eadward
Muybridge, proofed in cyanotype as strips from his Animal Locomotion
Series.492
Fig. 5.12 Eadward Muybridge

(Collection of the Smithsonian Institution)
There are 850 of these strips, which are believed to be contact proofs
of the original glass-plate camera negatives, whose whereabouts are now
unknown. A comprehensive public resource site now makes most of
these cyanotype proof strips available online.493 Many of these images are
particularly important because they reveal background details of
Muybridge’s photographic modus operandi which were cropped out of
the collotype prints in the published versions of his work. The unedited
pictures reveal the setup at his outdoor studio, with its corrugated rubber
track, the backdrop, props and cameras. In the animal series, a trainer
can often be seen to lead the animal or urge it along the track.494
155
In the collection of the George Eastman House there is a topographical

record, made as a survey for a railway company, which consists of
approximately 50 full-plate cyanotypes, possibly printed on commercial
blueprint paper, and entitled Elimination of Grade Crossing Photographs,
February 1907, Schenectady. Also on the topic of railways, there are
about 4000 cyanotypes, 6.5 x 8.5”, held by the National Railway Museum
at York, England, which are proofs of a large archive of ‘engineering-
works photographs’ dating from the period 1910-1953; these are glass-
plate negatives from the Great Eastern (later LNER) works at Stratford,
illustrating the stages of locomotive manufacture, testing and
maintenance, and many of the peripheral matters of a large engineering
industry, which include even portraits of the employees.495 The whole
archive is well-documented and the images are of excellent quality and
colour, generally 20/21E8.
All engineering projects depended on blueprinted plans, and some
were also nicely complemented by cyanotypes documenting the actual
progress of the construction.496 The ready availability of blueprint paper
'on site' must have provided an easy and inexpensive way of immediately
proofing the large format camera negatives made to record some of the
largest projects in history. The Irn-Bru Company of Glasgow has a slogan:
"Made in Scotland from Girders" which also serves as a most apt
description of that engineering wonder, the Forth Rail Bridge in Scotland.
Begun in 1883, this first steel structure in Great Britain took seven years
in the building.497 Besides the blueprints proper, in the Institute of Civil
Engineers in London, the gigantic construction of the Bridge during
1883-1890 was also documented photographically by one of the
project’s engineers, Evelyn George Carey, whose negatives were proofed
in cyanotype, as well as albumen, and the specimens are now kept in the
archive of the Canadian Centre for Architecture, in Montreal.498
Fig. 5.13 Evelyn George Carey, Construction of the Forth Bridge

(Collection of the Canadian Centre for Architecture)
156
The recent book by Michael Gray and Angelo Maggi describes the
achievement of the photographer:
"The building of the Forth Bridge was marked and controlled
through photography; the official photographer: Evelyn George
Carey, a young engineer and personal assistant to the designer,
Benjamin Baker worked on the project over a period of eight years
between 1882 to 1890. His pictures capture, clearly and lyrically,
the scale, tensions and inherent dangers of such a project." 499
An engineering enterprise of comparable magnitude was the cutting of

the Panama Canal. The National Library of Australia, Canberra, has an
album containing 37 full-plate cyanotypes, entitled Percement du Canal
de Panama,500 which makes an interesting comparison with an album of
57 albumen prints having the same title.501 These were probably
assembled from the work of French photographers during the period
1888-1893. Ferdinand de Lesseps’ attempt during 1881-90 ended with
his bankruptcy, and the project was taken over by the US Army Corps of
Engineers and completed ca. 1904-14.
The same Corps of Engineers was also responsible for the survey of
the Mississippi River and opening it up to modern commercial navigation.
This survey gave rise to an entire archive of topographical and
documentary studies of the river and its environs, photographed and
printed over the period 1883-92 by Henry Peter Bosse (1844-1903), a
German-born immigrant, trained in art and engineering, who became a
cartographer and draughtsman with the Corps of Engineers.
Bosse’s photographic archive, showing the survey, mapping and
maintenance of that great river, remained unknown until ca. 1988, when
an album of 169 of his cyanotypes was brought to light by historian and
antique dealer, Michael Conner, who published on the subject.502 The
album was sold at auction in 1990 and subsequently broken by dealers,
one of whom has published on some of its contents, now dispersed.503
Another album of Bosse’s work was discovered by a Corps historian, John
Anfinson, in the pilothouse of a river dredge; it is now preserved by the St
Paul District History Office of the Corps,504 and its contents can be viewed
online at the site of the Minnesota Digital Library.505
During the 1990s, five more cyanotype albums by Bosse were located
and are now distributed between the Rock Island Office of the Corps, the
Mississippi River Museum at Dubuque, Iowa, and the Mayo Foundation in
Rochester, Minnesota, the latter providing most of the images for Mark
157
Neuzil’s 2001 publication, Views on the Mississippi: The Photographs of

Henry Peter Bosse.506 Neuzil estimates that there is a total of about 350
original images known to be by Bosse, all originally made on 11”x14”
glass plates, of which only seven survive.
Fig. 5.14 Peter Henry Bosse
As the quality of his vision has come to be more widely appreciated,

Bosse’s prominence as a photographer has risen, so that he is now
regarded as one of the foremost landscapists of the USA in the late 19th
century. The quality of his work has been appreciatively articulated by
Earl Johnson:
“His vision was romantic, but he was scrupulous in his composition
and in his choices of subjects. While many of his photographs were
obviously shot to document the work of the Corps of Engineers to
tame the river, his vision displayed a much wider sensibility. He
includes many images of cities, bridges, harbors, and rail yards, as
if to explore the interaction between the river and the people who
live with it and use it. Bosse was trained as an artist (a few of his
paintings still exist), and his photographs display a highly
developed sense of composition. I think that he will be
remembered as one of the great photographers of his time.” 507
Both Neuzil and Johnson have observed that the paper used by Bosse in
at least two of his albums is watermarked Aloe's Satino Johannot et Cie
Annonay,508 and a fine example can be viewed online, courtesy of the J.
Paul Getty Museum, together with a description of this watermark.509
158
Evidently there was a market for ‘pictorial quality’ blueprint paper at

this time - at least in the USA - which was largely fulfilled by the A.S.
Aloe Company, the leading makers of surveying instruments and
precision machinery, who imported French paper rawstock specially made
for them by the famous mill of Johannot et Cie at Annonay.
Fig. 5.15 Advertisements for Blue print papers in the USA, 1894
The Aloe Company machine-coated it themselves with cyanotype

sensitizer, in St. Louis, Missouri. Their catalogue also carries the boast:
“Our facilities for making blue prints are superior to any in America”. 510
The New York Subway was constructed during 1901-09, and
extensively documented in cyanotype by Pierre Pullis.511
Likewise, above ground, the construction was recorded of the Boston
Elevated Railway Line. An eclectic online exhibition and anthology of
historical cyanotypes, featuring many engineering and commercial
objects, may be found on the Luminous Lint website.512
159
Fig. 5.16 Pierre Pullis,
5.11 Cyanotype currency notes and postage stamps

A curiously resourceful use of the cyanotype process arose in Southern
Africa during the Anglo-Boer War (1899-1902); the following account of
the episode is substantially condensed from the detailed study published
by the philatelic historian, Robert Goldblatt.513
In October 1899, shortly after the commencement of hostilities, a Boer
force laid siege to the town of Mafeking, in which were stationed about
1200 British and Cape troops. The beleaguered town was well-
provisioned and defended; it survived the bombardment and consequent
privations until, after 217 days, the siege was eventually lifted in May
1900 by British columns arriving from Kimberley. But by the end of
January 1900, a shortage of currency had arisen in the town, so to
provide for payments during the siege, the military commander, Colonel
Robert Baden-Powell, ordered the minting and issue of garrison siege
notes - popularly known as “good-fors” - in several denominations,
which would be redeemed on the resumption of civil law. Of these, the £1
note was designed by Baden-Powell personally and unlike the other
notes, which were printed, it was duplicated photographically onto glass
plates by Mr E J Ross, the local auctioneer who was a skilled amateur
photographer. Since there was insufficient photographic paper on hand
to print the issue, Ross decided to employ the recipe for cyanotype to
make the printing paper locally.
160
It is said that one of the two ingredients, potassium ferricyanide, was

in short supply in Mafeking, so native runners were appointed to the
unenviable task of smuggling the chemical through the Boer lines by
night. On the first attempt, the runners returned safely - but bearing
potassium ferrocyanide: the message had been badly written. At a second
attempt, the correct chemical was secured, but the unfortunate runners
were caught and shot, and their potassium ferricyanide confiscated.514
The third attempt was successful, and Ross was then able to prepare the
blueprint paper for sun-printing, followed by processing in the ‘Mafeking
Mint’, which was a bomb-proof subterreanean dark room. The £1 siege
note was issued in March 1900, as an edition of 683 individually
numbered specimens, each having taken fifteen to twenty minutes to
make, signed by the bank manager and countersigned by the paymaster.
Of these, only 44 were ever redeemed after the relief of the town and the
return to civil law.
The success of the cyanotype printing of the £1 siege notes must have
suggested to Baden-Powell that the same process could also be
employed to solve a growing crisis in the shortage of postage stamps for
internal use. Mafeking was defended on a perimeter of six outlying
fortifications, up to one mile distant, but traffic and communication
between the garrisons manning these posts and the town centre was
necessarily restricted. To provide an internal postal facility, Baden-Powell
set up a corps of messenger-cadets using bicycles, under the command
of Cadet Sergeant-Major Goodyear, who at the time had just achieved the
tender age of 12 years - the very model of an ideal future Boy Scout.
Postage rates per half ounce were 1d. within the town and 3d. to the
outposts, a small price considering the risks the messengers ran. Not
surprisingly, the demand for these denominations of postage stamp was
heavy and when the supply of official Cape stamps, overprinted ‘Mafeking
Besieged’, ran out, a special issue was printed locally by cyanotype. Dr. D
Taylor, another local amateur photographer, took a portrait of young
Goodyear mounted on his bicycle for the 1d denomination, and Ross
photographed Baden-Powell himself for the 3d. To these were added
appropriately designed ornamentation and the wording ‘Mafeking
Beseiged’ to form two ‘master dies’ which were then each duplicated
photographically twelvefold, and re-photographed in reduction onto
glass plates, from which the stamps could then be contact-printed by
cyanotype in sheets of 12 at a time, which were then gummed and
perforated.
161
Fig 5.17 Cyanotype postage stamps of Mafeking beseiged, 1900
The ‘Mafeking 1d and 3d Blues’ were issued on 9 April 1900, nearly

800 sheets of each denomination were minted by the local printer,
Townshend. These unofficial stamps echo the precedent of the celebrated
official Victorian twopenny blue – which was also printed in Prussian blue.
The idiosyncratic postage stamps and £1 banknotes that resulted from
this historic episode are now highly-prized - and highly-priced -
collectors’ items that have naturally intrigued philatelists and
numismatists over the years. The rarest varieties of the postage stamps
are deemed by collectors to be those of a pale grey colour, rather than
the deep blue, but in view of the ease with which cyanotypes can be
chemically bleached it might be imprudent to invest too heavily in such
‘rarities’.
The interest in these items is especially high in South Africa, where
there is also an historical awareness of Herschel’s presence at the Cape
over the years 1834 to 1838, just prior to the invention of photography,
and the subsequent visits there of other British photographers.515 The
debate rumbles on regarding the suggestion that Queen Victoria
personally considered it an act of lèse-majesty on the part of Baden-
Powell to have allowed his portrait to appear on a postage stamp, which
was a privilege normally reserved for the British Sovereign alone.
Certainly, the Queen gave no outward sign of displeasure, and the
defender of Mafeking became General Sir Robert Baden-Powell, now also
remembered as the founder of the Boy Scout Movement in 1907.
162
5.12 Linley Sambourne’s cyanotypes

Edward Linley Sambourne (1844-1910),516 an artist in black and white,
became the cartoonist-in-chief for the humorous British weekly Punch in
1900, having contributed to the magazine since 1867. To assist his
drawing he kept an archive of reference images for his illustrations that
has been preserved at Leighton House Museum, in the Royal Borough of
Kensington and Chelsea, and is in the process of being conserved and
catalogued at the item level.517 His working environment may be seen
nearby at his dwelling, Linley Sambourne House, which has also been
preserved by the Borough as a museum. Sambourne used photography
extensively to provide ‘model’ images for his illustrations; it is estimated
that he has left about 30,000 photographs - mostly 5x4 inches - perhaps
the largest collection known,518 and he proof-printed many of his
negatives of posed figures by the cyanotype process. The blue colour of
his prints was of little consequence for his line-drawings provided the
material could resolve adequate detail.
Fig. 5.18 Edward Linley Sambourne

(Collection of Royal Borough of Kensington and Chelsea)
When necessary for his illustrative purposes, he had enlargements

made by others in silver-gelatin. From the consistent size and accurate
regular rectangular shape of his cyanotype prints, it seems very likely that
he used small pre-cut sheets of a commercial blueprint paper of the kind
then currently marketed by Marion, Rouch, or Bemrose for pictorial
purposes rather than for plan copying. It is likely that he found cyanotype
a most convenient process because he did not possess a proper
photographic darkroom, with safelighting, but just used a curtained-off
area of his bathroom at home, in which he probably printed privately
163
while his wife was away, because many of his images were also blue in
the more pejorative sense. In 2002 an exhibition entitled Public Artist,
Private Passions, conveyed something of the voyeuristic nature of his
image archive.519
5.13 Some contemporary cyanotypists

There are now, in 2014, so many artists working with the cyanotype
process that this text, oriented towards the process rather than its
practitioners, cannot possibly do them justice. A few references to
exhibited and published work and websites follow, which will lead the
reader to a number of present-day cyanotypists, but with the frank
admission that the choice is the author’s personal one rather than being
representative or exhaustive. Many more afficionados of the cyanotype
medium can be found, for example, participating in the Facebook public
group for Cyanotype with over 3,400 members,520 on Flickr,521 and on the
AlternativePhotography.com website,522 representing over 500 artists,
whose editor, Malin Fabbri has published a thesis on the subject of the
contemporary professional use of cyanotype in graphic design.523
To begin our survey in the UK: photogram artist Angela Easterling was
inspired by the work of Anna Atkins, and has demonstrated the
cyanotype process for botanography in workshops at the renowned Eden
Project in Cornwall, and in collaboration with the Royal Botanic Gardens
at Kew, as well as using it on a human scale for making live figure
photograms in socially involving workshops.524 In Kent, Nick Veasey has
made very striking cyanotype prints from some of his extraordinary X-ray
photographs.525 Master printmaker, the late David Chow, was known
chiefly as an exponent of the palladium-platinum medium until his recent
tragic death, but began his exploration of alternative process printing
with studies of cyanotype flower photograms, bound into handmade
books.526
In Scotland there has been a significant resurgence of interest in
cyanotype, as exemplified by the plant photograms of Alexander
Hamilton, described in §5.7. The link between cyanotype and the
blueprinting practice of engineering drawings has been explored by
Roger Farnham and Harry Magee of Glasgow Print Studio, and was
brought to the public in a program of lectures and exhibitions occurring
throughout the city of Glasgow during early 2013.527 In Stockholm, Malin
Fabbri has set up, and edits the popular alternative photographic process
site AlternativePhotography.com, which is a valued resource for
164
information and research, representing more than 400 artists, over half
of whom have contributed to Fabbri’s published anthology of alternative
work in many processes, including cyanotype.528
In the USA, the artist John Dugdale of New York is celebrated for his
body of sensitive personal work in cyanotype, carried out in spite of his
serious visual impairment.529 Robert A. Schaefer Jr., has been a
distinguished practitioner and noted exhibitor of the cyanotype process
since the 1970s, and explains it thus:
“My favorite 19th century process is cyanotype because it enhances
the light surreal quality of my photography.” 530
Fig. 5.19 Ciurej & Lochman
Collaborative artists, Barbara Ciurej and Lindsay Lochman, have

assembled a fascinating series of composite images comprising plant
cyanotypes montaged over pigment prints of statuary-like high-key
portraits of women.531
Brian Pawlowski has written extensively on his blog of his journey of
discovery in making expressive cyanotypes by the ‘new’ method (§7.4),
and illustrates his work with beautiful examples.532 The innovative work
of John Metoyer in cyanotype has been very favourably reviewed by John
Wood, whose essay has been mentioned in §1.8.533 Brenton Hamilton 534
165
and Peter Nappi 535 are also noted practitioners with interesting websites
showing cyanotypes.
Jesseca Ferguson's cyanotype still-lives made from pin-hole
photographs collaged into book objects convey a museum-like ethos that
engages thoughtful examination.536
Fig. 5.20 Jesseca Ferguson
Possibly the most significant proof that cyanotype has finally achieved
an improved status of acceptance in contemporary artistic practice was
signalled by the curation in 2016 of an exhibition devoted solely to the
166
medium at the Worcester Art Museum, in collaboration with Clark

University, in Massachussetts. This was probably the first exhibition
devoted solely to cyanotypes in modern times, and was accompanied by
an excellent well-illustrated catalogue.537 The work on display ranged
from nineteenth century photographers such as Henry Bosse, Edward
Sheriff Curtis, and F. Holland Day, to contemporary artists, Jesseca
Ferguson and Christian Marclay. Organized in collaboration with a
seminar from Clark University, the exhibition was presented with
thematic emphasis on botanicals, landscape, abstraction, and
portraiture—areas that dominated much of the production of cyanotypes
in the early twentieth century and recur in contemporary work.538
167
6 Methodology of Contact-printing Practice

“Tui lucent oculi (“Your eyes sparkle with a light
Sicut solis radii, Just like the rays of the sun,
Sicut splendor fulguris Or the splendour of lightning
Lucem donat tenebris.” Illuminating shadows.”)
Carmina Burana
(Codex from the Monastery at Benediktbeuern, ca 1230)
Instructions for making traditional cyanotypes may be found in numerous

publications on photographic practice, but most of these accounts are
perfunctory, treating the process only as a curiosity. Recently, a few
publications have acknowledged the cyanotype more seriously as a fine-
print medium, to be ranked alongside other alternative processes.539 This
chapter and the following one are intended to provide all the information
needed to make fine cyanotypes, assuming no previous experience on
the part of the reader. As a preliminary, some consideration is given –
possibly for the first time - to the geometrical-optical effects of sun-
printing photograms from objects that are not flat like engravings and
photographic negatives, but rather three-dimensional. Then follows a
description of the material requirements and modus operandi that are
common to all the processes and formulae for making cyanotypes - and
indeed any other iron-based, plain-paper contact prints. But first, our
language stands in need of a word to encompass the class of possible
objects that can be used to intercept and modulate the light when
making photographs without lenses.
6.1 Diaphanes: a neologism

Throughout the first fifty years of photography, prints, as distinct from
camera negatives, were generally made without the aid of any optical
device employing lenses.540 A negative, or some other object, was simply
interposed between the sensitised surface and the light source, which at
that time was, faute de mieux, the sun or a bright sky. The print
sensitizer could be chosen from a variety of formulations, including
cyanotype. Close contact between object and surface ensures a sharp
outline, and any translucency in the object will transmit a gradation of
tone to the image, which can be both delicate and faithful. When the
object is a negative originated in a camera, the resulting photograph is,
of course, called a contact print, but the images derived from other
objects we now refer to as ‘photograms’.541 A photogram is a camera-
168
less image, usually negative-working, of the shadow cast by an object

onto sensitized paper. If the object is at all translucent, it may modulate
the transmitted light, and the photogram then ceases to be a mere
silhouette, but acquires some internal structure. Some of the
philosophical implications of this mode of photographic image-making
have been discussed recently by Geoffrey Batchen in terms of Derridean
concepts.542
The items used to generate photograms can be quite various, but a
usual requisite is that they should be quite flat: leaves, ferns, sea-weeds,
and other botanical specimens; feathers and insect wings; lace, silks, and
other thin fabrics; cliché-verre plates, painted glass, printed text,
handwritten notes, drawings on tracing paper, and printed images on
paper, such as engravings, mezzotints, and woodcuts.543 The most
extraordinary object ever used to register itself as a photogram must be a
thin frontal section, cut longitudinally with a microtome from the frozen
torso of a human male cadaver; this display of the internal organs of the
body was printed in cyanotype in the 1960s, and has been shown recently
in exhibition.544
In writing about the practices of early photography, I have become
increasingly conscious that the English language lacks any collective noun
to encompass the whole class of objects, such as those listed above,
which have been used to imprint their likenesses by interposing them in
the path of the light falling upon a photographically sensitised surface.545
To serve this semantic purpose, I propose a revival and re-definition of
the obsolete noun diaphane, in the hope that it may find acceptance
among photohistorians and possibly even lexicographers. The word has
completely fallen out of use in its original meaning.546 The Greek roots
(dia- fanhV, ‘through-showing’) seem appropriate,547 and there are
precedents for employing them in a photographic context, as evidenced
by Thomas Sutton’s obsolete diaphanotype process of 1856, and the
diaphanoscope, one of the many names contrived for an optical device
that could display transparent positive photographs, which are
themselves known today as diapositives.548 A diaphanometer was a
nineteenth-century instrument for measuring transparency, especially of
the atmosphere (see §1.6). The words photophane and lithophane are
similarly derived. The practice of photodiaphanie entailed the use of an
albumen stripping emulsion for making transparencies.549 Among the
genteel Victorian pastimes described in volumes such as Elegant Arts for
Ladies is the art of diaphanie: ‘the decoration of glass to the resemblance
169
of stained windows and painted transparencies’.550 In French, diaphane is

an adjective meaning ‘transparent’. Even if it can plead no other
recommendation, the reappropriation of the word ‘diaphane’ would at
least enable us to avoid the enforced solecism of having to refer to an
engraving, a piece of lace, or a plant, for example, as ‘negatives’.
6.2 Skiagraphy of photograms

The casting of shadows on a photochemically sensitized surface can be
described by the same geometrical-optical principles that are used to
determine the appearance of light, shade, and shadow forms in pictures.
This is the sub-science of ‘sciography’ or ‘skiagraphy’ 551 which has
existed since the early Renaissance, finding its origins in the elucidation
by Filippo Brunelleschi ca. 1413 of the principles of geometrical
perspective,552 which are elaborated in manuals of sciography,553 and
applied to the drawing of shadows in projection, mainly for the purposes
of architectural illustration. In these treatments sciography assumes that
the sun is a static, point source of light. In distinction, I will use the word
‘skiagraphy’ for its broader relevance to photography, because it can also
bring in the dimension of time, as well as space, to explain the
appearances of those historic photographs which were printed-out with
the lengthy exposures required by early proto-photographic
processes.554
The parameters most relevant to skiagraphy are the spatial resolution
of which the healthy human eye is capable, and geometrical aspects of
the sun’s light. Under optimum conditions,555 our vision can resolve
objects separated by an angle subtended at the eye of 1/3000 radian,556
and average vision can usually resolve 1/2000 radian;557 but the criterion
normally accepted for photographic purposes is less stringent than this:
an outline is deemed to be acceptably sharp when its ‘blurred’ region (the
penumbra) subtends an angle at the eye of 1/1000 radian (0.05° or 3
minutes of arc).558 This figure is taken as the limiting diameter of the
‘circle of confusion’, when a disc is just perceived as a point, and it is the
figure customarily employed in calculating the effective depth-of-field of
lenses.559 A viewing distance of 250 mm is the usual nearpoint for normal
vision, and is also the natural, comfortable distance for viewing a print of
about ‘whole plate’ or 10x8 inch size; at this distance, an angular
resolution of 1/1000 radian obviously corresponds to a linear resolution
of two points in the image separated by a distance of 0.25 mm.560
170
The angular diameter of the sun viewed from the the earth’s surface is
nearly 1/100 radian, or ca. 0.5°.561 The sun’s disc is therefore not exactly
a point source of light, as usually assumed by sciography, and any
shadow that it casts displays two distinct regions: the umbra and the
penumbra (figure 6.1).562
angle = 1/100 radian
penumbra
sun object umbra
penumbra
shadow plane
Fig 6.1 The structure of a solar shadow (not to scale).
The umbra is defined as the region of ‘totality’ where the sun’s disc
(or other light source) is completely obscured by the shadowing object,563
and it is the dominant feature of most shadows. The penumbra is the
region at the edges of the umbra where the illumination is partial, the
tones graduating from full shade to full light as a greater portion of the
sun’s disc appears uncovered; so the penumbra confers a blurred or
‘fuzzy’ edge to any shadow cast by a light source of finite size. Any solar
penumbra subtends an angle of 1/100 radian at the edge of the object
casting it; i.e. the width of a penumbra is approximately 1/100th the
distance of the object from its shadow. The apparent diurnal motion of
the sun is, thanks to the earth’s rotation about its axis, a full orbit of
360° over 24 hours; that is, an angular displacement of 15°/hour, as any
sundial will attest. Therefore, the sun travels through an arc in the sky
equal to its own apparent diameter in a time interval of about 2 minutes.
From these observations, four ‘Laws of Skiagraphy’ may be posited:
I. Sun-cast shadows appear blurred, with a penumbra, when viewed
from less than ten times the distance of the object casting them.
II. Solar exposure of a photogram of two minutes duration causes
shadow edges to blur additionally in the direction of travel by an amount
approximately equal to the limiting value in Law I.
III. The umbra of a narrow object of width, w, at distance, d, will be
completely displaced by penumbra within a solar photographic exposure
of duration ca. 4w/d hours.564
IV. The light scattered from a clear sky can ‘fill-in’ an umbra to the
maximum extent of ca. 3 stops less exposure than the direct illuminance
171
of the sun. The extent of ‘fill-in’ is proportional to the solid angle of

irradiating sky ‘seen’ by the sensitized surface.
When making photograms it is usually advised that, to obtain a sharp
image, the object and paper must be in close contact – but we now see
that this is not always necessarily so. The criterion for sharpness in a
projected photogram is easily determined based on the resolution: if we
assume that the finished image will be viewed at a minimum comfortable
distance of 10 inches (254 mm), then by the First Law of Skiagraphy, a
sun-printed image will appear sharp to the eye if the object that makes it
is distant from the sensitized paper by no more than 1/10th of this
viewing distance – i.e. 1 inch (25 mm).
Moreover, the exposure for such a projected image must be relatively
short (less than 2 minutes by the Second Law of Skiagraphy) if the sun’s
apparent motion is not to blur the edges significantly.565 Most cyanotype
and photogenic drawing photograms (using the old, insensitive
processes) were made with exposures substantially longer than two
minutes, which would have extended the depth of the penumbra and
caused considerable blurring if the objects were not in contact. This
explains the early photographers’ preference, in their choice of subject
matter, for objects that were already substantially two-dimensional,
flattened into close contact with the paper under a sheet of glass or mica:
eg. leaves, algae, feathers, silks, and lace.
However, the criterion does suggest that, provided sun exposures can
be kept short, as with the new cyanotype process (§7), there is no need
for an object to be in overall contact, and it explains why objects with
some three-dimensional quality can also produce interesting photograms
with apparently fairly sharp outlines.566 That an object can project a sharp
photogram without being in overall contact with the sensitized surface
leads to a further interesting effect on the ‘modelling’ of such images.
While totally opaque objects in close contact always yield simple
silhouettes, even incomplete contact can still maintain a sharp outline
with little or no penumbra, provided the object remains no more than an
inch or so away from the sensitized surface. Now, if there is a convexity
of shape, the skylight can penetrate exposed regions of the umbra,
causing a gradation of ‘fill-in’ light within the umbra, that appears in the
image as a ‘false shading’ for the shapes of curved objects (Fourth Law of
Skiagraphy). This phenomenon imbues the photogram with a 3-
dimensional quality,567 which is well-exemplified in the botanical studies
of Bertha Evelyn Jaques (§5.7),568 where the ‘modelling’ caused by the
172
thickness and curvature of the plant stems and florets, should be

compared with the ‘flat’ photograms by Anna Atkins. To achieve this
‘modelling’ effect in a photogram, three conditions must be met: a bright
sun, a clear hemisphere of sky which is ‘seen’ by the paper and not
obstructed by a high horizon, and a short exposure, if the outline is to
remain sharp.569 Obviously the specimen should not be flattened onto the
paper, but simply allowed to rest upon it – no printing frame is required,
but precautions may need to be taken to counter the effect of the wind.
This method of making photograms is exemplified below in Fig. 6.2.
Fig. 6.2 Mike Ware, photograms of apples and sea urchins made with
New Cyanotype exposures of ca. 1 minute to the Aegean sun.570
173
6.3 The processing environment

A benefit of the low sensitivity of alternative printing processes like
cyanotype is that you do not need a photographic darkroom or special
safe-lighting to work under. Ordinary curtains or blinds should subdue
daylight sufficiently for a ‘dim-room’, but for preference the working
area should be illuminated by incandescent (tungsten) light: a 40 watt
bulb, distant two meters or more, is safe for the time periods normally
needed. Avoid fluorescent light if you can; some types of tube have a
significant output of ultra-violet light which may cause fogging of the
sensitized paper.
You will need a clean, flat, dry surface for preparing the sensitized
paper and enough wet-processing space, preferably a sink, for one or
two photographic dishes of an appropriate size, together with a means
for washing your prints in running water, and an indoor drying facility
that need be no more sophisticated than a ‘clothes-line’ and pegs (you
should, however, select the design of your clothes-pegs very critically!)
The sensitizers and processing chemistry used are odourless and no
fumes are normally evolved, so there is no need for special ventilation in
the workplace. However, it is essential to observe clean working methods,
and any spilt substances must be removed immediately; remember that
the sensitizer will stain most surfaces and materials rather strongly.
Before commencing any practical work, it would be prudent to acquaint
youself with the hazards associated with the chemicals, which are
summarised at the end of this chapter. See my disclaimer on p. 6.
6.4 Materials and equipment

The following items of equipment will be needed:
Pencil
Ruler
Adhesive masking tape
Blower brush for dusting-off negatives
White blotting paper
Sheet of plate glass (4-6 mm thick and larger than the paper)
Jug to measure 1 litre (1000 cc)
Dropping pipettes 2 to 3 cc (or small syringes - the plastic disposable
type without needles - can also be used)
Small mixing glass (a liqueur or 'shot' glass is ideal)
Scales accurate to at least 1 g, or a chemical balance
174
Small spatula or plastic measuring spoon

Brown bottles for sensitizer solutions (ca. 100 cc is suitable)
A photographic dish (larger than the paper)
Means of print washing
Timer (a wristwatch will do)
Sheets of artists' paper suitable for coating* §6.4.1
Glass spreader rods or brushes for coating the paper* §6.4.2
Contact-printing frame* §6.4.3
Ultra-violet light source* §6.4.4
Suitable negatives* §6.5
*The specifications for these last five items are described in detail below.
6.4.1 Paper characteristics
Since the changes in industrial papermaking practice in the 1980s, it has
become increasingly difficult to source paper suitable for coating with
siderotype sensitizers. Paper chemist Dr. John C. Roberts of the University
of Manchester Institute of Science and Technology summarises this
historical shift in commercial practice as follows:
"For the period from around 1840 to the early 1970s paper was
usually made in an acidic environment at pHs of around 4-5. This
was because many grades required the use of rosin and aluminium
sulfate for the control of water penetration (sizing), and solutions
of aluminium sulfate exhibit a pH of around 4.5. Aluminium sulfate
has also been popular with paper makers because it assists the
flocculation of colloidal particles and therefore behaves as a mildly
effective retention aid. However, since the early 1970s there has
been a move away from acidic systems towards neutral and even
slightly alkaline pH. The advantages of operating at higher pH are
that there is reduced corrosion, greater strength arising from better
swelling of fibres at higher pH, the possibility of using high filler
additions and the energy savings associated with the easier drying
of filled paper. This change has had a profound effect upon the
whole of the chemistry of the aqueous fibre suspension." 571
The traditional cyanotype process can be applied, with some degree of

success, to almost any absorbent surface, including tee-shirts, trainers,
cushion covers, plywood and old grocery bags, (§4.6, 4.7). But if your
objective is high-quality cyanotype printing, it is best to work on paper
having the following characteristics:
1 High (>98%) alpha-cellulose content (i.e. a paper made solely from
cotton linters or linen fibres). Papers with a mixed furnish can absorb
sensitizer unevenly, causing granularity or blotchiness. The presence
of lignins will ultimately cause deterioration.
175
2 Internal sizing with Aquapel -or other equivalent alkylketene dimer-

the modern neutral sizing agent for fine papers. Alum-rosin sized
papers seem to work well, but their archival quality has been
questioned. Gelatin-sized papers are now less commonly available, but
can work well.
3 The paper must be free of alkaline buffering agents such as calcium
carbonate (chalk) which is often added to protect archival quality
papers from the effects of acid. Calcium carbonate reacts with the
sensitizer chemicals in a disadvantageous way, precipitating insoluble
calcium oxalate (also responsible for some kidney stones) and
hydrolysing the iron salt. Alkali also tends to destroy the Prussian blue
image (§9.2). The quoted pH of the paper should preferably be less
than 7.
4 A fairly smooth, hot-pressed (HP) surface is desirable - unless you wish
to make a highly textured print. ‘Not’ and ‘Rough’ surfaces may prove
difficult to coat and cause loss of resolution. However, some very
smooth, heavily calendered commercial papers do swell and roughen
on wet processing.
5 A weight of paper around 160 to 200 grams per square meter (gsm or
g/m2) is generally sufficient to ensure adequate wet strength for
smaller prints up to A4; larger prints benefit on heavier, 240-300 gsm
paper, but very heavy papers will be slow to wash completely. Thin
tissue can yield exquisite prints - if you have the skill to coat and
manipulate it in the wet baths.
6 The paper should not contain any other additives, such as clay fillers,
retention aids, wet-strengthening agents, optical brightening agents,
bleaches, alkaline buffers, dyes or white pigments such as titanium
dioxide, which is photoreactive (Appendix III.9).
7 A ‘wove’ mould paper texture is preferable to ‘laid’ for pictorial
purposes.
Increased awareness of the need for conservation of paper documents

and works of art has led to the formulation by the International Standards
Organization of specifications ISO 9706:1994 for Permanent Paper, and
ISO 11108:1996 for Archival paper, both of which require the presence of
an equivalent of 0.4 moles of alkali per kilogram of paper, i.e. an [OH–]
equivalent volume concentration in the paper of ca. 0.3 molar, which is
176
comparable with the [Fe3+] concentration of the applied sensitizer

solution, §9.4.5. For an average paper weight of, say, 250 g/m2, and a
typical coating weight of siderotype sensitizer, it can be calculated that,
per unit area, the permanent and archival papers contain an 8-fold molar
excess of alkaline reserve (equivalent to [OH]- ) available to react with and
hydrolyse the ferric salt. There is clearly great potential for
decomposition of the sensitizer chemistry if only a fraction of this alkali
in the sheet can contact the iron(III) salt, area for area, in the wet state,
although the chalk buffer is quite involatile and unlikely to migrate
through dry paper.
This new ISO specification accounts for the fact that, over the last few
decades, it has become increasingly difficult to find papers that work well
for siderotype printing, among the high quality cotton papers
manufactured for the fine-art market of watercolorists, sketchers and
printmakers and the archival document market, which all presumably
conform to the above ISO standards. Such papers now always boast that
they are "acid free" – implying that they probably contain an additive of
chalk filler which, unfortunately, will rapidly cause the decomposition of
any iron(III) sensitizer. Most contemporary siderotype practitioners have
been driven to the last resort of soaking their chalk-buffered papers in
dilute acid in order to destroy the calcium carbonate before coating them
with sensitizer. Unfortunately many practitioners think that oxalic acid is
suitable for this purpose, but calcium oxalate is just as insoluble as
calcium carbonate and can cause a ‘gritty’ texture. Hydrochloric acid can
be used, but recently it has been found very convenient and effective to
employ a 10% solution of sulphamic acid for this purpose.572
Nonetheless, it is a tedious and undesirable procedure that tends to
degrade the paper surface and strength.
A handmade paper with characteristics that have been specified by the
author for alternative printing is made by Ruscombe Mill at Margaux in
the Haut Médoc of France, and is called ‘Buxton’ paper.573 The Mill
describes Buxton paper as follows:
"It is designed to be chemically compatible with iron-salt-based
sensitizers and be archivally sound. The white paper is produced
from 100% security grade cotton fiber, peroxide bleached with no
optical brightening agents or alkaline buffering agents such as
chalk (calcium carbonate). The cotton is beaten to a high bonding
strength, with only a neutral reactive internal sizing agent [alkyl
ketene dimer] to ensure a good degree of impermeability. The
handmade sheets are heavily pressed several times in both wet and
177
dry states to reduce porosity and to create the desired surface. The
surface is lightly textured due to the impact of the felt mark
derived from the woven woollen felts which are used."574
The greatest imponderable in ‘plain paper’ printing is the effect of

manufacturers’ additives on the reactive chemicals of the sensitizer. Each
commercial paper has its own idiosyncrasies, which may vary from batch
to batch or be changed by the manufacturer without notice. There is no
substitute for personal trial and experiment. Recently, a few commercial
mills have produced some machine-made papers intended for coating
with siderotype sensitizers, in which the calcium carbonate is entirely
omitted. These include Arches Platine, Weston Diploma Parchment, and
Hahnemuehle Platinum Rag, which should all perform well with the
cyanotype process.
The two sides of a sheet of fine paper are often discernably different
in a way that reflects the method of manufacture: the so-called ‘wire’
side (‘screen’ side in the USA) shows, on close examination under bright
light, a very fine geometrical mesh pattern; the ‘felt’ side (‘nap’ side in
the USA) has a random texture. Paper may be coated on either side,
according to your taste. The watermark should be excluded from the
coated area. Always handle paper only by the edges, using both hands to
avoid creasing; never touch the picture area, either before or after
coating, and do not touch the reverse either, because moisture from the
fingers can rapidly diffuse through the sheet.
6.4.2 Using glass rods to coat paper
In place of the traditional brush, the more economic coating implement is
a glass rod or, preferably, thick walled capillary tube, which is usually
manufactured to higher standards of straightness than solid rod. The
external diameter should be between 6 and 12 mm, with a straight
central section equal in length to the width of the coating area. The end
portions of the rod should be bent at an angle, using a powerful gas
burner; a 'bicycle handlebar' shape: \_____/ is simple and ergonomically
effective in use. The centre portion acts as the spreader, and determines
the width of the coated area; the limbs serve as handles. This implement
must be kept scrupulously clean and free of grease film.
Rod-coating paper with sensitizer solution
Follow this procedure:
1 Cut or tear your sheet of paper to an appropriate size that allows
generous borders around the picture area - this not only looks well,
178
but the margins also facilitate handling and protect the image. Mark
the sheet of paper very lightly in pencil to locate the corners of the
area to be coated. Time and trouble may be saved by devising a card
template to guide this marking-up for standard formats. The
dimensions of the coated area should be 1 to 2 cm larger than the
negative to allow for irregularities and to make positioning easy. Avoid
including any watermark within the picture area.
2 Tape or clip the sheet lightly to a very flat level surface: a heavy glass
plate is ideal. Check with a spirit level that the paper is horizontal; this
is critical, so adjust it if necessary. Dust off the surface with a blower
brush.
3 First choose a syringe of appropriate size and ensure its plunger is fully
depressed, then, keeping the nozzle below the liquid surface, draw up
the mixed sensitizer into the syringe, some way above the volume
required. By carefully expelling sensitizer back into the bottle, adjust it
to the required volume: ignore the small air bubble in the syringe and
take the volume reading from the bottom of the plunger. A 10”x8”
print will require about 1.5 cc, or slightly less, and other sizes in
proportion to their area. Try to ‘fine tune’ this volume with experience
of a particular paper, so as to avoid excess sensitizer which may
puddle and crystallise.
4 Expel the liquid slowly and gently from the syringe (use two hands) as
you move it steadily from one side to the other across the width of the
coating area between the pencil marks at the top of the paper; reverse
and go back again if all the liquid is not expelled. For steadiness, touch
the paper lightly with the syringe nozzle and try to form an unbroken
strip of sensitizer, but do not be too slow or fussy - irregularities will
even out in the coating.
5 Take up the spreading rod with one end in each hand, place the
straight central portion onto the paper parallel to, and a little above,
the strip of sensitizer. Then, with slight pressure draw the spreading
rod down into the strip of solution. Pause briefly (a couple of seconds)
while the liquid distributes itself uniformly along the length of the rod,
then steadily draw the strip of solution down the paper, like a tiny tidal
wave running in front of the rod. The rod is not rotated, and very little
pressure need be applied. When you reach the pencil marks at the
bottom of the coated area, hop the rod over the strip of solution and
179
push it back to the top of the coating; then hop over the sensitizer
strip once more and draw it down to repeat the spreading. Just five
passes over the paper will suffice for its surface layer to become
saturated with a uniform coating of sensitizer. The first two passes
should be made quite rapidly (3 or 4 seconds each) to ensure complete
wetting of the surface, and the last three as slowly as possible (10 to
15 seconds each) to allow maximum absorption. At the end of the fifth
pass, drag the spreader well below the picture area, with its excess
sensitizer (when you have fine-tuned the exact volume to use with
your chosen paper, there should be very little waste).
6 Gently lift off the spreader, and soak up any residual liquid left at the
bottom of the coating with the edge of a clean strip of blotting paper,
otherwise crystals may be formed which can damage the negative.
Rinse the spreading rod immediately under the tap.
Occasionally a paper may be encountered that is difficult to coat by

this method, because the strip of sensitizer contracts into ‘beads’ and
blank areas are left in the coating. This may be corrected by applying two
strips of the sensitizer solution, one at the top and one at the bottom of
the coating area; if the continuity of the coat ‘drops out’ on the first pass,
the fresh strip of sensitizer picked up from the bottom will put this right.
It may also be helpful to place a thin sheet of sponge rubber or plastic
foam between the paper and the glass plate, to take up any variation in
thickness (e.g in a handmade paper) or unevenness in the glass rod.
Additional wetting agent (Tween 20 - see below) may also assist
evenness of coating.
Brush coating is the traditional method, for those who prefer a more
‘painterly’ style of coating, or who wish to apply the sensitizer more
selectively. It tends to be rather wasteful - although this is of little
consequence with an inexpensive sensitizer like cyanotype, it is a serious
disadvantage when one comes to coat platinotype. Brushes with metal
ferrules should be avoided. A broad Japanese 'Hake' brush, available from
most art shops, is very suitable. The sensitizer is poured along the paper,
just outside one edge of the picture area, and spread by means of a damp
brush, making long even strokes in a consistent direction to cover the
area, then repeating the brushing at right angles.
Sensitized paper is dried after coating, but the paper should be left
horizontal for a few minutes until the liquid is sufficiently absorbed for
180
the reflective sheen of the surface to disappear. The sheet may then be
hung up to dry in the dark for about an hour before use or left on a
screen. Alternatively some heat may be used - a uniform stream of warm
air at 40 °C for 10 minutes is adequate, and makes little difference to the
final result, provided that heat-dried paper is allowed to ‘rest’ for a half,
to one hour before exposure, otherwise a slight loss of density and
increase of contrast (§7.4.11) may occur. Heating directly, e.g. with a
hairdrier, tends to produce uneven results. The prevailing relative
humidity has very little effect on the results of printing in cyanotype, but
it can profoundly affect some of the other iron-based processes.
6.4.3 Contact-printing frames
With the exception of special cases described in §6.2, some form of
contact-printing frame is usually essential for applying enough pressure
to maintain the close contact between negative and paper that is
desirable with most UV light sources. The simplest option is a sheet of
plate glass (4 mm thick - not 2 mm picture glass, which may bend or
crack under pressure) and a flat baseboard covered with a felt blanket or
sheet of porous plastic to absorb the outgassing (see below) and take up
any surface unevenness. This sandwich is held together by strong clips.
The cyanotype processes give a print-out which is to some extent
solarized - that is, a fully detailed image is formed during the exposure
with reversed shadow tones; but subsequent processing modifies the
appearance of this. It is therefore a great advantage to be able to inspect
the progress of the printing (away from the UV source!) without
destroying the registration between negative and print. This can be
achieved with a hinged-back contact-printing frame of the traditional
design common in the nineteenth-century, when printing-out was the
standard practice. These can still be bought and are often of high
craftsmanship (and cost), but it is easy to construct one from a strong
picture frame of extruded aluminium with a deep rebate, and a hinged
backplate of heavy duty plywood which can be secured under pressure. It
is important that the sensitized paper should be directly backed in the
printing frame with a single loose sheet of felt, not divided and attached
to the backplate, for the following reason.
In all the siderotype processes, including cyanotype, one of the by-
products of the photochemical reaction is carbon dioxide gas (Appendix
III). It is important to the print quality that this gas should have a pathway
for escape by diffusing out from the sensitized surface through the back
181
of the paper, otherwise it may form a ‘bubble’ layer interposed between

sensitized surface and negative. This gap may be wide enough to
degrade the resolution or acutance of the image in regions of high local
contrast, especially if the illumination is a ‘light bed’, rather than nearly a
‘point source’ (see §6.4.4). The permeability of the printing paper should
be sufficient to allow passage of the CO2 gas through to its verso, there
to be absorbed by the felt blanket, or other material permeable to gases,
such as a breathable or porous plastic sheet. It has been observed that
backing the paper with an impervious plastic or rubber sheet may result
in ‘bubbles’ that cause an obvious blurring of the image produced by a
bank of UV fluorescent tubes, although it may not be apparent with
almost point sources, like HID lamps or the sun.
Very large contact prints demand more sophisticated technology such
as a vacuum easel, but the use of a felt blanket as a CO2 absorber behind
the paper sheet is still recommended.
Precious negatives may be protected during contact printing by
interposing a very thin sheet of archival polyester film - such as Mylar™
or Melinex™ - between paper and negative. The negative will be
particularly vulnerable if the paper is fully humidified. Such a sandwich
will, of course, diminish the edge sharpness (acutance) or resolution of
the print, depending on the geometry of the light source, as described in
§6.4.4: if a diffuse ‘light bed’ is used at a close distance, the protective
film must be no thicker than 20 micrometers (0.02 mm); if a small source
at greater distance is used, e.g. a NuArc UV printer or the sun, then more
robust film of thickness 50 to 100 microns (0.05-0.1 mm) can be
tolerated without perceptible diffusion of the image.
If you prefer clean edges and a white border around your image,
rather than the overexposed worked edges which some find fashionable,
then mask the interior of the glass front of the printing frame with a
window of ruby-lith, or other adhesive film that is opaque to UV and blue
light. Such masking brings three technical benefits: it avoids the
evolution of needless quantities of CO2 gas during the exposure that can
degrade the resolution, as described above. It also avoids the formation,
in the wet processing, of large quantities of redundant (non-image)
pigment, that may bleed into and stain the picture area. Moreover, the
unexposed but coated borders will provide a tell-tale check on the
effectiveness of the clearing process, as evidenced by the removal of all
traces of stain. If you do not mask your negative when printing but
expose the entire coated area, then you will never know for certain if the
182
print is properly cleared and the highlights are totally undegraded (see
§7.4.12). These are the disadvantages of ‘showing the brushmarks’ to
prove it’s a handmade print. Connoisseurs should be able to tell that fact
anyway, without it being so obviously displayed in a manner that distracts
from the picture content.
6.4.4 Ultra-violet light sources
Any light source with a substantial ultra-violet content will serve for
printing the iron-based processes. However, sources like 'sun-guns' and
quartz-halogen lamps also emit infra-red radiation, which has the
undesirable effect of heating the paper as the exposure proceeds. The
best sources are fluorescent-coated mercury-discharge tubes, emitting
mostly the so-called ‘long-wave ultra-violet’ (or UVA, with a wavelength
range of 320-400 nm) with a maximum output around a wavelength of
365 nm. Lamps with a peak output around 410 nm are not effective for
cyanotype, because this wavelength corresponds to an absorption
maximum in the spectrum of ferricyanide, which acts as an ‘internal
filter’. There is no advantage, and much additional risk, in employing the
more dangerous ‘short-wave’ ultraviolet mercury lamps which rapidly
damage eyes and living tissue.575 Long-wave UVA lamps are marketed for
graphic arts purposes as well as for domestic sun-tanning.
The geometry of the light source can affect the image acutance in
contact printing: if gaps are formed between negative and sensitized
surface, they may be wide enough to blur the acutance or resolution of
the image, to an extent that depends on the geometry of the light source.
The blur is the width of the penumbra (fuzzy zone) between light and full
shadow at any edge within the image, and depends on the ratio of size to
distance of the light source. If the largest linear dimension of the light
source is w, and its distance from the print is d, then w/d is the aspect
value of the light source. By the geometry of similar triangles:
blur = gap x aspect value
Light Source Aspect value w/d Blur in mm for gap =

0.1-0.5 mm
Sun 0.01 0.001-0.005
NuArc 2125 0.1 0.01-0.05
'Light bed' or open sky 4 0.4-2.0
183
The human eye at its near-point (ca. 250 mm) can resolve ~0.1 mm,
but 0.25 mm is taken to be acceptably 'sharp' in calculating the 'depth of
field' of lenses. Take 0.3 mm of blur as the onset of a 'fuzzy' or 'soft'
image. So the CO2 mentioned above is only likely to cause a visible
problem with 'light bed' sources, and blurring will be most conspicuous
where the local contrast is high.
The following UV sources are listed in ascending order of cost:
1 The sun. This is free, but an uncertain and variable source in many
locations. The sun is nearly a point source giving high acutance
because it subtends a 0.5 degree angle only. However, it also moves
through an arc of 0.5 degrees in 2 minutes, so exposure duration is a
consideration (see §6.2). In intensity the sun is about four times faster
than a small UV source like no. 4. There is a considerable heating
effect which may discolour the highlights in lengthy exposures.
The north summer sky is a diffuse source with a wide aspect like a
‘light bed’, no. 4. Its intensity is about 3 stops (8x) less than direct sun
with about half the speed of no. 4.
2 Small domestic sunlamps such as the Pifco 300 watt UV lamp No. 1012.
These should be used at a distance of about 30 to 50 cm from the
printing frame, providing about half the speed of no. 4.
3 Mercury discharge reprographic lamps, such as the Philips HPR 125W,
used at a distance of about 30 cm from the frame.
4 A convenient unit such as that once marketed by Gordon Audio-Visual

as a Diazo printer. This provides a ‘light bed’ of four Philips fluorescent
tubes, type TLADK 30W/05 UV, distant about 8 cm from the print, and
is adequate for an image size up to 25x30 cm.
5 A similar array can be made quite cheaply by purchasing from a
lighting supply house a commercial luminaire fitting, which is
equipped with all the sockets and control gear to take four 600 mm
fluorescent tubes. UVA tubes marketed as ‘insect attractors’ are
suitable, but not the ‘super actinic’ variety of plant or aquarium light.
Filtered BLB ‘black-light blue’ lamps will work, but less efficiently.
6 A domestic suntan bed - for large prints. The ‘facial solarium’ is a
smaller and more convenient version of this type of source such as the
Philips ‘Cleo’, which uses six UVA fluorescent tubes (actinic/09) with a
power of 20 watts each, and easily covers an area 30x35 cm.
184
7 If cost is no object, a commercial mercury-arc exposure system such as

those manufactured for the graphic arts and screen printing industry.
The NuArc is a popular model for alternative printing and is fitted with
a vacuum easel. The aspect of this UV light source is quite small,
subtending about a 5 degree angle at the platen, so it is good for
maintaining high acutance in the image. Exposure units like the NuArc
are fitted with a built-in light integrator which measures the intensity
continuously and multiplies it by the elapsed time, so accumulating the
overall dose of light which is read out in units of exposure, not units of
time. This is particularly necessary because the emission of the small
high intensity mercury arcs used tends to vary with time.
The considerations regarding image sharpness set out in §6.4.3
should be borne in mind when choosing a light source. Whatever source
you use, be sure to protect your eyes with appropriate UV-absorbing
goggles. All the sources depend on discharges in mercury vapour and
such light sources do not emit their full intensity on switch-on from cold
- they warm up over a period of time while the mercury vapourises, and
the light intensity increases with time, until a steady state is reached. If
that time interval is significant compared with the duration of exposure it
can lead to inconsistencies unless the lamps are 'warmed up' for 5
minutes or so in the absence of any sensitive material. Then they may be
briefly switched off and the print frame with negative and paper fairly
quickly inserted, then the exposure is made using warm lamps.
6.5 Making suitable silver-gelatin negatives

The light sensitivity of iron-based papers like the cyanotype is about a
million times less than that of silver-gelatine enlarging papers, so with
the technology commonly available they cannot be projection-printed,
only contact-printed to achieve a sufficient throughput of UV light. The
reproduction scale of 1:1 leaves the photographer with four options:
• to make very small prints from miniature camera negatives
• to photograph using a large format camera
• to make enlarged silver-gelatin internegatives in the darkroom
• to inkjet-print negatives on transparency film from digital files
Each of these strategies has something to recommend it, but
whichever you adopt, it will be essential to make your negatives to a
density range suited to the printing process. The UVA density ranges
required in the negatives (DD) should be the same as the logarithmic
exposure scales (DlogH) of the iron-based sensitizers, because in contact
185
printing there is no Callier Effect, where light-scattering increases the

contrast.
The traditional cyanotype process has an exposure scale DlogH ~ 0.9-
1.2 (3 or 4 ‘stops’) which matches the density range of camera negatives
developed to a contrast suited to a Grade 3 silver-gelatin enlarging
paper. For the new cyanotype process the DlogH scale may vary typically
from about 1.6, equivalent roughly to a Grade 0 silver-gelatine paper, to
a DlogH around 2.4, which is a far longer scale than any silver-gelatine
paper. To yield a full tonal range in a new cyanotype print, negatives
must be made with a correspondingly long density range. This will be
obtained by giving a good continuous-tone camera film normal
exposure, but developing it for 75% to 100% more than the normal time
period recommended for silver-gelatine enlargement printing. Trial and
error can be lessened by consulting the film manufacturers' published
data for the variation of Contrast Index (Kodak's C.I. or Ilford's G) with
development time or with developer concentration. As a guide, negatives
for cyanotype printing should be developed to a C.I. value in the range
0.7 to 1.3, depending on subject and interpretation. These C.I. values are
more or less incompatible with normal silver printing in an enlarger, so
you cannot expect to use the same negative for both. A convenient
compromise is to make your original negatives in the usual way for silver
printing, ensuring adequate exposure to give plenty of shadow detail,
and then prepare internegatives of higher contrast from them.
Much of what was originally written here 15 years ago about making
enlarged internegatives has become obsolete as a consequence of silver-
gelatin materials ceasing to be manufactured. However, I will let it stand
in the hope that equivalent materials may still be found. The user of
35mm and roll film formats is best advised to enlarge the original camera
negative to give an internegative, which has the added advantage that the
precious original is not put at risk by the rigors of contact printing. There
are several ways to do this:
1. Direct reversal duplicating film makes a negative directly from a
negative. A convenient (but expensive) method was to make
internegatives by enlarging onto Kodak Professional Direct Duplicating
Film Type SO-339, which was available in 5"x4" and 10"x8" formats
only but is now discontinued. Dodging and burning may be carried out
at this stage if desired. Very vigorous development is necessary to
achieve an adequate maximum density in this material. As a starting
186
point for experiment, develop it in the Kodak high contrast developer

D19 (undiluted) for 6 to 8 minutes at 22-24 °C with continuous
agitation. This film has orthochromatic sensitivity, so a red safelight is
necessary. See the Kodak literature for further information. N.B. SO-
339 is a very slow material and needs a heavy exposure under the
enlarger. Even modest enlargements may require a minute's exposure
at full lens aperture. Remember that it is a reversal material - more
exposure gives less density. Remember always to make your
internegatives as mirror images: i.e. place the original negative
emulsion up in the enlarger so the duplicate will make a ‘right-reading’
image when exposed emulsion-to-emulsion on the print.
2. The longer route is via an interpositive, using normal negative-working
continuous tone films such as Kodak Gravure Positive Film 4135 or
Agfa Gevatone N31p or N33p, which are available in large sheet sizes.
The choice of size for the interpositive will depend on your available
enlarger formats and budget, but to minimise the degradation in
quality introduced by the extra copying step, the larger you make the
interpositive, the better. (Contact-printing an interpositive from a
35mm negative can introduce serious dust problems.) The interpositive
should be printed quite heavily to ensure no loss of information; the
highlights should appear somewhat grey and the tonal separation
should still be visible in the shadows. Development to a Contrast Index
value of one is recommended, and dodging and burning may be
carried out at this stage so that the results can be readily judged. The
negative is then made from this interpositive by enlargement or,
preferably, contact, onto the same film, using more vigorous
development to raise the contrast. Kodak HC-110 is a convenient
developer, offering control of contrast by varying concentration as well
as time. Agfa market a continuous-tone orthochromatic sheet film,
GO210p, which has the advantage of a variable contrast that can be
adjusted by yellow or blue filters.
3. By starting with a positive transparency from the camera (such as Agfa
Dia-Direct), an enlarged internegative can be obtained in one step on
the films specified in 2. If the original small-format image is a colour
transparency, projection onto these orthochromatic materials (which
are insensitive to red light) will not faithfully reproduce the tonal
balance. To maintain this, a panchromatic sheet film, e.g. Ilford's FP4
187
or HP5, will be necessary, possibly with a colour head or a filter to

bring the enlarger light source to a daylight colour balance.
4. The sheet film on which a negative is projection-printed may be
reversal-processed to give another negative directly. Directions for
carrying out this rather lengthy procedure are to be found in the
standard photographic manuals. Kodak market a reversal-processing
kit for their T-MAX films.
5. Paper internegatives are a very economical but low-quality option.
Obviously a thin-based paper stock will be needed to avoid long
printing exposure times. The ‘Resin-coated’ (RC) variety is said to work
well, and methods have been suggested for delaminating the emulsion
layer from the base. If exposures become intolerably long, a paper
negative may be waxed to improve its translucency: beeswax dissolved
in turpentine at a concentration of 50% and used warm is a convenient
waxing agent. Alternatively the wax may be ‘ironed in’ between sheets
of blotting paper.
Variations in paper texture, including any lettering on the back, can be
cancelled out by giving white light exposure to the back of the paper,
in addition to the normal image exposure to the emulsion side. A test
strip is necessary to determine this additional compensating fog
exposure.
6.6 Inkjet-printed digital negatives

The modern technology of digital photography, with all its capacities for
image manipulation, provides new possibilities in the realm of negative
preparation for alternative process printing, since the scope for adjusting
the image and its characteristic curve is almost unlimited.
6.6.1 Summary and purpose
Herein follows a stepwise sequence of explicit instructions for making
monochrome digital photographic negatives, using a personal computer
and ink-jet printer, starting from original medium format camera
negatives or transparencies, or from digital picture files. The software
used for this image manipulation was fairly basic: Photoshop CS™
(Version 8) and Epson Scan, using an iMac, running OS 10.4. No
additional software or ‘latest upgrades’ are required. The files are printed
out onto ceramic-coated transparency material (such as Pictorico OHP
Transparency Film or PermaJet Digital Transfer Film)576 using an Epson
188
photo-quality ink-jet printer. This procedure provides internegatives,

suitably matched in optical density at UVA wavelengths, for contact
printing in any of the ‘alternative’ photographic processes. This digital
negative workflow differs significantly from the customary practices in
several respects:
• For calibration, 100-step-tablet negatives having intervals of 1%
relative opacity, are inkjet-printed onto the film, identical in all
their print parameters and materials with the actual negatives to be
output. All the colours of the printer's ink-set are used, blended in
a smooth greyscale.
• The Standard Printing Exposure (SPE) is found by printing these
100-step-tablet negatives in the process of choice with the
standard printing setup, so that the clear filmbase (relative opacity
of 0%) produces ‘nearly’ the maximum print density (Dmax) that
the process is capable of, while some tonal separation is still visible
in the deepest shadows (0-10%).
• The correct ink density range (at UVA wavelengths) for the negative
to match the exposure scale of the chosen process is found by
adjusting the printer driver settings: trial 100-step-tablet negatives
are output and test-printed at the SPE, until the 99% relative
opacity step in the negative prints as ‘just white’ background of the
paper used for the process.
• Having established the 'black and white' end-points for the
process, when making the negatives appropriate for analogue
printing, the image levels are redistributed in Photoshop to ensure
that mid-grey tones also print correctly. This is achieved simply by
using the gamma slider in the Levels window, which applies a
suitable 'generic curve' automatically; making the laborious
derivation of a 'personalised' curve generally unnecessary.
6.6.2 Calibration of equipment and materials
The optical density range of the ink laid down on the film depends on:
• the particular make and model of printer
• the particular ink set being used
• the choice of settings in the printer driver software
• the film substrate used to receive the negative image
If any of these are changed, it is necessary to re-calibrate, as follows
1 Output the digital file for the 100-step-tablet to your printer, loaded
with the identical film that you will use for all your negatives.
189
As a starting point, the ink density range may be approximately

provided by the printer’s ‘premium photo-quality’ media settings for
glossy or semi-glossy paper, preferably an 'Advanced Black & White'
setting if one is provided.
Use 'matt black' ink rather than 'photo black' if there is a choice.
The software for driving modern printers now usually includes a means
of adjusting the maximum ink density; this control may be buried
rather deeply in the options. Record its setting before outputting the
100-step-tablet files. Let the test negative dry and cure overnight
before use, as inks have been observed to change density with time.
2 Assemble the standard printing setup for your chosen process. Take
care that the light source, distance, and other conditions of printing
and wet-processing are maintained constant throughout.
3 Use your 100-step-tablet negative to make trial exposures. Increase
the exposure until the test print is perceptibly overexposed.
['Overexposure' is judged by the analogue photographer's negative-
working criterion of shadow tones beginning to 'block up' and become
indistinguishable – it is NOT judged by the digital photographer's
positive-working criterion of high values being 'blown out'.]
Call the exposure for this test print, E. If you are making timed
exposures to a constant light source, E is the duration in minutes or
seconds. If you employ an exposure system with a light integrator, E
will be measured in the instrument's arbitrary 'exposure units'.
4 Examine the test print carefully when it is dry. The lowest part will
appear dense and ‘blocked up’ with no resolution of the sequential
steps. Compare each step with the lighter one immediately above it -
i.e. the one resulting from 10% less transmittance in the negative
(=10% more opacity, corresponding to a density difference of ~0.05, or
1/6th of a 'stop'). Try to locate the first step, starting from the bottom,
that is just perceptibly resolved in tone from the one above it; this
lower step gives the transmittance in this negative needed to produce
your effective maximum print density. Call this transmittance P %.
5 Calculate your Standard Printing Exposure (SPE) by taking P % of E, the
exposure for the test print made in 3:
SPE = E x P/100
The SPE should apply approximately for all prints made in the chosen
process, with your digital negatives, provided that the light-source and
190
other printing and processing conditions are kept the same.

Note that the SPE is not sharply-defined, but can be varied slightly,
because the effective maximum print density depends on how far up
the shoulder of the characteristic curve of the printing process you
choose to place it. However, the higher the effective maximum density
is placed, the greater proportion of the 256 image levels will need to
be allocated to increasing separation in the shadow tones, so there will
be fewer left for the lighter tones of the image, which compresses the
contrast. Furthermore, if you seek to produce a 'maximum black' by
extending the SPE excessively, the printer inks may not have sufficient
UVA blocking opacity to produce a 'paper base white' at all.
6 Make a test print of the 100-step-tablet negative at the SPE. Inspect
the test print carefully, when fully dry, to find the step that appears
‘just white' - i.e. the first one that has no upper boundary and which
precedes the step with the first perceptible print tone.
7 If no steps print white, adjust the printer driver settings to lay down
more ink when making the negative.
If several steps print white, adjust for less ink in the negative.
8 Return to Step 1 and make a new 100-step-tablet negative using the
new settings, and re-test it through Steps 2 to 7. The ideal aim is to
get 'just white' for the first step of the tablet - that having 1%
transmittance or 99% opacity – but it should be satisfactory to get
within 2% or 3% of this.
Experiment carefully with your printer driver settings: change the
maximum optical density of the ink (a control usually found in the
"colour management" window) and note its value when you make each
new test negative. Print each one in your process with the same SPE
until you come close to the ideal highlight tones.
6.6.3 Fine-tuning negative opacities
This is only necessary in case your printer driver controls do not allow
you to get the first white step close to 1% transmittance or 99% opacity. If
it is still significantly different from this ideal value, then it will be
necessary to reset the maximum opacity of the negative file to a value
equal to the opacity observed for the first white step, before you output it
on transfer film, so that the negative will print the white level correctly.
The Layer Opacity Slider in the layers palette provides a convenient
means of making this fine adjustment. All the other opacity values will
191
then be scaled proportionally. Setting the Layer Opacity to less than 100%
makes the tonal separation a bit coarser in principle, because the tonal
scale of the image will be resolved across fewer than the full 256 levels. If
the printer driver settings result in a white value close to 100%, it’s
unlikely the eye will notice this, so this adjustment in the following
Workflow should be no more than ‘fine-tuning’.
6.6.4 The 'gamma slider' for tonal correction
There usually is a need to redistribute the levels of a digital image file
before it can provide a useful negative. If this is not done, a print made
with the SPE will usually appear too dark in overall tonality, even though
the black and white end-points print correctly. In effect, what is needed
is a transformation of the linear scale of Relative Opacity %, used for
digital files, into the logarithmic scale of Optical Density needed for a
negative to yield a visually acceptable print by any analogue photographic
process having a typical (D/logH) characteristic response.
If we are prepared to accept an approximate transformation, it
becomes unnecessary to use experimentally-derived curves, with all their
burdensome and inaccurate measuring of many experimental points,
because essentially the same result comes from re-setting one parameter
in Photoshop: this parameter is the middle slider in the Levels Histogram
window, which controls the gamma (or contrast) value shown in the
central box, and which always has a default value of 1.0 when the window
is opened. The gamma needs to be increased to a value in the region of
1.8 to 2.2 in the positive image, which – be warned! – will then appear
horribly overexposed in the digital sense, or ‘blown out’ on-screen. This
adjustment usually suffices to re-map the relative distribution of levels
from most positive digital image files to provide, on inversion, a
negative with sufficient density to be printable by analogue processes.
This one simple adjustment with the gamma slider effectively applies a
built-in ‘generic curve’ for transforming all digital negatives to analogue
status. In practice, it is found to agree with individually-derived personal
correction curves for platinum-palladium, which have been scrupulously
plotted by expert workers from point-by-point reflectance density
measurements of their actual step-tablet test prints.
192
Fig 6.3 100-step: Gamma = 2.2 on left; default Gamma = 1.0 on right
193
6.6.5 Workflow for digital negatives

Summary of the sequence of ‘Things that may need to be done’
Phase I: Acquisition and digitization of image

1 Turn off Colour Management in Photoshop
2 Open Image File – if already available, or –
3 Scan Source at appropriate resolution – if necessary – and Save
Phase II: Preparation of positive image file

4 Preliminary Image Adjustments – if needed
5 Crop & Re-size Image
6 Monochromatize – if an RGB colour image
7 Burn & Dodge – if desired (a crude tool)
8 Set Black & White Levels
9 Adjust Local Contrast – in areas where needed
10 Retouch Flaws – where necessary
11 Save ‘Perfected’ Positive Image
Phase III: Preparation of negative image file

12 Apply Tone Correction Curve to Positive by resetting Gamma
13 Reverse Handedness – if image reads correctly on screen
14 Invert Tonality from Positive to Negative
15 Sharpen Image with Unsharp Mask
16 Mask Print Border – if desired (expensive in ink)
17 Flatten Layers & Save ‘Adjusted’ Negative Image File
Phase IV: Printing the negative image onto film

18 Fine-tune Negative Opacity to Match Exposure Scale - if needed
19 Connect Printer, Load Film & Check Inks
20 Select Printer & Page Setup
21 Reduce Image File to 8 bits per channel if necessary
22 Set Printer Driver & Maximum Ink Density appropriately
23 Output the Negative file onto film
24 Number the Negative & allow ink to cure
194
Workflow for Making Digital Negatives in Photoshop CS
Phase I: Acquisition and digitization of image

Key Menu or Tool Procedure
Keys are Shortcuts to Tools and Menus: F6, F7, F8 are re-defined: Edit>Keyboard Shortcuts
1) Turn Off Colour Management in Photoshop

zÝK Photoshop>Colour Settings>In order to turn Colour Management OFF:-
RGB: ‘Monitor RGB – iMac’
CMYK: ‘Euroscale Coated v2’ or suitable choice
Gray: ‘Gray Gamma 1.8’
Spot: ‘0% Dot Gain’ or as low as possible
Colour Management Policies: All OFF
2) Open Image File – if already available

zO File>Browse…>Open> Acquire raw image, preferably 16 bit/channel
Negative or positive, preferably RGB
If colour profiles mismatch Discard any embedded colour profile
3) Scan Source – if necessary

File>Import>Epson Scanner Use Epson Scan App in Professional Mode
Settings: Save the following settings:
Document Type: ‘Film (with Film Holder)’ if negative or slide
Film Type: ‘Positive Film’
Image Type: ‘48-bit Color’
Scanning Quality: Best
Resolution for 360 ppi negs: Final Format: A5 A4 A3
(Use higher if heavy crop) Source Format:
10x8 in. 300 425 600 ppi
5x4 in. 600 900 1250
9x6 cm 900 1300 1850
6x6 cm 1300 1850 2700
35 mm 2100 3050 4350
Check Document Size: 10x8 in. 9.69x7.68 in. 246x195mm
5x4 in. 4.72x3.70 in. 120x94 mm
9x6 cm 3.25x2.20 in. 82x56 mm
6x6 cm 2.20x2.20 in. 56x56 mm
35 mm 1.417x0.945in. 36x24 mm
Target Size: ‘Original’; Trimming: Off
Click Zoom button in Preview panelfor enlarged preview image of 6x9 etc
Set marquee cropping as little as possible
Click Histogram button to open histogram adjustment window
Check Output Levels full range 0 to 255
Set Input Levels B&W sliders on histogram; RGB channels; avoid clipping
Click & Hold Show Output button to see if there’s any clipping bars at ends
Tone Adjustment none: linear, centre slider gamma = 1.00
Auto Adjustments none: turn off any
Densitometer is useful to check B&W levels before scan
Click Close button in histogram window
Click Scan button to scan image and import into Photoshop
zÝS File>Save As>‘Raw Scans’ Save scan as .tif file in folder ‘Raw Scans’
no compression
195
Phase II: Preparation of positive image file

4) Preliminary Image Adjustments – if needed
Double Click ‘Background’ layer to select it in Layers palette
Click OK button in New Layer windowConverts Background to Layer 0 for adjust
IF not suitably oriented:-
Image>Rotate Canvas>… to rotate image for normal viewing
IF a negative:-
zI Image>Adjustments>Invert to invert tonality to positive for easy viewing
z0 (zero) View>Fit on Screen Gives optimum image window size
OR:-
Z Select ‘Zoom’ tool Click ‘Fit on Screen’ button
5) Crop & Re-size Image

Image>Image Size… Check re-sizing parameters correct selected
Tick boxes for Scale Styles; Constrain Props; Resample: Bicubic
C Select ‘Crop’ Tool to frame and re-size the image:-
Click Tool preset picker bar top leftSelect tool to re-size image at 360 ppi
(allowing margin of ~0.15”) A5: 8.0x5.6 in 203x142mm 2880x2016px
A4: 11.4x8.0 in 290x203mm 4104x2880px
A3: 16.3x11.4in 414x290mm 5868x4104px
zR View>Rulers to show rulers, if desired
Ý Shift Lock to replace crop icon cursor with crosshairs
zÝ ; View>Snap to switch ‘snap to edge’ on/off as desired
Click & Drag within image to generate a marquee defining image
z’ View>Show>Grid to generate grid – if useful for placement
« Keys move whole marquee finely: it rotates about
Registration point - can be dragged
IF desired to correct perspective:-
Tick perspective box in toolbarto transform marquee:-
Click & Drag corners of marqueeonto a would-be rectangle for the image
Pull edge centres of marqueeto resize - keeping corners within frame
(only possible if crop undimensioned)
√ Click √ button to execute Crop

OR:-
Ø Click Ø button to cancel and re-set
6) Monochromatize – if an RGB colour image

F8* Image>Adjustments>Channel Mixeropens window to convert to monochrome
Click Load… button for preset channels file
Select ‘mono.cha’ file saved file from Documents:
Click Load L = 30%R + 59%G + 11%B (visual response)
Tick Monochrome box Output Channel: Gray
Click OK button to apply channel mix and close window
OR if desired to set channels manually:-

Click each RGB channel Study them for noise, or use as filter, then:-
¯ Arrow Keys to set channels as preferred; total must sum to 100%
Click OK button to apply channel mix and close window
Image>Mode>Grayscale to revert image to grayscale (smaller files)
196
7) Burn & Dodge – if desired (a crude tool)

O Choose 'Burn' or 'Dodge' Toolas required - alt click switches between
Tool Options Bar>Brush Preset:Brush Diam: 100-1000 pixels; Hardness: 0
Tool Options Bar>Range: Select from: Shadows, Midtones, Highlights
Tool Options Bar>Exposure: Exposure slider: ca. 10-20% for intensity
Click & paint with tool to burn or dodge areas of image. Can repeat
IF correction needed:-
Edit>undo or step back in History Palette and delete
8) Set Black & White Levels

zL Image>Adjustments>Levels opens Histogram window: Set B&W levels:
Alt Click B&W level sliders view effect on image areas for Dmax Dmin points.
Keep all image levels: don’t clip
Click OK button to execute Levels adjustments – or Cancel
9) Adjust Local Contrast – in areas where needed

zalt0 View>Zoom In or Zoom OutMagnify area, and select it:-
EITHER use ‘Magnetic Lasso’ Tool:-
L Choose ‘Magnetic Lasso’ ToolFeather: ~5 px; Antialias: off; Width: 10 px
Edge Contrast: 10-20%; Frequency: 100
Click cursor on edge to select startMove mouse round area to be selected
Can back-delete anchors; Click sets anchor
[] Keys change Width in use Use wider for smooth edges
Click to get marquee when returned to the starting point
OR use ‘Magic Wand’ Tool:-
W Choose ‘Magic Wand’ Tool Contiguous; ±Tolerance = range of levels
Select>Grow Increases range
zL Image>Adjustments>Levels Accesses Histogram window:
Re-set Slider for Centre LevelRe-set gamma <1.00 to increase contrast
Check Preview box to see effectMatch densities using eyedropper if need be
Click OK button in Levels window to apply contrast or Cancel
zD OR ESC to deselect Lasso, or Click within marquee
10) Retouch Flaws – where necessary

zalt0 (zero) Zooms magnification to 100%. Seek defects:
Click & Drag in scroll bars Scan frames from top left, raster-like
J Choose ‘Healing Brush’ Tool Blending mode: Replace; Source: Sampled
Click Brush menu Aligned; Brush Size: Off; Brush Space: 25%
Brush Diameter: ~12 pixels; Hardness: 0
Alt Click cursor on nearby area to locate sampling of replacement density
Click & paint Repair defects with Healing Brush. Continue.
z0 (zero) to revert Zoom to normal when finished
11) Save ‘Perfected’ Positive Image

Layer>Flatten Image Flattens image to Background Layer (small)
zÝS File>Save As>‘Positives360’ Save positive as .tif file in appropriate folder
no compression; Mac byte order
197
Phase III: Preparation of negative image file

12) Apply Tone Correction Curve to Positive by Resetting Gamma
Double Click Background Layer to select it in Layers palette
Click OK button in New Layer windowto convert ‘Background Layer’ to ‘Layer 0’
EITHER use Centre-point slider:-
zL Image>Adjustments>Levels opens Levels window & Histogram
Re-set Slider for Centre LevelSet Gamma to ca. 2.2 Record value used
Check Preview box to see effectPositive image looks ~2 stops overexposed
Click OK button to apply curve & close window – or Cancel
OR use a Stored Curve:-
zM Image>Adjustments>Curves opens adjustment curves control window:-
Alt Click on grid to improve coordinate grid to 10 divisions
Load a selected .acv Curve options stored for adjusting as desired by:
¯ Keys move any selected control pointopen or close highlights, mids or shadows
OR Generate a Custom Curve: - by placing cursor on image area to modify:-
zM Image>Adjustments>Curves opens adjustment curves control window:-
zClick the image area to set a control point on the Curve
¯ Keys darken or lighten this area, respectively
zClick in another area to set a point to be further adjusted likewise
Click & Hold eyedropper in an areato see where that area lies on the Curve
ctrl-Tab moves selected point through control points on Curve
shift ctrl-Tab moves through in reverse
shift-Click enables multiple points to be selected
Click OK button to apply Curve & close window – or Cancel
13) Reverse Handedness – if image reads correctly on screen

F7* Edit>Transform>Flip horizontalLaterally reverses picture to a mirror image
14) Invert Tonality from Positive to Negative

zI Image>Adjustments>Invert Inverts image tonality to negative scale
15) Sharpen Image with Unsharp Mask

F6* Filter>Sharpen>Unsharp MaskAmount: 100–250% usually 200%
Radius: 0.5–1.5 pixels usually 0.8 pixel
Threshold: 2–4 levels usually 3 (skin ~6)
Tick Preview box to see effect of USM when image clicked
Click OK button in USM window to apply USM – or Cancel
16) Mask Print Border – if desired (expensive in ink)

Image>Canvas Size… 8.0x5.6 in A5: 8.28x5.85 in 210x149 mm
11.4x8.0 in A4: 11.7x8.28 in 297x210 mm
16.3x11.4in A3: 16.65x11.7 in 421x297 mm
W Choose ‘Magic Wand’ Tool Antialiased; Contiguous; Tolerance 0
Click with Wand in border area to select border area to be masked
ÝF5 Edit>Fill… Colour: Black; Blending Mode: Normal
zD to deselect border mask
17) Flatten Layers & Save ‘Adjusted’ Negative Image

Layer>Flatten Image Flattens image to Background Layer (small)
zÝS File>Save As>‘Diginegs’ Save negative as Tiff file in folder ‘Diginegs’
Click OK button Tiff Options: No compression; Mac byte order
198
Phase IV: Printing negative image onto film

18) Fine-tune Negative Opacity to Match Exposure Scale - if needed
If White ~99% Opacity skip this whole box:ink density will be adjusted in printer settings 22)
Double Click Background Layer to select it in Layers palette
Click OK button in New Layer windowto convert ‘Background Layer’ to ‘Layer 0’
Click Image Layer Opacity button to reveal Opacity Slider in Layers palette
Adjust Image Layer Opacity slider to a value matching process to be used:
step % prints just white when 0 % prints near Dmax
Record the step% value used See: calibration of equipment & materials
Click Image Layer Opacity button to close the Opacity Slider
Layer>Flatten Image Flattens image to Background Layer
19) Connect Printer, Load Film & Check Inks

Connect Printer & Switch ON Load Printer with 1 sheet of transfer film
Launch Printer Utility Ensure correct Printer selected from List
Click Status Monitor to check ink levels & recharge if needbe
Choose Nozzle check if printer long-dormant: head clean
Close Printer Utility window
20) Select Printer & Page Setup

zÝP File>Page Setup… Format for printer & paper size & Aspect
Click OK button in Page Setup window
Click in Doc: edge area to see fit of image on page
Alt Click in edge area to see size of image, resolution, etc.
21) Reduce Image File to 8 Bits per Channel if necessary

Image>Mode>8 bits/channelto prepare for sending to printer – if 8 bit
22) Set Printer Driver & Maximum Ink Density & Output the Negative
zaltP File>Print with Preview…Check page settings:- Check centred image
Colour Management: Document; Profile: Same as Source
Click Print… button to access Print Settings window:-
Printer: Check printer choice correct
Presets: Use saved Printer Driver settings file:
Print Settings>Media Type: Appropriate paper choice (photo or matte)
Colour: Colour ON
Advanced Mode: Premium Photo Quality: highest resoln dpi
High speed OFF; Mirror image OFF
Colour Management: Mode: Epson Standard sRGB
Colour Controls: All sliders centred; gamma = 2.2
Advanced B&W mode: Neutral & ‘Normal’ tone
Adjust maximum ink density to give White~99%
Extensions: Normal paper (not thick)
Paper Configuration: Colour density & head drying time – defaults
Click Print button to print the negative on transparency film
Don't 'Save Changes' on closing file
23) Number the Negative & Allow Ink to Cure

Number the negative indeliblyAllow negative to dry in dust-free environment for at least 12
hours before use. Do not stack.
199
6.6.6 Short version of the workflow

The previous detailed workflow was designed to cope with most
eventualities, but it has necessarily become so lengthy that it is “hard to
see the wood for the trees”. If we are working with preformatted and
‘perfected’ positive images, the following are the only essential steps
needed to prepare a tonally-corrected and density-adjusted negative.
Numbers refer to the original ‘Things that may need to be done’ boxes.
1 zÝK Photoshop>Colour Settings> Colour management OFF

4 Double Click ‘Background’ in Layers palette
Click OK New Layer 0
6 F8* Image>Adjustments>Channel Mixer
Load ‘mono.cha’ file: L = 30%R + 59%G + 11%B
Image>Mode>Greyscale
8 zL Image>Adjustments>Levels
Alt Click Sliders & Set B&W levels
12 Set Middle Slider to Gamma ≈ 1.8 - 2.4 for desired contrast
13 F7* Edit>Transform>Flip horizontal
14 zI Image>Adjustments>Invert
15 F6* Filter>Sharpen>Unsharp Mask
18 IF maximum ink density cannot give White for 99% Opacity step (22):
Adjust Image Layer Opacity to % Opacity of first 'white step'
Layer>Flatten Image
19 Launch Printer Utility: Ensure Printer connected, loaded, & turned on
Click Status Monitor to check ink levels etc
20 zÝP File>Page Setup… Format for Printer, Page, Size & aspect ratio
22 zaltP File>Print with Preview…
Click Print… Print Settings: Advanced B&W mode, Neutral tone, Normal
Adjust Maximum Ink Density to the value that gives ~99% Opacity in the
negative printing as the 'white step'.
23 Click Print
* reassigned function keys
200
7 Formulary for Cyanotype Sensitizers

Gold is for the Mistress,
Silver for the maid,
Copper for the craftsman
Cunning at his trade.
“Good,” said the Baron, sitting in his hall,
“But Iron, cold Iron, is master of them all.”
Rudyard Kipling (1865-1936)
Cold Iron, from Rewards and Fairies (1910)
This Chapter offers a choice of three cyanotype formulae:

• first, the traditional cyanotype, which is easy to prepare,
inexpensive, of uncertain quality, and non-toxic;
• second, a widely-recommended elaboration of this recipe; and
• third, my new formulation that provides finer results but
demands a little more care and expense in the preparation of
the sensitizer, involving less benign chemicals.
The traditional recipes are fully described in the useful practical guide
for beginners by Malin and Gary Fabbri,577 and in many other sources.
They are well-suited in their short exposure scale to the small density
range of normal negatives, such as would be made for printing in
commercial gelatin-silver halide papers of contrast grade 3, or in gum
dichromate pigment processes. In contrast (literally!), the new cyanotype
process has a wide exposure scale that provides a good match to the
long density range of negatives specifically prepared for salted paper,
argyrotype, platinum-palladium, chrysotype and similar siderotype
alternative processes.578 The contrast of your negatives should be
adapted to your preferred choice of process, not vice versa.
7.1 The traditional cyanotype process

This is the time-honoured formulation; it is made up as two separate
solutions, which are combined just before coating.
Sensitizer chemicals needed
Ammonium iron(III) citrate (green variety) 25 g
Potassium ferricyanide K3[Fe(CN)6] 10 g
Tween 20 (non ionic surfactant) 20% v/v solution 1 cc
Purified water 200 cc
201
The only innovation suggested here is the addition of Tween 20

surfactant, which can greatly improve the quality of the print with many
papers. The chemicals are not dangerous and can be safely handled by
children under supervision. Ammonium iron(III) citrate is also known as
ammonium ferric citrate; and is a very variable substance: the ‘green’
form with an iron content of about 14% is preferable to the ‘brown’ form
with 19% or more iron. However, any variety will work up to a point. If the
‘brown’ form is the only one available it may be advantageous to add
about one third of its weight (ca. 8 g) of citric acid to the solution, to
bring the citrate : iron ratio close to that for the green form (see
Appendix III.5). For the potassium ferricyanide GPR Grade (General
Purpose Reagent: 98-99% purity) is adequate.
7.1.1 Making up the sensitizer solutions
Two stock solutions A & B are required; their concentrations are not
critical so they do not need to be made up with great precision:
Stock solution A:
Dissolve 25 g of the solid ammonium iron(III) citrate in ca. 70 cc of
purified water at room temperature and make up to 100 cc.
Within a week or two in many environments, the surface of this solution
will be covered by a growth of mould. This may be inhibited at the outset
by adding a few crystals of thymol to the bottle, which float on the
surface without dissolving, and should simply be avoided in extracting a
sample of the sensitizer. (N.B. Thymol is a harmful chemical.)
Stock solution B:
Dissolve 10 g of the solid potassium ferricyanide in ca. 80 cc of purified
water at room temperature and make up to 100 cc.
These solutions must be kept in brown bottles, in a cupboard or box.
7.1.2 Mixing and coating
Before use, mix equal volumes of A and B to give a total volume of
sensitizer appropriate for the amount of coating to be done. The
surfactant Tween 20 is added at this point – see below. The mixed
sensitizer only has a relatively short life. It is best to dedicate a separate,
labelled syringe or pipette to each solution, and to use a third one for
applying the mixture, in order to avoid cross-contamination during the
work. As a guide to the volume needed for coating, about 1.5 cc should
suffice for a 10x8 inch print, coated by rod; but brush coating may
consume possibly twice this.
202
Although it is not part of the traditional formula, I would strongly

recommend adding a few drops (2 to 5) of the 20% solution of Tween 20
wetting agent to each 10 cc of sensitizer, to improve its absorption by
the paper fibres; the appropriate amount depends on the porosity of the
chosen make of paper and should be arrived at by trial.
7.1.3 Printing exposure
Expose to sunlight or a UV lamp: the image prints out, with the
uncovered regions of sensitizer first darkening to blue and then reversing
back to a pale blue-grey. The image shadows should be somewhat
reversed too, giving it a solarized look, and the highlights should be
distinctly green. The exposure is correct when the image appears rather
‘blocked up’. Lamp exposure time may be around 20 to 30 minutes.
Proponents of this version of cyanotype suggest that it is almost
impossible to overexpose, because so much image substance will be lost
in the wet processing!
7.1.4 Wet processing and reoxidation
Immerse the exposed print in gently running water (or several changes of
static water), face down, until the yellow sensitizer has entirely
disappeared from highlight areas (20 minutes should suffice). There will
be considerable leaching-out of blue pigment, so the extended washing
necessary to fully clear the paper will significantly ‘reduce’ the image, and
truncate the tonal scale in the high values. The ‘reversed’ shadow tones
will darken with time and the image colour will slowly intensify during
drying over several hours, due to aerial oxidation of Prussian white to
Prussian blue. If you wish to see the full density immediately, then
immerse the print for about half a minute in dilute (0.3%) hydrogen
peroxide (see §7.4.9) before the final wash. This does not make any
difference to the final densities. Under no circumstances should you use a
bath of potassium or ammonium dichromate for this reoxidation,
although it has been recommeded in the past (see §7.6). Hang the print
to dry at room temperature, or in a uniform stream of warm air.
7.2 An improved cyanotype recipe

A widely-circulated ‘improved’ formula follows, but it is not known where
it originated.579 It is claimed that the highlight areas remain whiter.
However, if you are prepared to go to this degree of trouble to introduce
an admixture of oxalate and dichromate, which are poisonous, it would
probably be more worthwhile to employ my new cyanotype method,
203
which uses oxalate entirely. This sensitizer should not be placed in the
hands of children.
Sensitizer chemicals needed (GPR Grade (98-99%) is adequate)
Ammonium iron(III) citrate (green variety) 27 g
Oxalic acid (COOH)2.2H2O 1 g
Ammonium dichromate (NH4)2Cr2O7 0.2 g
Distilled water 200 cc
Preparation of sensitizer
Solution A:
Dissolve 25 g of the solid ammonium iron(III) citrate in ca. 70 cc of
distilled water at room temperature; add 0.5 g oxalic acid, stir well to
dissolve, and make up with distilled water to a final volume of 100 cc.
Solution B:
Dissolve 9 g of the solid potassium ferricyanide in ca. 80 cc of distilled
water, add 0.5 g of oxalic acid and 0.2 g of ammonium dichromate, stir
well to dissolve and make up to 100 cc with distilled water.
These solutions must be stored in brown bottles in a dark box.
Mix equal volumes of solutions A and B just before use.
7.3 Process shortcomings and their remedy

Herschel’s standard negative-working cyanotype formula has such an
elegant simplicity and economy that it may seem presumptuous to seek
to improve on it. In 1906 John Tennant, the editor of The Photo-
Miniature, could confidently write:
“...the blue print process reached finality long ago as far as any
important improvement is concerned, and there is little that is new
to tell about it.” 580
But any contemporary practitioner today, accustomed to siderotype

processes such as palladiotype, who tries the traditional cyanotype
formulation for pictorial purposes is likely to notice some of the following
comparative drawbacks:
1 Ammonium ferric citrate is a notoriously ill-characterised chemical and
can vary widely in composition, depending on the source. Even after
170 years of chemical manufactury, it continues to cause problems of
unpredictability, inconsistency and poor keeping qualities. No two
practitioners can be sure that they are using the same chemical.581
204
2 The two ingredients have to be stored as separate stock solutions,

because when mixed the resulting sensitizer has an inconveniently
brief shelf-life. The solution of ammonium ferric citrate proves to be
an excellent nutrient medium for the growth of moulds, which will
usually cover the surface within a week or two; a preservative fungicide
such as thymol can be added to prevent this.
3 The traditional sensitizer is poorly absorbed by cellulose paper fibres
and some tends to remain on the surface or in the pores; being
deliquescent, it may cause a tackiness which can damage negatives if
the humidity is high. Much of the image washes off on wet processing.
Adding a surfactant to penetrate the fibres can help to counter this.
4 Printing is rather slow compared with other iron-based processes such
as the palladiotype; exposures of thirty minutes or more to a typical
ultraviolet light source are not unusual for traditional cyanotype,
compared with about two minutes for palladiotype.
5 A significant proportion of the image substance, ‘soluble’ potassium
Prussian blue, tends to be peptized and wash out during the wet
processing, due to poor absorption and lackof retention by the paper
fibres. This results in a serious loss of gradation in the high values,
truncating the tonal scale, and conferring an artificially high contrast.
The only remedy that is offered to counter this drastic weakening of
the image is to overexpose it grossly, which adversely affects the
quality and the ease of judging exposure. Some workers even
recommend double coating with sensitizer.
6 In heavily exposed print areas, the traditional sensitizer tends to
produce excess ferrous iron which diffuses into, and stains adjacent
highlight areas with Prussian blue. This is the defect already referred to
as ‘bleeding’.
7 The traditional process has an exposure scale of only about 0.9 – three
‘stops’ - comparable with that of a Grade 3 or 4 silver-gelatin printing
paper so, like the gum dichromate process, it requires only a very short
density range in the negative. The process is quite unable to render the
long density range (2 or more) of the negatives needed for other
siderotype processes.
All the drawbacks listed above can be overcome, in principle, by replacing
ammonium ferric citrate with ammonium ferric oxalate as the
photosensitive agent, so employing the ferric salt that was shunned, or
205
overlooked, by Herschel (§2.10) - whichever may have been the case.

Ammonium ferric oxalate offers the following remedies for the
disadvantageous problems listed above:
1 It is a well-characterised substance, widely available all over the world
in an inexpensive, pure crystalline form that is chemically consistent,
keeps well in the dark, and is highly soluble in water.
2 It does not permit any mould to grow on the solution. The mixed
sensitizer has a shelf-life of several years.
3 Its solution penetrates the fibres of cellulose paper much more readily
than the citrate, and it mordants better onto fabrics. It is non-
deliquescent. Image substance is better retained.
4 It is much more light-sensitive than the citrate by a factor of 4 to 8, so
exposures are correspondingly shorter, by two or three ‘stops’. This
can be important when sun-printing 3-dimensional objects (§6.2), or
working with poor light sources.
5 It yields an ‘ammonium blue’ of very good colour, which is more
resistant to peptization and alkalies. There is little loss of image
substance on wet processing, so it renders delicate gradation in the
high values and a high Dmax verging on black. It is readily toned by a
variety of reagents.
6 It can be mixed in chemically equivalent amounts with ferricyanide to
avoid excess iron(II) and therefore suppress image ‘bleeding’.
7 Because there is little or no image loss the exposure scale of the new
process is around 2.2 – an excellent match for negatives of long
density range, also intended for salt printing, platino-palladiotype,
argyrotype or chrysotype. One negative can serve all processes.
However, as the early experimentalists discovered, see §4.4.4, the use

of ammonium ferric oxalate introduces one difficulty: if it is mixed with
potassium ferricyanide solution at concentrations high enough to provide
a good ‘pictorial’ sensitizer,582 the sparingly soluble salt potassium ferric
oxalate tends to crystallize out, rendering the mixture useless as a
sensitizer.583 A ‘gritty’ sensitizer produces pale spots on the print, and
crystallization may also occur within the coated paper, giving rise to quite
pretty, but totally unwanted fern-like patterns.
206
It is true that ammonium ferric oxalate was used in all the formulae
for the later commercial blueprint papers, as described above in §4.10,
but this was only made possible by keeping them sufficiently dilute in
potassium ferricyanide, 5% or less, to avoid this crystallization; so they
only yielded a maximum density rather too low for pictorial purposes.
In theory, there is an obvious answer to this crystallization problem: to
use ammonium ferricyanide instead of potassium ferricyanide; but the
former is not commercially available, is described as a “rare chemical”
and is rather troublesome to make.584 In 1994, I developed a new, simple
procedure for preparing an ammonium salt cyanotype sensitizer that
overcomes all the drawbacks besetting the traditional method.
The desired ammonium ferricyanide is prepared in situ by the simple
expedient of using the low solubility of potassium ferric oxalate to
eliminate most of the potassium ions from the solution. This is achieved
by mixing very concentrated, hot solutions of ammonium ferric oxalate
and potassium ferricyanide. When the mixture cools, the sparingly
soluble potassium ferric oxalate crystallises out, is filtered off and
rejected, leaving a solution of ammonium ferric oxalate and ammonium
ferricyanide, in the necessary proportions for making Prussian blue when
the former is photoreduced. Of course the separation of potassium from
ammonium cations is not 100%: some K+ remains in the solution and a
little NH4+ will be in the crystalline product, so the conditions of time and
temperature adopted for the crystallization in the following recipe must
represent a compromise.
The appropriate molar quantities are calculated as follows: what little
evidence we have (IR spectra – see Appendix II.7) suggests that the
cyanotype product Prussian blue may be closer to the "insoluble" form
Fe4[Fe(CN)6]3 i.e. it contains a 4 : 3 molar ratio of Fe(III) : Fe(II), rather than
the 1 : 1 molar ratio of Fe(III) : Fe(II) which assumes that the form of
Prussian Blue precipitated is (NH4)[FeFe(CN)6]. What follows makes that
assumption, but the relative proportions are not critical.
Precipitation of most of the potassium ions requires a molar ratio of
1 : 1 for the two components, so in the former case the overall molar
ratio of ammonium ferric oxalate to potassium ferricyanide needed for
both precipitation and for the photochemistry would be 7 : 3. Taking the
relative molecular masses of the two components, this molar ratio
corresponds to a weight ratio of:
7 x 428.07 : 3 x 329.26 = 2996.49 : 987.78 = 3.034 : 1
207
which comes to a rounded weight ratio of 3 : 1 for the two

components, and accounts for the formula used in §7.4 below. The
filtrate is then diluted appropriately to provide a stable, pre-mixed
sensitizer solution, which contains 0.4 molar ammonium ferric oxalate
and 0.3 molar ammonium ferricyanide, and which offers much greater
printing speed and a long storage life without mould growth. The
coatings are non-deliquescent, unlike those made from ferricitrate, which
reduces the likelyhood of damage to negatives during contact printing.
The image colour and gradation are finer than the traditional process,
and the ‘ammonium’ Prussian blue has less tendency to peptize, so the
print-out image is a good indication of the final print. The only
disadvantages of this method are a slightly more complicated method of
preparation, (which would not be the case if ammonium ferricyanide
could be bought – see above!) a moderate toxicity of the sensitizer
solution (but no worse than kallitype or palladiotype), and a somewhat
greater expense for the chemicals.585 Despite these slight complications,
a number of competent practitioners have found that the results more
than repay the trouble taken. The preparation and procedure are
described in detail next.
208
7.4 The new cyanotype process

Chemicals needed for making and processing 100 cc of sensitizer
(sufficient for ~65+ prints 10x8 in.)
Ammonium iron(III) oxalate (NH4)3[Fe(C2O4)3].3H2O 30 g
Ammonium dichromate (NH4)2Cr2O7 0.1 g
Tween 20 0.25 cc
Citric acid (optional) 2g
Nitric acid, concentrated (~65%) or hydrochloric acid 100 cc
Purified water 100 cc
Apparatus for making 100 cc of sensitizer

Pyrex glass beakers 2 x 100 cc
Measuring cylinder 100 cc
Scales or chemical balance sensitive to 0.1 g
Glass stirring rod
Conical filter funnel ca. 8-10 cm diameter
Filter paper Whatman Grade 1 ca. 15-20 cm diameter
Brown glass bottle 100 cc
Hotplate (or bath of very hot water)
Thermometer 0-100 ºC
Tungsten lighting to work under, not fluorescent or daylight.
Equipment and materials needed for cyanotype printing (see §6.4)

Paper (100% cotton or linen, sized but unbuffered)
Glass coating rod or brush
Glass plate
Spirit level
Drafting tape or clips
Blotting paper strips
Plastic syringes or pipettes, calibrated 2 cc and 5 cc
Printing frame (hinged-back)
UVA light source
Timer
Plastic measuring jug 1-2 litre; Stirrer
Processing dishes (2)
Print tongs or plastic gloves
Drying line and pegs or drying screen
209
7.4.1 Making up the sensitizer

N.B. This sequence of instructions should be carried out under dim
lighting (40-60 watt tungsten), not fluorescent tubes or daylight.
1 Weigh 10 g of potassium ferricyanide into a 100 cc pyrex glass beaker.
Add exactly 20 cc of purified water using a measuring cylinder.586
Heat to ca. 70 °C (158 °F) and stir until all the solid is dissolved.
Keep the solution hot.
(A thermostatted hotplate is best for heating the solutions, but a basin
of very hot water will do.)
2 Meanwhile, weigh 30 g of ammonium iron(III) oxalate into a second
(100 cc) pyrex glass beaker. Add exactly 30 cc of purified water using
a measuring cylinder. Heat to ca. 50 °C (122 °F) and stir until the green
solid is dissolved.
(Ignore any small residue of colourless crystals.)
3 Add 0.1 g of ammonium dichromate to (2) and let it dissolve.
(Potassium dichromate may be used alternatively. This small amount of
dichromate may be omitted to satisfy prevailing safety regulations, but
the contrast and storage-life of the sensitizer will be diminished, and
some Prussian blue may be formed as a sediment.)
4 Add the hot potassium ferricyanide solution (1) to the ammonium
iron(III) oxalate solution (2), stir well, and leave the mixture in a dark
place to cool to room temperature (ca. 20 °C, 68 °F) and crystallize.
(For this quantity of solution, the cooling and crystallization will take at
least two hours, but it may be better left overnight, especially if a
larger scale of preparation is used.)
5 Separate the green liquid from the green crystals by decanting and
filtration: use a conical filter funnel with a Whatman grade 1 filter
paper, running into a 100 cc measuring cylinder.
(The volume of filtrate should be ca. 60 cc. The yield of green crystals
should be ca. 15 g. Dispose of these crystals safely by washing them
away with hot running water. They are not an environmental hazard.)
6 Make up the filtered solution with purified water to a final volume of
100 cc in the measuring cylinder. Mix well and transfer the sensitizer
solution to a well-stoppered, labelled and dated brown bottle.
(The sensitizer solution is toxic if ingested and it will stain skin, etc.
Stored in a cool dark place, its shelf-life is six months to a year without
dichromate, and several years with dichromate added.)
210
1. 10 g potassium ferricyanide + 20 cc water
2. 30 g ammonium iron oxalate + 30 cc water
211
2. 1 and 2 dissolving 3. 0.1 g ammonium dichromate
4. Add 2 to 1
212
4. Mixture cooling & crystallising in dark
5. Filtration and making up to 100 cc
213
7.4.2 Use of dichromate

If your scales cannot weigh out 0.1 g of the solid dichromate, then make
a 20% w/v solution of ammonium dichromate by dissolving 4 g of the
solid in purified water at room temperature, and making up to a final
volume of 20 cc. Mix thoroughly. Add 0.5 cc of this solution to the
ammonium iron(III) oxalate solution in 2, above.
You should recognise that this optional use of dichromate does
introduce a toxic substance into the working environment (see Notes on
Hazards §7.5) but in very small amount, especially compared with other
users (e.g. gum-dichromate printers). Ammonium dichromate is a strong
oxidising agent, intended here as a preservative: its addition can prevent
any impurities in your chemicals from causing the formation of Prussian
blue in the sensitizer over time, and usually gives it a shelf-life of several
years. However, the presence of this reagent will also tend to increase the
contrast of the sensitizer significantly (see the D/logH curves in Figs 7.1
and 7.2 below). Omitting the dichromate altogether may shorten the
shelf-life of the sensitizer to a few months.
7.4.3 Use of surfactant
The new cyanotype sensitizer solution is much more readily absorbed by
paper fibres than the traditional sensitizer using ammonium ferric citrate,
so the addition of a wetting agent is usually quite unnecessary. With a
few, hard-sized papers such as Buxton, however, the wetting agent can
improve the ease of coating and the retention of Prussian blue by the
paper fibres. In such cases, I prefer Tween 20 (a non-ionic surfactant)
which may be added to the sensitizer solution to produce a final
concentration of ca. 0.1% to 0.25%. A stock solution of concentration 20%
v/v is useful for this: if you find it necessary, add one or two drops (0.05
to 0.1 cc) per 10 cc of sensitizer and mix well just before coating. The
appropriate amount will depend upon the paper, so do not add it to the
bulk of the stock sensitizer solution. Tween may interact unfavourably
with some gelatin-sized papers. The Tween solution is not stable
indefinitely and should be replaced if it shows a deposit.
7.4.4 Choice of papers
Use only papers that are not alkaline-buffered with chalk (calcium
carbonate). Any paper that claims it conforms to the ISO 9706 standard
for permanent paper will contain alkali (see §9.4.5), and alkalies are
hostile to cyanotype chemistry. The best results will be obtained on
unbuffered papers such as:
214
• ‘Buxton’ or 'Herschel' handmade by Ruscombe Mill

• Weston Diploma Parchment
• Crane’s Platinotype
• Arches Platine
• Atlantis Silversafe Photostore
• Wyndstone Vellum
• Awagami Masa
• Canson Lavis Technique (is ISO 9706 compliant, but works well
with added citric acid)
If buffered papers are unavoidable, such as Fabriano Artistico, or
Whatman Watercolour, the chalk should be destroyed by pre-treated in a
bath of dilute (5% v/v) hydrochloric acid, or 5% - 10% w/v sulphamic acid,
for about 10 minutes, then washed.
For prints up to 10 x 8 in. or A4 in size, a paper weight of 160 gsm
(grams per square meter, g/m2) is adequate. For larger prints of A3 size,
a heavier weight of 240 gsm, or more, will minimise cockling and bellying
of the coated sheet due to the stresses set up by the hydroexpansion of
the cellulose fibres in the wetted area. The sheet will contact the negative
better, and be more robust in wet handling.
7.4.5 Drying
Let the sensitized paper dry at room temperature in the dark for about an
hour. Shorter drying times are possible, but very humid paper may
damage precious silver-gelatin negatives, and not lie flat due to fibre
swelling.
Alternatively, allow 5-10 minutes for the sensitizer to soak in, until
the paper surface loses its reflective sheen, then heat-dry it with an air
stream at about 40 °C for ca. 5 minutes. The heat-drying of a cyanotype
paper does not appear to influence image colour but it may increase
contrast slightly, see §7.4.11. Rapid drying can reduce any chemical
fogging due to impurity in the paper; but note that over-rapid drying may
worsen the loss of image substance during the wet-processing
procedure.
The storage life of coated paper depends on the purity of the paper
base, as mentioned above, so use the sensitized paper within a few hours
of coating, if possible. It will keep longer in a cool, dark desiccated
enclosure. The coated side should remain light yellow: if it turns green or
blue the highlights are chemically fogged, so either reject it and find a
better paper, or try adding citric acid as described below, for your next
215
coatings. The new cyanotype sensitizer provides a sensitive test of paper

quality, especially if the coated paper is left for some hours in the dark at
normal relative humidity: any change of the yellow coating towards a
green or – worse - blue colour signals the presence of impurities or
additives in the paper that are hostile to this process (and possibly to
other siderotype processes as well).
7.4.6 Citric acid prevents fogging
Chemical fogging (“greening”) of the coating due to paper impurities may
often be prevented by adding citric acid to the sensitizer, before coating,
to a final concentration of ca. 2%. Add one drop (0.05 cc) of a 40% w/v
solution of citric acid per cc of sensitizer. Do not add citric acid to the
stock of sensitizer, because it will shorten the shelf-life.
7.4.7 Negatives and exposure
Expose the sensitized paper within a few hours of coating, if possible.
Negatives should have a long density range (in the UV): at least 1.8 and
as much as 2.4, to produce a full tonal range in the print - as for my
other siderotype processes.
Whether your light source is the sun or a UVA lamp (§6.4.4), exposure
will be much shorter than that needed for the traditional cyanotype
recipes - this new sensitizer requires about two to five minutes exposure
under a typical light source. With an 800 W HID UV lamp and digital
negatives my exposure time is ca. 60 seconds.
Since this is substantially a print-out process, a traditional hinged-
back contact printing frame (§6.4.3) will enable inspection of the desired
result: the exposure is continued until the high values appear green, the
mid-tones are blue, and the shadow tones are substantially reversed to a
pale grey-blue, giving the image a solarized appearance. The exposed
print may be processed immediately. If the maximum detail in the high
values is desired, the print may be left in the dark at average humidity for
about 20 hours ‘darktime’ before wet processing, but the increase in
tonal scale is slight – about one stop - see the results in §7.4.11.
7.4.8 Wet process and contrast control
It is important to use a dilute mineral acid for the first development bath
to achieve a consistent outcome:587 a deep maximum density and a well-
shaped characteristic curve, having a very long printing exposure range
of about 2.1 to 2.4 density units (7 or 8 ‘stops’) - see Figs. 7.1 and 7.2.
The print should be processed initially by immersion in a bath of
dilute nitric acid (ca. 0.1% to 1% v/v) for 5-10 minutes. For safety in
216
handling, it is suggested that a stock solution of 10% v/v nitric acid is

made up, by diluting the concentrated (~65% w/v) acid 10-fold (i.e. 100
cc of acid added to water and made up to 1 litre). Take care to avoid the
corrosive fumes! This intermediate solution is then diluted a further 10 to
100 times before use. The final 100-1000x diluted bath is not
dangerous, but you should still protect your eyes. By varying the acid
concentration of this processing bath between 0.1% and 1% v/v, some
control of the contrast may be achieved at the development stage, with
ΔD ~ 2.6 – 2.1. Fig 7.1 shows the characteristic (D/logH) curves for a new
cyanotype sensitizer (without added dichromate) developed in water and
in 1% nitric acid; the result using 1% hydrochloric acid is identical. These
development baths tend to accumulate Prussian blue, which will
eventually begin to stain the paper, so they should be replaced after a
few prints have passed through them; typically, 1 litre will process five to
ten 10 x 8 inch prints. If there are safety objections to using nitric acid,
then 0.1% to 1% v/v hydrochloric acid may be used. If there are safety
objections to this, then 1% to 5% w/v sulphamic acid could be used, or as
a last resort, 1% to 5% w/v citric acid. Acetic acid is not recommended
because it tends to form insoluble coloured basic ferric acetate.
The softest gradation will be obtained, ΔD ~ 2.6, with a more open
gradation in the shadows but a somewhat reduced maximum density, if
the exposed paper is simply processed with gently running tap water for
at least twenty minutes; alternatively, several static baths of water may be
used. To avoid undue loss of image substance, the water should not be
significantly alkaline, preferably having a pH below 7, and a spoonful of
citric acid may be added to ensure this. However, this processing method
can still result in some loss of image, depending on the paper and sizing
combination used, and the shape of the characteristic curve that it yields
is rather unusual and may not suit all negatives – see Fig 7.1.
7.4.9 Washing and reoxidation
Finally wash the print gently in running water for at least 20 minutes.
Alkaline water (pH >7) must not be used, nor hard water, containing
calcium salts, which will damage the Prussian blue image. The criterion
for completed processing is the total removal of the yellow stain of
sensitizer from unexposed areas - the print should be held up to a bluish
light, or viewed through a blue optical filter, to check that no yellow
stains remain in the interior of the paper; if any stain persists, further
217
washing is required. Unlike prints made by the traditional recipe, there

should be very little loss of image substance during this washing.
Reflectance Density (red)
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
3 2 1
–log(relative exposure)
Fig. 7.1 Characteristic curve for New Cyanotype without dichromate.
1% acid developed (left) and water developed (right)
The reversed shadow tones usually begin to regain their values quite
rapidly during the wet processing (nitric acid particularly assists this), and
they will fully recover during drying for about 24 hours in the air.
However, if you are anxious to see the final result immediately, then
immerse the print in a bath of 0.3% hydrogen peroxide for no more than
half a minute before transferring to the final wash. This solution of
hydrogen peroxide is made up freshly by diluting 50 cc of the commonly
available 6% solution of hydrogen peroxide (the so-called ‘20 volume’
solution) to 1 litre of water. This treatment makes no ultimate difference
to the print densities in the final outcome. Under no circumstances
218
should you use a bath of potassium or ammonium dichromate for this

reoxidation, although it has been recommeded in the past – see §7.5.
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
3 2 1
–log(relative exposure)
Fig. 7.2 Characteristic curve for New Cyanotype with 0.1% dichromate
1% acid developed (right) and water developed (left) 4x exposure
7.4.10 New cyanotype: workflow summary

1. Unbuffered paper: choose side, mark up coating area
2. Prepare sensitizer: add citric acid (to 2%) and/or Tween (to 0.25%)
3. Coat: ca. 1.5 cc per 10x8 in. area: 4-6 ‘passes’ of coating rod
4. Dry: 1-2 hours at room temperature, or hot air for 10 minutes
5. Negative: density range ~2-2.4 in the UVA
6. Expose: UVA source till highlights green and shadows ‘reversed’
7. Develop: 0.1-1% v/v nitric acid for 5-10 minutes, raising contrast
8. Wash: in non-alkaline water for 20-30 minutes
9. Drain for 10 minutes and Dry flat
219
7.4.11 Effects of heat-drying and ‘darktime’

The effect of heat-drying new cyanotype sensitizer coatings before
exposure is found to slightly shorten the tonal scale by about one stop –
from 2.3 to 2.0, but it may also slightly coarsen the texture. Heat-drying
the paper before coating was found to give no benefit at all. It has been
claimed that results from the traditional cyanotype process can be vastly
improved in quality by giving the exposed print a lengthy ‘dark-time’ of
ca. 18 hours before processing.588 I have tested this claim for the new
cyanotype process on Buxton paper with a 2 minute exposure followed by
19 hours of darktime at 58% RH and ~20 °C. The test results using
Stouffer 3110 step tablets are shown in Fig 7.3. Each is in duplicate.
1 Control: no heat-drying or darktime
2 Darktime only: 19 hours at 58% RH
3 Coating heat-dried only: at 42°C for 5 minutes
4 Coating both heat-dried and given darktime
1 2 3 4
Fig 7.3 Effects of heat drying sensitizer and ‘darktime’ exposure
I concluded that leaving the exposed print in the dark before wet
processing lengthened the exposure scale by about one stop, from 2.3 to
220
2.6, but weakened the Dmax perceptibly, flattened the shadow tones
slightly, and caused some edge fog, with a rim at the tideline. It seems
only to be worth resorting to this darktime treatment if essential
highlight detail would otherwise be lost.
Since heat-drying the coated paper shortens the exposure scale by
one stop, then a combination of both heat-drying the coating and giving
darktime gets one back to the original exposure range of ~2.3, but with a
loss of Dmax, so it is best to do nothing!
7.4.12 Diagnosis of fogged highlights
A common fault observed in many handmade prints is highlight fogging –
the appearance of tone due to image substance (Prussian blue) being
deposited where there should be none. There are at least six possible
causes of this fog:
(1) The sensitizer is already decomposed

(2) There is a hostile chemical in the paper, reducing the sensitizer
(3) The safelighting is inadequate under the conditions of working
(4) The mask or negative is not dense enough in its maximum value
(5) The wet chemistry is faulty, chemically reducing the sensitizer
(6) The clearing procedure is inadequate, leaving residual iron, etc.
so there now follows in Table 7.1 a systematic method of diagnosis.

Note that it is important for reference and comparison, to have a margin
of clear paper which has not been coated with sensitizer, and a border to
the image that has been coated but remained masked and unexposed, as
described in §6.4.3.
221
Procedure:
Compare a border region of the sensitized area – coated, but masked
during exposure – with the margin of uncoated paper, and answer the
following questions YES/NO :-
Fog apparent on coating? YES ® Sensitizer is decomposed (1) or

NO ¯ Very hostile chemical in paper (2)
Fog appears during drying? YES ® Fog if dried in total darkness?
YES® Hostile paper chemical (2)
NO ¯ NO® Faulty safelight (3)
Fog appears after pop exposure? YES ® Mask/neg not dense enough (4)
NO ¯
Fog after wet dev processing? YES ® Sensitizer decomposed (1) or
Faulty safelight (3) or
Mask/neg not dense enough (4)
NO ¯ or Wet chemistry faulty (5)
Stain of sensitizer after wash? YES ® Clearing process inadequate (6)
NO ¯
Stain in uncoated areas of paper? YES ® Wet chemistry contaminated (5)
Notes
pop refers to printout processes, dev to development processes.
It is important to distinguish Fog (unwanted residual image substance)
from Stain (unwanted other residual chemicals, especially ferric salts).
These are usually distinguishable by different colours – the former grey,
the latter yellow.
Fault (4) can be detected by including a small area of high UV blocking –
'Rubylith' – for comparison with the maximum density of the negative.
Table 7.1 An algorithm for the six most probable causes of fogging
222
7.5 Cyanotype composites with other processes

It is possible to make superimposed multiple prints which combine the
characteristics of the cyanotype with the same image rendered in other
processes. The easiest option, because of the inertness of the image
substance, is to print and completely finish a platinum-palladium print,
then recoat the dry print with cyanotype sensitizer, and make a second
print in register using the same negative. For a cyanotype over platinum
or palladium the new cyanotype process is more suitable because it has
about the same exposure scale as the platino-palladiotype process and a
single negative will serve well for both printings; whereas the classical
cyanotype process cannot render the density range of a negative made
suitably for printing in platinum-palladium.589 The reverse order of
process, with the cyanotype made first, is not possible with platinotype
because of the effect of oxalate on Prussian blue: traditional platinotype
developer will rapidly bleach and dissolve any Prussian blue image. On
the other hand it should be possible with palladiotype because the
developer – acidified sodium citrate – will not attack Prussian blue
(although it does render it light-sensitive). A cyanotype over palladium
print was exhibited in 2016 by the Worcester Art Museum in their
landmark Cyanotypes show: it was a print by Edward Steichen entitled
Jean Simpson in Profile which is now thought to be of this type.590 XRF
showed that the predominantly brown image is palladium, with iron
detectable in small areas of residual cyanotype blue at the edges; the
surface presents a slight sheen and a fine craquelure that suggests it may
be a Japine paper. It is worth noting here that a degree of Prussian blue
toning can also be imparted to a finished platinotype, without a second
printing, by the catalytic toning procedure described in §9.12. This
addition of blue to the highlights of a platinotype was sometimes referred
to as a “Moonlight effect” toning, and could also be carried out by
including potassium ferricyanide in a dilute oxalate developer with added
glycerol.591
The chief difficulty in multiple printing is maintaining registration
between the successive images in the face of possible dimensional
changes of the paper substrate when it passes through the expansion
and contraction that accompanies two cycles of wetting and drying. For
this purpose a precise pin-register system is desirable, and papers may
be pre-conditioned by one or two cycles of soaking in pure water and
drying.
223
7.6 List of chemicals and hazards

The following information has been condensed from the Materials Safety
Data Sheets for the substances involved in making various cyanotype
sensitizers.592 The quoted Hazard Ratings are on a scale running from 0
to 4, in which:
0 = no hazard, 1 = slight, 2 = moderate, 3 = severe and 4 = extreme.
When handling any chemical, it should be standard practice always to
wear the minimum laboratory protective equipment of safety glasses, a
lab coat or smock, and latex gloves. Additional precautions are indicated
where necessary. During the use of any chemical, avoid splashing into
your eyes, skin, and clothing. Wash your hands thoroughly after handling.
When a chemical is not in use, keep it in a tightly closed, properly
labelled container, in a place inaccessible to children.
Ammonium dichromate
(NH4)2Cr2O7
RMM = 252.07
CAS 7789-09-5
Aldrich #22,481-2; Alfa #13444
Hazard Rating
Health 4 extreme (cancer causing)
Flammability 1 slight (involatile, but solid is combustible)
Reactivity 3 severe (oxidizer)
Contact 3 severe (corrosive)
Harmful if dust is inhaled, contact may cause burns or external ulcers.
Strong oxidizer, contact with other materials may cause fire.
Do not get in eyes, on skin, on clothing. Wash thoroughly after handling.
Unfortunately, this substance has a hazard rating of 4 on the health scale,
because it is a known human carcinogen: external contact can cause
ulcers and dermatitis. In spite of this hazard, ammonium dichromate (or
the corresponding potassium or sodium salts) has long been used on a
large commercial scale in the printing industry, and for tanning of
leather, occasionally giving rise to documented cases of occupational
illness. It is also employed in large amounts for the Gum Bichromate and
Carbon printing processes. The very small quantity used in the
sensitizers described above does not represent a serious risk if the
chemical is handled sensibly and proper precautions taken, but if the
224
hazard is still considered unacceptable, its use may be avoided altogether

as indicated above in the preparative section.
You are strongly advised against the use of a dichromate bath to ‘fix’ (i.e.
re-oxidise) cyanotypes, which was once commonly recommended
practice, especially for prints on fabrics. The unpleasant consequences to
an artist who used a dichromate bath for quilting fabrics are described in
some detail in the medical literature.593 Victims of dichromate exposure
can become particularly sensitized to it.
Ammonium iron(III) citrate

Fe content: 14-18% (green form); 19-28% (brown form)
RMM = ca. 290/mole Fe (green); ca. 391/mole Fe (brown)
CAS 1185-57-5
Aldrich #22,896-6; Alfa #A11199 (brown)
Also known as: iron ammonium citrate; ammonium ferric citrate; ferric
ammonium citrate.
Hazard Rating
Health 1 slight
Flammability 0 none
Reactivity 1 slight
Contact 0 none
This substance presents very low risk; it has even been prescribed as an
iron tonic and is present in small amounts in a well-known soft drink!594
Unprescribed consumption is not recommended, however, because
excess iron in the body can cause medical problems (siderosis).
Ammonium iron(III) oxalate (trihydrate)

(NH4)3Fe(C2O4)3.3H2O
RMM = 428.07
CAS 13268-42-3; 14221-47-7
Aldrich #23007-3; Alfa #42112; Bryant #F335; Spectrum #F1002;
Fluka #12302; Citychemical #F393
Also known as: iron ammonium oxalate; ammonium ferric oxalate;
ferric ammonium oxalate; ammonium trisoxalatoferrate(III) trihydrate.
Hazard Rating
Health 3 severe
Flammability 1 slight
Reactivity 2 moderate
Contact 3 severe
225
Like all soluble oxalates (see oxalic acid, below), it should be regarded as
highly toxic, i.e. capable of causing death or permanent injury (especially
kidney damage) if taken by mouth. However, several grams would have to
be ingested to cause such extreme damage. The chemical takes the form
of non-volatile crystals, so there is little risk of inhalation, but it is an
irritant.
Hydrochloric acid, concentrated (ca. 36% w/w HCl )

HCl
RMM = 36.46
CAS 7647-01-0
Aldrich #25,814-8; Alfa #L13091
Also known as: muriatic acid
Hazard Rating
Health 3 severe
Flammability 0 none
Reactivity 2 moderate
Contact 3 severe
Concentrated acids cause severe burns; may be fatal if swallowed or
inhaled. Do not get in eyes, on skin, on clothing. Keep in tightly closed
container. Loosen closure cautiously. Use with adequate ventilation. Wash
thoroughly after handling. In case of spill, neutralize with soda ash or
lime and place in dry container. Causes damage to respiratory system
(lungs): inhalation of vapours may cause pulmonary edema, circulatory
system collapse, coughing and difficult breathing.
Emergency and first aid procedures: call a physician. If swallowed, do not
induce vomiting; if conscious, give water, milk, or milk of magnesia. If
inhaled, remove to fresh air. If not breathing, give artificial respiration. In
case of contact, immediately flush eyes or skin with plenty of water for at
least 15 minutes while removing contaminated clothing.
to skin, eyes and mucous membranes.
Nitric acid, concentrated (ca. 69% w/v HNO3)

HNO3
RMM = 63.01
CAS 7697-37-2
Aldrich #25,811-3; Alfa #33260
226
Hazard Rating
Health 3
Flammability 0
Reactivity 2
Contact 3
Concentrated acids cause severe burns; may be fatal if swallowed or
inhaled. Do not get in eyes, on skin, on clothing. Keep in tightly closed
container. Loosen closure cautiously. Use with adequate ventilation. Wash
thoroughly after handling. In case of spill, neutralize with soda ash or
lime and place in dry container. Causes damage to respiratory system
(lungs): inhalation of vapours may cause pulmonary edema, circulatory
system collapse, coughing and difficult breathing.
inhaled, remove to fresh air. If not breathing, give artificial respiration. In
case of contact, immediately flush eyes or skin with plenty of water for at
least 15 minutes while removing contaminated clothing.
to skin, eyes and mucous membranes.
Oxalic acid (dihydrate)

(COOH)2.2H2O
RMM = 126.07
CAS 6153-56-6
Aldrich #24,753-7 or #O-875-5; Alfa #33262; Bryant #O226; Fluka
#75702 or #27725; Merck #818242 or #100495
Also known as: ethanedioic acid.
Hazard Rating
Health 2 moderate
Reactivity 1 slight
Contact 3 severe (corrosive)
Harmful if swallowed, inhaled or absorbed through skin. Avoid breathing
dust. Very destructive of mucous membranes. Ingestion may cause severe
burns to mouth, throat, and stomach and may cause headache, nausea,
vomiting, gastrointestinal irritation, unconsciousness, convulsions.
May have adverse effect on kidney function and may be fatal.
May cause congenital malformation in the foetus: possible teratogen.
227
inhaled, remove to fresh air. Rinse mouth or gargle repeatedly with cold
water. In case of contact, immediately flush eyes or skin with plenty of
water for at least 15 minutes.
Potassium ferricyanide
K3Fe(CN)6
RMM = 329.26
CAS 13746-66-2
Aldrich #24,402-3; Alfa #33357
Also known as: potassium hexacyanoferrate(III), red prussiate of
potash.
Hazard Rating
Health 1 slight
Flammability 0 none
Reactivity 1 slight
Contact 1 slight
This substance is classified as a low toxic hazard, but may act as a skin,
eye and lung irritant. It is incompatible with concentrated acids, which
may release deadly hydrogen cyanide gas. Used correctly there is no risk.
Tween 20™ (polyoxyethylenesorbitan monolaurate)

C58H114O26
RMM = 1227.54
CAS 9005-64-5
Aldrich #27,434-8; Alfa #43309; Bryant #T355; Fluka #93773 or
#63158; Merck #822184 or #817072
Also known as: polyoxyethylene (20) sorbitan monolaurate;
Polysorbate 20™.
Hazard Rating
Health 1 slight
Reactivity 0 none
Contact 0 none
This is a viscous, odourless liquid, soapy and innocuous, which presents
no serious risks, provided contact with the eyes is avoided.
228
7.7 Environmental issues and disposal

Concientious practitioners of cyanotype often ask about the appropriate
disposal of their waste treatment solutions, and this can become quite a
crucial question in environmentally sensitive situations for those
operating “in the wild” or with septic tanks, land drainage, etc. In general
the processing solutions are very dilute; and apart from dichromates (see
below), there is nothing in the cyanotype sensitizer or processing baths
when dilute that can be described as hazardous - compared with all the
other chemicals we use and encounter in everyday life. Treating the
effluent as “chemical waste” is quite unnecessary from an environmental
viewpoint, and an expensive and needlessly troublesome consumption of
vital resources.
The components of the Herschel cyanotype formula are pretty benign:
ferric ammonium citrate is not poisonous (it is taken medicinally, and
used (E381) in a celebrated Scottish soft drink ‘Irn-Bru’); its components,
ferric ion and citrate ion, are widely naturally-occurring. Ferricyanide ion
would quickly reduce in the environment to ferrocyanide, which is also
non-toxic; it is a permitted food additive (E536) used as an anti-caking
agent in table salt. Oxalates, though poisonous when concentrated, are
naturally present in rhubarb, spinach, and sorrell at the 1% level.
If the waste seems very acidic (test with pH paper) it can be
neutralised with a cheap alkali like lime, or chalk. Nitrate itself is of very
low toxicity: ammonium nitrate is spread on farmland in massive
quantities as an agricultural fertilizer.
Prussian blue is very insoluble and non-toxic. Humans can eat – and
excrete - grams of it per day with no ill effect. It is a valuable antidote to
poisoning by thallium and caesium. It has been spread on land
contaminated by Cs137, and fed to reindeer, see §3.3.
Despite these common-sense observations, one occasionally
encounters – especially on the unmoderated internet - alarmist claims
such as those in the article "Food for Thought: the Hidden Dangers of
Cyanotype",595 which are exaggerated, confused and inaccurate.
This article is an egregious example of the irrational "anti-chemical"
mindset, arriving at a ridiculous conclusion: "From a prudent perspective
it would seem advisable to avoid Historic photographic processes
altogether." This is justified with the misleading quotation: "Potassium
ferricyanide, although only slightly toxic by itself, will release hydrogen
cyanide gas if heated, if hot acid is added, or if exposed to strong
229
ultraviolet light (e.g., carbon arcs)." It is necessary to qualify this: the

heating has to be very strong and the acid very concentrated to get much
HCN. Neither of these things are done in cyanotype. The 'traditional'
cyanotype process, using only ammonium ferric citrate and potassium
ferricyanide (§7.1), will not emit any significant, perceptible or harmful
amount of hydrogen cyanide during a normal exposure and processing.
From all points of view, there is no safer photographic process, and I
have no misgivings placing it in the hands of 5-year-olds.
However, cyanotypes can also be made by exposing a paper coated
with potassium ferricyanide alone (see the "proto-cyanotype" of §4.1.1);
the exposure to UV to get an image is much longer than the 'traditional'
process - about 20-50x - i.e. hours instead of minutes. In this case a
perceptible odour of hydrogen cyanide (and/or possibly cyanogen - also
toxic) can be detected when the print frame is opened. This may be the
origin of the alarmist views in the document. But the quantity of HCN so
formed is not life-threatening and there is no particular reason to make
cyanotypes this way.
The only chemical in this work that is recognised as hazardous is
potassium or ammonium dichromate (§7.6), which is a listed human
carcinogen. But it does not need to be used in traditional cyanotype at all.
In the past it was occasionally employed as a processing bath – at least
once with harmful consequences – and this practice is now discontinued.
Dichromate is recommended in very small amount in my New Cyanotype
as a preservative - but can be entirely omitted (§7.4.2). Regarding
disposal of dichromates, there are some methods reported in the
literature for eliminating the toxicity of dichromates first by reducing the
Cr(VI) to Cr(III) with reducing agents such as sodium sulfite,596 sodium
thiosulfate,597 iron filings,598 or sugar,599 but all these have the
disadvantage of requiring rather acidic conditions; a preferable reductant
for hexavalent chromium(VI) is ascorbic acid (vitamin C),600 which is
harmless and inexpensive and works at neutral or even slightly alkaline
pH.601
7.8 Suppliers of chemicals and equipment

Ready-made sensitizers, chemicals and equipment for the cyanotype
processes, and in some cases complete kits, can be obtained from the
following suppliers:
230
Fotospeed
Jay House Ltd
Fiveways House
Westwells Road
Rudloe
Corsham
Wilts
SN13 9RG
UK
Tel: +44 (0)1225 810 596
Fax: +44 (0)1225 811 801
Silverprint Ltd
12 Valentine Place
London
SE1 8QH
UK
Tel: +44 (0)171 620 0844
Fax: +44 (0)171 620 0129
WWW: http://www.silverprint..co.uk/
Phototec
Peter Loffler GmbH
Gewerbehof 1
26209 Hatten
Germany
Tel: +49 04481 1534
Fax: +49 04481 1895
Artcraft Chemicals Inc

PO Box 583
Schenectady
NY 12301
USA
Tel: +1 800 682 1730
WWW: http://www.nfinity.com/~mdmuir/artcraftB.html
231
Bostick & Sullivan Inc

PO Box 16639
Santa Fe
NM 87506
USA
Tel: +1 505 474 0890
Fax: +1 505474 2857
WWW: http://www.bostick-sullivan.com/
Photographers Formulary Inc

PO Box 950
Condon
MT 59826
USA
Tel: +1 800 922 5255
Fax: +1 406 754 2896
Fine-art papers may be obtained in the UK from:
John Purcell Paper

15 Rumsey Road
London
SW9 0TR
UK
Tel: +44 (0)171 737 5199
Fax: +44 (0)171 737 6765
232
8 Methods for Toning Cyanotypes

"To gild refinéd gold, to paint the lily,
To throw a perfume on the violet,
To smooth the ice, or add another hue
Unto the rainbow, or with taper light
To seek the beauteous eye of heaven to garnish,
Is wasteful and ridiculous excess."
William Shakespeare
King John, Act 4, Scene 2
Many procedures can easily be devised to change the composition of a

Prussian blue image in order to modify its colour, but most of these
prove to be ‘more curious than valuable’ and have little reproducibility
and even less permanence. Handbooks on cyanotype often quote recipes
for a variety of toning methods,602 but their descriptions are usually
tempered by the author’s reservations as to the worth of such a
procedure, and sometimes even by a plea not to ‘torture’ the humble
cyanotype in this way! The process historian, W. Russell Young III says:
“My best advice for toning cyanotypes is this: If you want a color
other than the Prussian blue natural to the cyanotype process, use
some other printing method!” 603
However, it is becoming increasingly evident that many present-day

practitioners, who are attracted by the ease and low cost of cyanotype,
are still dissatisfied with its natural colour, and feel driven to seek toning
procedures that convert their images to brown or black, rather than
adopting the naturally black processes such as platinotype, palladiotype,
or even kallitype.604
There is little evidence for how long such toned cyanotypes can be
expected to last: the identification of historical toned cyanotypes in
photograph collections presents considerable curatorial difficulty, if
visual inspection is the only non-destructive method that is permissable.
In a survey of several major collections conducted by photograph
conservator Megan Gent in 1992, no curator that she interviewed could
report having definitely identified a toned cyanotype. However, Gent has
discovered some prints in informal Royal Family snapshot albums in the
Royal Archives at Windsor dating from 1906 that are probably toned
cyanotypes: these are single layer prints of a purplish-brown colour
233
indicative of toning with tannic acid, and they share the same album
pages with conventional blue cyanotypes.605
The toning processes which are acknowledged to be the most reliable,
and least impermanent, appear to fall into five categories, most of which
have a commonality in commencing with the hydrolysis of the Prussian
blue to iron oxide/hydroxide before further treatment.
8.1 Hydrolysis of Prussian blue

The print is hydrolysed with dilute (1% to 5%) alkali, which may be
aqueous ammonia, sodium carbonate (washing soda), sodium or
potassium hydroxide (caustic soda or potash), calcium hydroxide (lime
water) or sodium tetraborate (borax). The Prussian blue lattice is broken
up by hydroxide ion, and the ferric iron is converted into insoluble
yellow-brown ferric hydroxide, while the soluble and stable ferrocyanide
ion is eventually washed away in the solution. Prompt treatment with acid
can partially restore this breakdown, but it soon becomes irreversible.
Trisodium phosphate (sometimes called TSP) is also very alkaline and
will bring about hydrolysis of Prussian blue, but the product is believed to
be the golden yellow pigment, ferric phosphate, which is air-stable, and
insoluble in water but can be dissolved by acids. It is fairly permanent in
its own right, and it is said to be responsible for the famously enduring
and non-rusting “Iron Pillar of Delhi”, the surface of which is coated with
an invisible but impervious layer of the phosphate. This method of toning
is quite popular for making yellow images on fabrics.606
Exposure just to ammonia vapour, rather than solution, does not fully
hydrolyse the Prussian blue but imparts to it an impermanent, though
striking, violet colour; this toning is reversible within a few hours simply
by restoring the print to air, when the occluded ammonia gas is
presumably lost, and the pigment reverts to its original blue colour.607
8.2 Ferro-gallate and tannate

The image is converted to ferric hydroxide as above, and this is reacted
with gallic or tannic acid to form ferric gallate or ferric tannate ‘ink’.608
As was described in §4.4.2, John Mercer can be credited as the originator
of this toning process. The print is hydrolysed as in §8.1 and then treated
in a bath of gallic or tannic acid (1% to 5%), with some precautions to
minimise staining:
1 If the print is dry, presoak it in water for 1 minute.
234
2 Immerse the print in 1% v/v ammonia (ammonium hydroxide) for ca. 5

minutes until the Prussian blue image is bleached to pale yellow (ferric
hydroxide). Caustic soda or caustic potash (sodium or potassium
hydroxide) 1% w/v, or domestic washing soda (sodium carbonate) ca.
5% w/v, can be used instead of ammonia.
Curtailing the time in this bath can yield interesting split-tone effects.
3 Rinse in water for half a minute.
4 Immerse in 1% v/v acetic acid for one min. to neutralise residual alkali.
5 Rinse in water for half a minute.
6 Tone in 1% tannic acid solution for 5 to 10 minutes.
7 Wash in water for 20 minutes.
This treatment generally intensifies the image, and imparts a rich
purplish-brown colour. Staining of the paper base is a problem, however,
and the colour is very sensitive to alkali. The best purplish-brown colours
are said to result when the print is immersed first in gallic or tannic acid
for a few minutes, rinsed, then treated with alkali, and finally, as an
option, it may be returned briefly to the gallic acid bath.609
8.3 Assorted beverages

In recent years it has become fashionable to tone cyanotypes with
“natural products” containing tannins and similar catecholic substances
that react chromogenically with iron salts. Teas, black and green, black
coffee of several varieties, and red wine have all been used both with, and
without preliminary bleaching and hydrolysis, to yield rather muddy
brown and black images.610 A common by-product of this kind of toning
is a stained paper-base which some apparently find attractive. The
working texts should be consulted for further details and examples.611 It
has not yet been established what the life-expectancy of such images
may be, but the author, Mrhar, admits that some of the procedures are
“unreliable and capricious”, and other workers, such as Lukasz
Brzezinski,612 have noted significant colour shifts in their toned prints
within a year or two.
8.4 Iron-mordanted dyes

The Prussian blue is again converted to ferric hydroxide as in §8.1, which
then acts as a mordant for organic dyestuffs. This procedure was also
invented by Mercer in his quest for generating ‘chromatic photographs’,
235
and is particularly well-suited to toning images on textiles. The Prussian

blue image is usually hydrolysed with lime (calcium hydroxide) or other
dilute alkali as above, then washed and treated in a suitable dye bath,
when the resulting ferric hydroxide serves as a mordant for vegetable
dyes such as gall-liquor, logwood, quercitron bark, murexide, cochineal,
or madder. When using madder dye, Mercer found that if the Prussian
blue was left intact and unhydrolysed, a very rich purple colour was
produced.613 Out of a whole range of chemically inventive processes,614
perhaps Mercer’s most remarkably circumambulant achievement was to
make a cyanotype as described above, then to convert the image into
silver ferrocyanide by treatment with silver nitrate, after which it was
reduced to silver metal by sodium stannite, and finally substantially
converted by a chloroaurate bath into gold in the form of Purple of
Cassius. Examples of portraits made by this process on cloth, may be
seen in fine condition, in the Mercer Archive at Lancashire County Record
Office, see §4.4.2.615 The Museum of Science and Industry in Manchester
also has a photographic portrait of Mercer, printed from a negative taken
ca. 1856 by John Benjamin Dancer, the famous instrument maker and
photomicrographer; it is printed on cotton, probably by Mercer himself,
and has a deep violet colour, which suggests it may be a madder-toned
cyanotype.616
Forty years later, in 1898, Mercer’s toning process was re-invented -
apparently unwittingly - by Stewart Carter, and reported to the New York
Section of the Society of Chemical Industry.617 Cyanotypes, made in the
usual way were treated first with dilute caustic soda, then, after washing,
with sodium phosphate (a non-toxic alternative to Mercer’s arsenate), at
which point it was ready to receive the hot 70-80 °C aqueous 3-5% dye
bath, which could be selected from resorcin green, gallocyanine, alizarin,
or logwood.
8.5 Black toners

The conversion of a cyanotype into a black silver image would seem to be
a singularly pointless exercise when the same end can be achieved
directly; however, it has been done.618
“To get the desired black color, bathe the print in distilled water
and then immerse it in a 2% solution of silver nitrate. After the
bleaching of the image, re-wash the print in distilled water and dry
it. Following drying, expose the print to the vapors of ammonia for
a short time and develop it in a solution of oxalate of iron (ferrous
sulfate developer as used for wet plate collodion).” 619
236
The soaring expense of silver photography during the 20th century

motivated the quest for other, cheaper ways to obtain a black non-silver
image by toning the blueprint. An early process, due to Planchon,
converted the image into copper(II) sulphide,620 but there are better
methods, as will be described.
The reaction products of bleaching the Prussian blue in a sulphite
bath, can be used to catalyse the oxidative reaction of hydrogen peroxide
with certain polyphenols, such as catechol, resorcinol and quinol, to form
dark-coloured, near-black dyestuffs. This method was developed in the
1970s by Reginald Heron,621 with the well-intentioned aim of
transforming the strident blue image of the inexpensive cyanotype into a
more photographically-acceptable black, thus providing a very economic
non-silver medium of practice, especially for students who did not have
access to a photographic darkroom. Heron’s developing bath is a
complex mixture, with a composition finely-balanced according to the
characteristics of the original blueprint material that was commercially
available at the time. The process is said to produce a “fair imitation of a
platinum print”. Thanks to its chromogenic nature, the colour can be
modified, but the resulting dyes, like all those of chromogenic
photography, are not totally stable. According to Heron:
“Samples have survived the 25 years to the present day, but the
original (reasonably attractive) black has faded to a passable (and
inevitable) brown.” 622
8.6 Heavy metal incorporation

Treatment of a Prussian blue image with a solution of a heavy metal salt
may cause the incorporation of the metal cation into the Prussian blue
lattice, and thus modify its colour. Not all metal cations cause a
perceptible effect: it is conjectured that the metal may need to have two
accessible oxidation states to intervene in the electronic charge transfer
transition giving rise to the colour. Thus lead (+2,+4), thallium (+1,+3),
nickel (+2,+3) and copper (+1,+2) all exert a perceptible effect on the
cyanotype colour, as described next, whereas alkali metals (+1) and
alkaline earth (+2) cations, of a similar size and charge, do not.
8.6.1 Lead
In a procedure attributed ca. 1900 to Oscar Bolle, a cyanotype is treated
in a hot (30-40 C) bath of lead(II) acetate solution (S.G. 1.24), to shift the
colour towards a more violet blue, sometimes rather inaccurately
described as “lilac” or “lavender”.623 The effectiveness of this toning bath
237
has recently been found to be critically dependent on the pH of the lead

acetate solution.624 This is often distinctly acidic (pH ~5) when freshly
made up from solid to a working strength of ~5% w/v, but for successful
toning, it is particularly important that the lead acetate solution should
not be acidic; adjustment of its pH to ca. 7.5 to 8 by dropwise addition of
strong ammonia solution is recommended, using a pH paper to check the
alkalinity. Approximately 20 cc of ammonia (S.G. ~0.9) may be needed
per litre of 5% lead(II) acetate toning solution. A slight precipitate (of lead
hydroxide) doesn't matter, as it can always be filtered off. With this
marginally alkalized solution, the toning can be carried out at room
temperature and will be complete within 1-2 minutes.
The presence of lead in this toned image substance has been detected
by X-Ray Spectrometry (EM-EDX),625 and the lead signal is seen to
correlate with the iron signal from the Prussian blue, suggesting that the
toning effect may be due to the incorporation of lead(II) ions into the
voids of the microporous Prussian blue structure (see §3.3 and App II.2).
The result may also depend on the cyanotype formula; it works well
with the new cyanotype recipes (§7.4). The colour change is permanent,
and it stabilises the Prussian blue somewhat against light fading, by an
exposure factor of about 4x compared with untreated images, but the
vulnerability of the pigment to alkali is found to be undiminished.626 The
lead toning can apparently be reversed by treatment in a bath of dilute
nitric acid, which removes the lead.
It should be born in mind that lead(II) acetate is a heavy metal salt
which is seriously toxic by ingestion and it should be handled carefully
and stored safely; it is said to taste sweet (its commonplace name is
sugar of lead), but experimental confirmation of this claim is definitely
not recommended for this cumulative poison. It is also said – disputedly -
to have been one of the causes of the downfall of the Roman Empire! 627
8.6.2 Thallium
In the same way that thallium(I) is incorporated when Prussian blue is
used as an antidote to this insidious poison (see §3.3), it can also be
used as a toner of cyanotype. A bath in thallium(I) sulphate solution of ca.
5% w/v does itself cause a slight but agreeable colour shift in a cyanotype
to a ‘cornflower blue’, but this cannot be recommended as a toning
practice for the general public owing to the severe toxic hazard presented
by this heavy metal (see §3.3).
238
Fig. 8.1 Varieties of toned cyanotype

8.6.3 Nickel
This method follows from the experiments reported by Holtzmann,
described in §9.2.628 Immerse the cyanotype in a 5-10% solution of a
nickel(II) salt (e.g. the sulphate, chloride or nitrate) for an hour or so. The
colour shift with nickel(II) salts is rather slight - towards a more
greenish-blue. This treatment has the beneficial effect of making the
239
cyanotype much more resistant to alkaline hydrolysis, by at least an order

of magnitude (10x) in exposure, but the application of the nickel solution
may cause a slight loss of highlight density. Nickel(II) salts are listed
carcinogens. Consult the MSDS.
Fig. 8.2 Examples of cyanotypes toned by heavy metals.

Clockwise from the top left: untoned, lead(II), nickel(II) and thallium(I)
8.6.4 Copper
This method of converting the image colour to a purplish-brown derives
from a U.S. patent of 1889.629 The stock reagent is a deep blue solution
of tetraamminocopper(II) nitrate, made by dissolving 1 part of copper
nitrate in 4 parts of water and treating the solution with one part by
volume of 30% (S.G. 0.90) aqueous ammonia solution, sufficient just to
re-dissolve the blue precipitate of copper hydroxide. The toning bath is
made by diluting this blue stock solution sixteen times with water and
adding up to one further part of 30% ammonia, depending on the colour
desired, which is described as a “Bartolozzi red”, and is believed to
contain some copper(II) ferrocyanide.
240
8.7 Insoluble metal ferricyanides

This rather curious – and apparently little-explored - toning procedure
was described by Humphrey Desmond Murray of London in 1933 UK and
1935 US patents;630 it imparts a two-tone appearance to the image by
colorizing the high values. The idea is that, immediately after exposure of
a conventional cyanotype sensitizer and the printing-out of the primary
Prussian blue image, the first wet processing bath should contain a salt
of a metal that forms an insoluble ferricyanide – generally lightly-
coloured – with the unreacted ferricyanide ion in the sensitizer. Thus the
high values of the print receive a different colour from the blue shadows,
depending on the chosen metal cation, and further action of light is
prevented, i.e. the print is also ‘fixed’. Murray described the use of salts
of the following metals at 10-20% concentration, with the results
indicated for the highlight colour: Manganese(II) (salmon pink), cobalt(II)
(brick red), nickel(II) (bright yellow), zinc(II) (pale yellow), and cadmium(II)
(pale yellow). He also advised that the first bath should contain an
oxidising agent such as a persulphate or a chlorate, coupled with the use
of a mildly acidic buffer, to ensure full initial development of the Prussian
blue.
8.8 Blue toning of silver prints

Finally, this is an appropriate point to mention those toning procedures
for silver-gelatin materials which transform the silver image, wholly or
partially, into Prussian blue, thereby generating a print with the
cyanotype colour, but differing in that the image is not in the paper
fibres, but is embedded in a gelatin binder layer which confers a more
lustrous and reflective appearance to the surface. The chemical
transformation of the silver image into ferric ferrocyanide is
accomplished by first oxidising the silver to silver ferrocyanide using a
bath of alkaline potassium ferricyanide, then treating it in an acidic bath
of ferrous sulphate, followed by fixation in thiosulphate.631 It is also
possible to use a single direct toning bath of acidified ferricyanide and
ferricitrate combined, which results in some degree of intensification, but
the image is said to be less permanent.632 The same effect can be
achieved simply by adding gelatin to a standard cyanotype sensitizer,633
or - wastefully - by coating with cyanotype sensitizer a silver-gelatin
photographic paper which has had all its silver halides dissolved and
removed.
241
9 Vulnerability & Conservation of Cyanotypes

“But of course these tints must be used with judgement,
and no one but a vandal would print a landscape in red,
or in cyanotype.”
Peter Henry Emerson (1856-1936)
Naturalistic Photography for Students of the Art (1889)
The chemical properties of Prussian blue described in Chapter 3, and at

more depth in Appendices II and III, identify three distinct directions in
which cyanotypes are vulnerable to attack. The weaknesses that need to
be addressed by conservation science are: photochemical reduction,
alkaline hydrolysis, and aqueous peptization. These pathways of
destruction may more conveniently be referred to as ‘fading’, ‘bleaching’,
and ‘dispersing’, respectively. Each of these damaging reactions leads to
a loss of Prussian blue from the image, but each is chemically distinct in
its causes, products, and remedies. In the first, the cause is light, the
product is Prussian white, and the reaction is completely reversible. In the
second, the cause is alkali, the product is yellow hydrated ferric oxide
(colloidal, becoming insoluble) and ferrocyanide ions (which are soluble),
and it is only partially reversible. In the third, the Prussian blue is not
changed chemically but dispersed as nanoparticles which are washed
away by water - a loss which is quite irreversible. These pathways of
vulnerability are summarised in Table 9.1 and will now be considered
individually in more detail.
Fading Bleaching Dispersing

Description Photochemical Alkaline Aqueous
reduction hydrolysis peptization
Cause Visible and UV Any substance Water and high
light and a of alkaline pH ionic strength
reductant (pH >7) solutions
Product Prussian white Hydrated ferric Colloidal sol of
(ferrous oxide and Prussian blue in
ferrocyanide) ferrocyanide water
Reversibility Fully reversible As ferric oxide Pigment is
by air in the dark ages it becomes washed out from
over a few days irreversible image irreversibly
Table 9.1 Pathways of destruction for cyanotypes
242
9.1 Fading of cyanotypes by light

Notwithstanding the extravagant claims made for its stability by the
discoverer of Prussian blue, its discolouration to a greenish hue under the
action of sunlight was noted in a handbook of painting as early as 1757,
quite soon after its adoption as a paint pigment. The historical evidence
in the mid-eighteenth century for the awareness of this deterioration has
been compiled into an interesting review by Jo Kirby of the Scientific
Department of the National Gallery.634 It is likely that impurities in the
early preparations of Prussian blue may have been responsible for the
variability observed in its behaviour. Even the reversal, in the dark, of the
decolorization was noted in 1834. The light-fading of pigments is
naturally a central issue for conservation scientists concerned with
paintings, and the fading of Prussian blue in oil paintings, especially
when in admixture with large quantities of an extender such as lead
white,635 has received close attention at the National Gallery.636 A
conspicuous recent example is the demonstration of fading of Prussian
blue in an 1838 oil painting by Danish painter Christen Købke – a
decolorization that quite alters the character of his romantic
landscape.637
From the mercantile viewpoint, when used as a textile dyestuff the
pigment also revealed its susceptibility to light at an early stage. In 1801,
Desmortiers described his researches into the fading of fabrics dyed with
Prussian blue in his Recherches sur la décoloration spontanée du bleu de
Prusse,638 in which he emphasised the importance of allowing access to
the air in order to reverse the fading and restore the blue colour.
Subsequently, around 1840, Michel Chevreul found that Prussian blue in
an air-tight chamber became paler and liberated cyanogen in the direct
rays of the sun, but in the dark it absorbed oxygen and recovered its
colour.639 These phenomena must have been widely known at the time
because they were also reported in the standard English chemistry
texts,640 one of which observed that ‘the recovery of the colour is owing
to the formation of the so-called basic Prussian blue’.641 The chemistry of
this reaction is considered in Appendix III.
Cyanotypes and commercial blueprints also acquired a reputation
early on for fading partially in the light; and, although some contrary
opinions were expressed, it is now generally accepted that they are
susceptible,642 but the degree of fading has not been fully quantified.643 A
number of photograph conservators have cited evidence for the fading of
243
cyanotypes. In 1989, Ian and Angela Moor described densitometric

measurements on cyanotypes exposed to daylight, employing “facsimile
material produced using Herschel’s original formulation”. They found a
“loss of density ranging from 4% after 15 minutes exposure to 10% after
45 minutes exposure”, but they do not report data on the absolute
density changes. They attributed the fading to “the presence of residual
soluble ferric salts, not removed during the washing procedure.”644 They
found that the regain of the density on dark storage was not fully
reversible, only 7% of the original 10% loss being recovered.
An extensive programme of densitometric monitoring of a variety of
original historic photographs has been carried out by John McElhone,
photograph conservator at the National Gallery of Canada, Ottawa.
Specimens were displayed under stringent museum conditions, and
periodically densitometered in order to determine how ‘safe’ these
conditions are for the exhibition of the various types of image; this work
and its curatorial implications have been described elsewhere.645
McElhone’s recent unpublished studies have included original cyanotypes
by Atkins (1854) and Curtis (1868). Highlight, mid-tone and shadow
densities were measured with a reflectance densitometer for the three
primary colours at intervals during time periods extending from two to
four years of display in glazed mounts.646 Illumination was by 50-60 lux
incandescent tungsten, filtered to remove the UV. Within the instrumental
precision and reproducibility of the measurements, no significant effects
(i.e. with permanent density changes ∆D ≥ ±0.02) were observed during
total accumulated exposures that amounted to about 100 kilolux hours.
This important finding should offer some reassurance to curators and
conservators that historic cyanotypes can be safely displayed under
proper conditions. However, these harmless tests, having caused no
change in the objects, do not establish the threshold at which damage
does begin to appear, and that is the purpose of the present work with
facsimile material.
9.1.1 Experimental: preparation of specimens
In the present work, five different cyanotype formulations were tested, to
represent the main categories of sensitizer used for the negative-working
process (see Chapter 4). For brevity and ease of reference these formulae
will be allocated the names of Smee, for Herschel’s ‘proto-cyanotype’
using potassium ferricyanide alone (§4.1.1); Herschel, for his own
modified recipe (§4.1.2); Lietze, for the usual 19th century recipe using
244
brown ammonium ferric citrate (§2.12); Valenta, for the usual 19th
century recipe using green ammonium ferric citrate; and Ware, for the
new cyanotype using ammonium ferric oxalate (see Table 9.2).
Process Description Ammonium Potassium

Name ferric citrate ferricyanide
Smee Herschel’s ‘proto-cyanotype’ none 16%
Herschel Herschel’s ‘standard’ recipe 6% 12%

(brown)
Lietze Typical C19th recipe 14% 12%
(brown)
Valenta Typical C20th recipe 14% 12%
(green)
Ware New cyanotype 16% 10%
as described in Chapter 7 ammonium
ferrioxalate
Table 9.2 Varieties of cyanotype – final concentrations in sensitizer
The paper generally used for coating in the comparative tests was
Atlantis Silversafe Photostore, of weight 120 grams/square meter. This is
a pure cotton conservation paper, of high purity, sized with Aquapel™
which is a neutral-curing alkyl ketene dimer. It contains no added buffer,
filler, or other agent that might interfere with the photochemisty.647 Such
a pure paper differs rather from the papers of the 19th century, so
comparative tests were also performed on gelatin-sized papers, both
contemporary (Fabriano 5 HP) and of the type used 150 years ago
(Whatman’s Turkey Mill 1840 and J. Whatman 1849, made by
Hollingworth and Balston, respectively).648 Coating was carried out by a
glass rod, and the volumes of sensitizer used were closely controlled to
give comparable coatings. The sensitized paper specimens were all dried
at room temperature in the dark.
The samples to be tested were made by conventional cyanotype
printing and direct sunlight, using a glazed printing-out frame with a
hinged-back, mimicking as accurately as possible the manner of
exposure that would have been employed in the 19th century. Only the
Ware formulation was printed by means of an artificial ultraviolet light
source, which is now universally used, employing four Phillips
TLADK30/05 coated fluorescent tubes (having a maximum emission at
245
360 nm) at a distance of 8 cm. The test prints were made through a
calibrated Kodak No.3 step tablet, which produced a maximum of 21
steps of density in half stop intervals of exposure, which were large
enough to be accurately densitometered. Exposures were usually
sufficient to make the first two steps indistinguishable, so ensuring that
the maximum density of which the sensitizer was capable had been
reached. A direct comparison was made for the Herschel formula to see
if the sunprinting employed in the nineteenth century produced any
qualitative differences from the artificial UV source. The heating effect of
the sun exposures was notable, and it was found that prolonged
exposures to the sun resulted in a purple veiling of the highlights which,
however, usually disappeared in the wet processing. Processing was
appropriate to the sensitizer in use, usually by means only of gently
running tapwater, and the specimens were dried in the dark.
Smaller control specimens were printed through Stouffer Graphic Arts
T2115 step tablets649 at the same time as the main K3 test piece, and
were retained in a light-proof box, that had access to the atmosphere of
the same environment through a set of baffles. These could be checked
for any long term changes not due to illumination and were also useful
for sacrificial experiments.
9.1.2 Experimental: fading apparatus and measurements
The light source used to induce fading was a commercial luminaire
fitting, equipped with four 20 watt standard artificial daylight fluorescent
tubes, of length 60 cm, General Electric type F20W/AD. This formed the
lid of an enclosed box fitted with two fans to ensure a gentle constant
stream of air to cool the lamps and maintain the samples in a constant
environment. The temperature during exposure was maintained at 19 ±
1°C and the relative humidity at 50 ± 5 %. The fluorescent tubes were run
initially for a period of time before testing to ensure constant emissivity.
The illuminance at the plane of the samples was measured with a Weston
Master V exposure meter fitted with an ‘Invercone’ attachment in its
incident light mode, and checked with a spotmeter V and standard grey
card. Both readings corresponded to an illuminance of approximately 4
kilolux.
The optical densities of the specimens were measured to a precision
of 0.002 by means of an X-Rite model 310 densitometer in diffuse
reflectance mode, which was checked by reference to a standard density
plaque, and recalibrated when necessary to absolute density readings
246
with an accuracy of ± 0.01 over the range 0.06 to 1.74, before and after
each set of experiments. All densities recorded refer to the ‘red’ channel
of the reflectance colour head, which corresponds to the waveband where
the absorption by Prussian blue has its maximum value.
9.1.3 Results: characteristic curves and the ‘fade’
The characteristic (or D/logH) curves of the typical negative-working
cyanotype sensitizers are shown in Figures 9.1a-9.1f, in which the diffuse
reflectance density is plotted vertically against the logarithm of the
relative exposure horizontally. In each case the characteristic curve was
remeasured after an exposure to artificial daylight of ca 2 kilolux hours,
and this is also plotted on the graph. It can be seen that all types of
sensitizer have suffered a loss of density in consequence of the light
exposure, amounting to about 0.1 to 0.2 density units.
If Di represents the initial optical reflectance density of any step of the
pigment layer in a step-tablet test, and Df is the measured density of the
same step immediately following a fading exposure, the extent of fading
is denoted by the change DD = Di – Df. The measured values of DD have
probable confidence limits of ±0.01 units of density, and they are
generally less than 1, so it will be convenient, both typographically and
with regard to numerical precision, to discuss the fading henceforth in
terms of a whole-number parameter, which I shall call the ‘fade’, which is
defined as 100DD. We will find it convenient to have a single parameter to
express the measured susceptibility of a particular Prussian blue layer to
fading. The maximum value of the fade for a given exposure is the
parameter we seek, since it represents the greatest loss of image density
within the entire tonal scale, and therefore the greatest loss of picture
information.
When Di = 1, the value of the fade is also the value of the % relative
fade, 100DD/Di , but it is not appropriate to make general use of the
relative fade as the parameter to discuss the results for two reasons:
psychometric experiments show that the JND is approximately constant
over a wide range of densities; and the consequence of division by the
very small density values in the highlights is to put a numerical bias on
the results because it gives undue prominence to small changes.
247
Smee Herschel

Reflectance Density (red) 1.6 1.6
1.4 1.4
1.2 1.2
1.0 1.0
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
3 2 1 3 2 1
–log(relative exposure) –log(relative exposure)
Herschel on gelatin-sized paper Lietze
1.6 1.6
1.4 1.4
1.2 1.2
1.0 1.0
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
3 2 1 3 2 1
Valenta Ware
1.6 1.6
1.4 1.4
1.2 1.2
1.0 1.0
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
3 2 1 3 2 1
Fig. 9.1 Characteristic curves for cyanotype sensitizers
248
9.1.4 Results: variation of fade with initial density

To see how the fade varies across the tonal scale, typical results for the
step-tablet tests printed by a variety of cyanotype sensitizers are plotted
in Figs 9.2a-9.2f where the vertical axis is the fade, plotted against the
initial reflectance density, Di, of the steps which runs horizontally. The
fade is seen to reach a broad maximum value (which varies between
sensitizers) in the mid-tones of the density scale between Di values of ca
0.5 to 1.0, and to fall off towards both ends of the scale. This trend is
broadly true for all the cyanotype sensitizers examined. The fact that the
maximum fade occurs over quite a wide plateau of mid-tones is a
significant, and perhaps surprising, finding. It shows that the fade is not
proportional to the concentration of Prussian blue - which might have
been expected if the fading were an intrinsic property of the substance.
Values of the maximum fade, and the densities at which they occur for
some of the specimens investigated are summarised in Table 9.3.
Sensitizer type Maximum fade Over initial

100∆D density range
Smee 13 0.90-0.95
Herschel on aquapel-sized paper 11 0.55-1.05
Herschel on gelatin-sized paper 12 0.80-1.20
Lietze 10 0.65-0.85
Valenta 10 0.35-0.85
Ware 22 0.85-1.15
Table 9.3 The maximum fade for cyanotype sensitizers
The near-constancy of the maximum fade for a given specimen over

quite a wide density range makes further investigation of the other
variables easier, because only the density of one or two typical mid-tones
need be followed and not the whole scale of 21 steps. This finding also
emphasises the importance of always carrying out cyanotype fading
studies on mid-tones, where the loss of picture information will be
greatest, and not just on a region of maximum density, where there may
be a large ‘reserve’ of Prussian blue in which little change will be
observed.
249
Smee Herschel
Fade (100∆D)
Fade (100∆D)
15 15
14 14
13 13
12 12
11 11
10 10
9 9
8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5
Initial Density Initial Density
Herschel on gelatin-sized paper Lietze
Fade (100∆D)
Fade (100∆D)
15 15
14 14
13 13
12 12
11 11
10 10
9 9
8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5
Valenta Ware
Fade (100∆D)
Fade (100∆D)
23
15
14 21
13 19
12
17
11
10 15
9 13
8
11
7
6 9
5 7
4
5
3
2 3
1
1
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5
Fig. 9.2 Variation of fade with initial density
250
The fact that the maximum fade of a cyanotype tends to occur in the
middle tones may also account for some of the disparities in the
anecdotal evidence that has been reported concerning the fading of
cyanotypes and blueprints. If the original image has a high contrast, with
little or no intermediate tone, then the visible fading effect will be
confined to the extremes of the tonal scale where it is minimal. This is
particularly the case with ‘line’ images, such as contact prints from
engravings, or photograms of opaque objects, where the image has only
two densities in the limiting case: the density of the paper base plus fog,
and the maximum density of which the sensitizer is capable. As can be
seen from Table 9.3 the extent of fading varies little with sensitizer
composition, except the sensitizer prepared from ammonium ferric
oxalate, Ware type, which shows a deeper initial fade. As will become
apparent, this is not in fact a disadvantage.
9.1.5 Results: variation of fade with exposure time
To ascertain how the fade varies with fading exposure time, test strips
were exposed in the artificial daylight source and monitored at regular
time intervals; only 4 of the 21 steps were selected for density
measurement so that the time taken to perform this did not significantly
detract from the exposure times. When the densities did not undergo any
further decrease with exposure, the samples were removed to dark
storage at ambient temperature and humidity (in the range 18 ± 2 ºC and
50 ± 10 %RH), with full access to the atmosphere, and their densities
remeasured periodically.
The results are shown in Figure 9.3a-9.3f for one step only (that
giving rise to the maximum fade) of the five sensitizers, in which the
maximum fade is plotted vertically against time along the horizontal axis.
It may be seen that the fade quickly reaches a maximum value (i.e. a
minimum in the measured density) after only about a half to one hour’s
exposure to 4 kilolux. Following this it tends to decrease slightly but
steadily with continuing exposure up to 24 hours, when the exposure
was terminated. In one case the decrease in the fade (i.e. the regain of
density during the fading exposure) was quite marked, as illustrated in
Figure 9.3f. The knowledge that the maximum fade was reached within
one hour’s exposure at this illuminance greatly simplified subsequent
monitoring of repeated fading tests. The regain of density in the dark was
also monitored in these samples and found to be effectively complete
(final densities ≥99% of the initial densities) within five days.
251
Herschel Lietze
Fade (100∆D)
Fade (100∆D) 23
15
14 21
13 19
12
17
11
10 15
9 13
8
11
7
6 9
5 7
4
5
3
2 3
1
1
10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80
Time/minutes Time/minutes
Valenta Ware - water developed
Fade (100∆D)
Fade (100∆D)
23
15
14 21
13 19
12
17
11
10 15
9 13
8
11
7
6 9
5 7
4
5
3
2 3
1
1
10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80
Ware - citric acid developed Ware - HCl developed
39
Fade (100∆D)
Fade (100∆D)
31
37
29
35
27 33
25 31
23 29
27
21
25
19 23
17 21
15 19
17
13
15
11 13
9 11
7 9
7
5
5
3
3
1 1
10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80
Fig. 9.3 Variation of fade with exposure time (at 4 kilolux)
252
9.1.6 Results: reversibility of repeated fade and regain

In view of the kind of exposure that cyanotypes receive when on
exhibition, it was thought important to determine the reversibility of the
fading, when subjected to repeated exposures. When the step densities
had regained an almost steady value in the dark, the process of
measurement and exposure was therefore repeated.
Figures 9.4a-9.4f show schematically the fluctuating density of one
step (taken from those within the region of maximum fade), over several
cycles of fade and regain. The fading exposure was 4 kilolux for half an
hour, the regain period varied from one to several days in the dark. The
most striking feature is the diminishing oscillation of the depth of fade,
while the regain of density is usually greater than 95% within one day,
and 99% within five days or more, i.e. it is complete within the limits of
experimental error.
It is apparent from these results that the tendency of a cyanotype to
fade may be dependent on its previous history. They are most susceptible
when freshly-made, and, for some processes, each exposure to daylight
levels of illumination, followed by regain, diminishes this susceptibility.
The permanent loss of density incurred overall as a result of repeating
the process is generally very small, representing a fade at most in the
order of 3 or 4 (i.e. ≥96% regain).
It is clear from the experimental results that the fading of cyanotypes
does not obey the photochemical Law of Reciprocity,650 but varies with
the absolute exposure illuminance. At low light levels (ca 50 lux), we have
the evidence of McElhone’s monitoring program on cyanotypes to show
that even illumination for 2000 hours, giving an exposure totalling 100
kilolux hours, causes no detectable fading. The probable reason for this
behaviour is that the rate of density regain, due to air re-oxidation of the
Prussian white, is superimposed on the effect of fading, and when the
illuminance is so low that the rate of fading is less than or equal to the
rate of regain, no fading at all will be observed, provided sufficient air
can diffuse in to the vicinity of the print. This is the situation of
cyanotypes exposed to stringent gallery illumination of 50 lux. When the
illuminance is much higher, such as daylight at ca 4000 lux in the present
tests, an identical exposure of 100 klux hours is reached in 25 hours.
This incurs a measurable maximum fade of between 10 and 20,
depending on the sensitizer.
253
Herschel Lietze
Fade (100∆D)
Fade (100∆D)
22
14
20
12 18
16
10
14
8 12
10
6
8
4 6
4
2
2
1 2 3 4 5 1 2 3 4 5
Number of fade cycle Number of fade cycle
Valenta Ware - water developed
Fade (100∆D)
Fade (100∆D)
22
10 20
18
8 16
14
6 12
10
4 8
2 4
1 2 3 4 5 1 2 3 4 5
Ware - citric acid developed Ware - HCl developed
Fade (100∆D)
Fade (100∆D)
38
30
36
28 34
26 32
24 30
28
22
26
20
24
18 22
16 20
14 18
16
12
14
10 12
8 10
6 8
6
4
4
2 2
1 2 3 4 5 1 2 3 4 5
Fig. 9.4 Reversibility of fade and regain in cyanotypes
254
Conservators Jennifer Koerner and Karen Potje at the CCA have tested
a variety of 22 blueprints to a 36 klux light source, and found that all
samples reacted similarly, showing a perceptible fading within 5 minutes
exposure – a light dose of 3 kilolux hours.651
To demonstrate the important influence of the air on the outcome of
fading, J. Dunbar performed an experiment in which cyanotype
specimens were enclosed in an anoxic environment (a Marvelseal™ packet
containing Ageless™ oxygen scavenger) and given a 24 hour fading
exposure, which caused them to turn totally white. When opened and
exposed to the air, they regained their full blue colour.
To determine the effect of very high illuminance on the fade, samples
(of the Herschel and Smee sensitizer types) were given the same fading
exposure as above, but to direct sunlight, i.e. ca 100 kilolux for one or
two hours, which was found rapidly to cause a deep fade, as much as 40
or so, before a steady state was reached. Thus the same exposure of 100
kilolux hours, but delivered at three widely differing levels of illuminance,
has three quite different outcomes in the degree of fading.
9.1.7 Results: effect of conditions on fading
To see if the fading of Prussian blue in cyanotypes was wavelength-
sensitive, gel filters were placed around fresh specimens of Ware new
cyanotype (which are the most sensitive) in the light exposure box, while
still allowing access to the atmosphere; identical test strips were exposed
side-by-side under red and blue filters, and compared with a strip
exposed without filter.
Property Control Red filter Blue filter

Transmittance of red light % 100 51 12
Transmittance of blue light % 100 10 84
Maximum fade = 100∆D 16 12 12
Initial density = Di 0.85 0.87 0.83
Time to maximum fade in mins 20 45 45
Table 9.4 Effect of wavelength of light in fading cyanotypes
The results are summarised in Table 13, where it may be seen that
there is no significant difference between the fading experienced by the
test strips exposed to red and blue light, indicating little or no
wavelength dependence of the effect over the visible spectrum. Since
Prussian blue has absorption bands in both the red and blue regions of
255
the spectrum, presumably either is effective in promoting fading. The

rates of fading are somewhat slower than the control which was exposed
to unfiltered light, as might be expected from the attenuation of the
illumination by the filters. It follows that there is no useful form of
filtered ‘safelighting’ that can completely protect cyanotypes against
fading.
To determine the effect of relative humidity on fading and regain,
identical specimens of Ware new cyanotype were desiccated (both by
pre-heating to 110 °C for 3 hours, and at room temperature) in an
enclosure dried by silica gel to a relative humidity (RH) of ≤ 5%, and then
given a fading exposure of 2 kilolux-hours within their desiccator vessel,
which had a thin glass lid, versus a control which was also exposed under
similar glass but with full access to the ambient atmosphere (RH ≈ 55%).
It was found that the initial desiccation of the specimens had no
perceptible effect on their densities. The values of the maximum fade for
the desiccated specimens were 16 and 12, respectively for the heated and
unheated specimens, compared with a value of the maximum fade of 15
for the control specimen. The specimens were allowed to regain density
in the dark, respectively, in the desiccator vessel and the ambient air.
After 4 weeks the fade had returned to values of 5, 3, and 2, respectively.
On removal of the desiccated specimens to ambient atmosphere and
allowing them to regain for a further 5 days, the fade values were 2, 1
and 2, i.e. the samples had returned to their original densities within
experimental error, with a regain of ≥98%. It may be concluded that
complete desiccation of cyanotypes has no significant effect on their
tendency to fade, but it may somewhat inhibit their rate of regain. Since
Prussian blue has a great avidity for water, attracting and retaining a
proportion of water molecules within its structure, the effect of relative
humidity on fading of cyanotypes under all normal circumstances is
inconsequential.
The thoroughness of washing a newly-made cyanotype has a
considerable effect on its susceptibility to fading. To demonstrate this, a
Ware formulation new cyanotype, acid-developed, but which had only
received a perfunctory wash of 4 minutes in running water before drying,
was observed to suffer a maximum fade as high as 42 under a 2 kilolux-
hour exposure. In contrast, a similar test-piece which had been washed
in gently running tapwater for 20 minutes experienced a ‘standard’
maximum fade of only 16 under the same exposure. This provides
further evidence that fading is promoted, at least partly, by impurities left
256
in the print as a result of imperfect processing, as originally suspected by

Ian and Angela Moor. However such impurities cannot usually be fully
washed out retrospectively, especially in old specimens, and prolonged
washing in water can cause considerable loss of image by peptization of
the Prussian blue, so a compromise in this regard must be reached.
To discover if there is any mitigation of fading if oxidants are
incorporated in the print, specimens of Ware cyanotypes were treated,
after development, with oxidants to restore fully the densities of the
Prussian blue. Solutions of ferric chloride and ammonium dichromate
(strength ca 3% w/v) were used, the latter being a common treatment for
commercial blueprints at one time. Step tests were then densitometered
and test-faded in the usual way. The results showed that these
specimens were not immune from fading, indeed they showed a
maximum fade of 17, comparable with the fade of 16 in an identical test
that had not been treated with re-oxidants but simply allowed to regain
density in the air. The only difference was that the maximum fade was
reached more slowly - in about twice the time taken by the untreated
specimen. This result suggests that there is no significant benefit in
using the dichromate bath that has been recommended in the past. The
experiment also shows that inorganic ferric salts, i.e. ferric ions per se,
neither promote nor significantly inhibit the fading.
Of the many toning procedures suggested for cyanotypes, most de-
stabilise the image, and cannot be recommended. In the lead toning
method of Oscar Bolle (§8.6.1), the processed cyanotype is immersed in a
5% solution of lead(II) acetate, preferably at a pH between 7 and 8, which
shifts the blue towards a beautiful deep violet. Specimens treated in this
way were found to exhibit improved resistance to light fading, by a factor
of about four. Analysis by X-ray spectrometry strongly indicated that the
lead(II) ions are incorporated in the Prussian blue lattice. However, the
strong colour shift and the toxicity of the lead acetate bath debar this as
a conservation teatment, but artists originating cyanotypes might well
consider incorporating this procedure into their practices if they find the
colour acceptable.
9.1.8 Conclusions: mechanism of light-fading
The chemical reaction of the fading process (Appendix III.10) is
essentially a light-induced reduction of Prussian blue to Prussian white,
which necessarily entails the acquisition of an electron from some other
substance, and an extra cation to balance this increased negative charge.
257
For every reduction, there must be a corresponding oxidation.652 The

question naturally follows: what is there in a cyanotype print that can be
oxidised? If this component can be identified, then it may be possible to
prevent or inhibit fading. There are several candidates, which fall into
three categories:
1. A component of Prussian blue itself.
2. Other constituents of the cyanotype print, including the substrate.
3. An impurity incorporated within the Prussian blue, arising from the
circumstances of its preparation.
Let us examine each of these categories in turn:
1. The inherent stability of Prussian blue to light
The Colour Index describes the light-fastness of Prussian blue as
‘excellent’,653 and the manufacturers of artists’ pigments list Prussian
blue as ‘durable’. Kirby has reported a fading test on Winsor and
Newton’s powdered Prussian blue pigment, which was painted out as a
moderately intense, simple wash with gum Arabic. The specimen was
exposed continuously for 180 days to a standard artificial daylight source
of 10 kilolux, which amounts to a total exposure of over 40 megalux-
hours. Little or no loss of density or colour change was observed in the
specimen.654
We also have the evidence of scientific investigations of the behaviour
of pure Prussian blue in highly controlled environments, where it is also
found to be very stable. The pure substance (as its ‘insoluble’ variety) in
aqueous suspension shows no tendency to decompose under prolonged
irradiation with visible light, establishing that Prussian blue does not
significantly photo-oxidise water to oxygen, or its own bound cyanide ion
(to form cyanate or cyanogen) under these conditions.655 This is further
discussed in Appendix III. Moreover, if Prussian blue were inherently
light-sensitive we might have expected proportional behaviour for the
fade, rather than the sub-proportional behaviour, which is more
indicative of the presence of an impurity.
Set against these observations is the early nineteenth century report
by Chevreul which suggests that very heavy exposures to UV-rich light,
e.g. direct sunlight, will result in some decomposition of Prussian blue,
but these are not conditions that will prevail in any gallery or archive, and
they have not therefore been extensively investigated in the present
work, which has adopted exposures to illumination that are likely to be
encountered in exhibition and everyday handling.
258
2. Possible effects of the substrate and binder

In a cyanotype print, as opposed to a pure Prussian blue sol, the system
is complicated by the presence of substances from the paper
environment: primarily the cellulose substrate, any non-cellulosic
constituents such as lignin, the sizing agent, and any other
manufacturers’ additives to the paper pulp. The surface contact between
the Prussian blue particles and these substances may be rather restricted,
but they cannot be excluded from consideration as potential reducing
agents.
A number of experiments were carried out in the course of this
research to determine if the substrate itself could promote photochemical
fading. Cellulose is an oxidisable material,656 so its possible contribution
to fading was checked by preparing cyanotype test prints on a variety of
other, non-oxidisable absorbent substrates, such as glass plates coated
with silica or alumina (as used for thin-layer chromatography), fibreglass
filter paper, and bisque-fired ceramic tile bodies. The general
observation in all these tests was that cyanotypes formed on such non-
reducing substrates faded in light as rapidly as those prepared similarly
on pure cellulose paper. Thus, cellulose itself is not implicated as a major
contributor to the fading process in cyanotype images.
The lignins that are present in inferior grades of wood-pulp paper are
readily oxidisable substances and could promote the photoreduction of
Prussian blue, but no tests have been made in this respect because there
is every reason not to use such papers for making cyanotypes, and they
are easily avoided. However, some commercial blueprint papers of the
past may have contained appreciable amounts of lignin, which could
contribute to their deterioration. The presence of gelatin size does not
appear to increase the fading, as judged by the foregoing results.
3. Impurities destabilising Prussian blue
The most likely cause of light-sensitivity in cyanotypes is the
incorporation within the image substance of impurities which derive from
the sensitizer; the availability of readily oxidisable substances will
facilitate the rapid photoreduction to Prussian white. This may not simply
be a case of residual impurities left in the paper fibres by imperfect
clearing of the chemicals during the wet-processing procedure; but there
may also be a trapping of the impurity within the Prussian blue lattice,
and once trapped no amount of washing may remove it. The
susceptibility to fading is therefore likely to be highly dependent on the
circumstances under which the Prussian blue is formed, i.e. the
259
formulation and concentration of the sensitizer and the processing of the

print.
The variability and history-dependent character reported for the light-
fading of cyanotypes therefore becomes comprehensible, but there have
been few proposals as to which substances or formulations are more
likely to promote fading. It has been suggested that residual ferric salts,
presumably the hydrated oxides, in the print can be a cause, but these
are inherently oxidising rather than reducing agents and the
experimental evidence cited above shows no such effect.
Fading of Prussian blue requires the simultaneous presence of several
contributing factors: light; a reducing agent (electron donor) in intimate
contact with the solid; and cations to migrate into the lattice and balance
the increased negative charge brought about by reduction. Molecules that
are small enough to be occluded within the voids of the Prussian blue
lattice should be particularly effective in promoting the photoreduction.
To test this possibility, pure Prussian blue coatings on paper were
prepared by successive coats of potassium ferricyanide and ferrous
sulphate, then part-coated with solutions of oxidisable anions such as
citrate and oxalate, and allowed to dry. The density of the impregnated
Prussian blue paper was compared with the control region after a period
of dark storage to ensure that there was no reduction in the dark, then
the whole sheet was given a standard fading exposure. The reductant-
coated area was found to fade in the light much more rapidly than the
control area, showing that these oxidisable organic anions do indeed
promote photoreduction of the pigment. It may be significant that the
anions effective in promoting fading in Prussian blue are also those that
normally render iron(III) photoactive in its salts, and are consequently
present in the cyanotype sensitizer.
To improve the light fastness of cyanotypes, therefore, the presence of
excesses of these free anions, such as citrate or oxalate, should be
avoided. Since an acid development bath is desirable in the processing of
new cyanotypes (Ware formula employing ferrioxalate sensitizer), it is
better to use a very dilute mineral acid such as hydrochloric or nitric,
rather than citric or oxalic acids, which have been previously
recommended.
If, in the light-fading of the Prussian blue, the products of oxidation
of the impurity are lost from the cyanotype (for example as carbon
dioxide gas, in the case of oxalate), the impurity should be progressively
and irreversibly destroyed as the dose of light accumulates, but the
260
Prussian white will still be fully re-oxidised by air. The quantity of

impurity should therefore diminish with successive cycles of fading and
regain, and eventually the susceptibility to light fading should disappear
when all the impurity is ‘burnt out’. This is the behaviour observed during
the experiments on cyclic fade and regain described above, although the
fading never dropped to zero, so there must be some irremovable cause
also. Ferrocyanide ion is cited as a common impurity in Prussian blue; the
oxidation-reduction equilibrium between these two substances is finely
balanced; the oxidation product of ferrocyanide is ferricyanide, which will
not be lost from the system, so although fading and regain may be
reversible in this case, the susceptibility to fading cannot be ‘burned out’.
9.1.9 Regain of colour by faded Prussian blue
It has long been appreciated that processing solarised or faded
cyanotypes in a dilute dichromate bath re-oxidises them rapidly, but it
has also been claimed that it improves their resistance to fading;657
presumably because this powerful oxidant will destroy most reducing
impurities. There is also the (as yet unproven) possibility that some
dichromate or chromate may be incorporated into the Prussian blue
lattice, where its oxidising presence might continue greatly to inhibit
photoreduction. However the strong yellow colour of dichromate would
tend to make the hue greenish if it were so incorporated. The effect of a
dichromate re-oxidation bath on the fading susceptibility of new
cyanotype (Ware formula) was tested as described above and found to be
ineffectual.
Regain of the Prussian blue in a faded layer requires an oxidant
(electron acceptor), and probably some water to assist cation migration
out of the lattice, and it will be further assisted by acid which enhances
the action of most oxidising agents, such as oxygen, dichromate and
hydrogen peroxide, and neutralizes the hydroxyl ions produced by them,
so avoiding hydrolysis. The amount of gaseous oxygen needed to
accomplish the re-oxidation of an average cyanotype is surprisingly
small: a calculation for a typical coating of maximum optical density
given below shows that, for an area of A4 size, the oxygen contained in
about 0.7 cc of air should suffice to completely re-oxidise the Prussian
white to Prussian blue:
An area of 1 m2 of a Prussian blue layer of optical density 1.55
contains ca. 5x10–4 moles; this corresponds to 11.2 cc of O2 gas at STP.
(Although molecular oxygen only constitutes one quarter of the
261
atmosphere, one mole of dioxygen can oxidise four moles of Prussian

white, so the factors approximately cancel, see Appendices II.9 and III.8):
4Fe[Fe(CN)6]2– + O2 + 2H2O ® 4Fe[Fe(CN)6]– + 4OH–
An A4 sheet has an area of 1/16 m2, hence the volume of air needed
to restore a completely faded A4 print is 11/16 = 0.7 cc.
9.1.10 Recent spectrometry of Prussian blue fading
Fifteen years after the work described above was concluded, further new
scientific investigations were made of the fading of Prussian blue by
Claire Gervaise and co-workers, using sophisticated modern X-ray
spectrometric techniques (XANES), employing the Soleil proton
synchrotron.658
These new investigations have been carried out on samples of
Prussian blue prepared by the direct reaction of an iron(III) salt with
ferrocyanide; this form does not suffer from the same vulnerabilities
typical of the kind of Prussian blue occurring in cyanotypes, which is
sensitized by impurities such as citrate or oxalate. In general the directly
prepared pure substance is very much more stable to light, as explained
in §9.1.8 and Appendix III.10, so this new work is operating in quite a
different energy regime. To obtain recordable significant fading, Gervaise
and colleagues used light-fading exposures to an HMI lamp of irradiance
1200 W/m2 (ca. 0.82 Mlx @ 555 nm) of 5 days duration (120 hours), i.e.
their exposures were in the order of 100 Mlx hr. These exposures are
therefore about four to five orders of magnitude greater than those
employed earlier in this work with cyanotypes (2 klx hr as in §9.1.2). The
new work has also highlighted the importance of the cellulosic substrate
in providing oxidisable material which facilitates the reductive fading by
supplying electrons and K+ cations to maintain charge neutrality.
Interestingly, cotton and linen based samples were found to differ in their
response, and so does silk. Lignin-based papers are very reactive and
suffer deterioration of the Prussian blue even before light exposure.
These workers also found that Prussian blue is faded by X-ray irradiation.
9.2 Bleaching of cyanotypes by alkali

Despite the false claim by the discoverer, Diesbach, that his blue was
stable to lime, painters soon realised that this, and indeed any other
alkali, rapidly destroyed Prussian blue and the pigment could not
therefore be used for painting frescoes.659 Alkaline sensitivity is the most
262
serious drawback with this pigment, and there have been many
endeavours to diminish it.
An extensive investigation of the effect of alkali on Prussian blues has
been conducted by Holtzman,660 who has also sought treatments to
protect the pigment from this type of deterioration. He found that a
buffer at pH 9.4 completely decolourises (i.e. destroys by irreversible
hydrolysis) Prussian blue in 1 to 10 minutes, depending on the precise
method of preparation.661 This is not a highly alkaline pH, and is the
same, for example, as that of a saturated solution of calcium carbonate,
the buffer commonly incorporated in archival papers and boards. A
stronger alkali is a solution of sodium carbonate of strength 0.25 molar,
which has a pH of ca 10.7; this is high enough to destroy the Prussian
blue of a cyanotype in less than half a minute.
Holtzman especially recommended the incorporation of nickel ions
into the Prussian blue as a means of improving its resistance to alkaline
environments.This is best done at the time the blue is prepared, by
forming it in the presence of nickel salts, but it can also be incorporated
retrospectively in high proportion by soaking the Prussian blue in a bath
of a nickel(II) salt. Holtzman found that Prussian blue so treated was
‘stable indefinitely’ at pH 9.4 (or, at least, for 4 months - the duration of
the experiment), and could even withstand the pH 10.7 solution of
sodium carbonate for 4 to 5 hours.
I tested the nickel treatment on Herschel and Ware cyanotypes in case
it could offer a possible conservation procedure for improving the
stability of cyanotypes towards alkaline hydrolysis. The process of
soaking a cyanotype test strip in a solution of 10 % w/v nickel(II) sulphate
for one hour caused a perceptible shift in the hue towards a more
greenish-blue: from 21D8 to 23D8 in the Herschel specimen, and from
20E8 to 21E7 in the Ware specimen (see also §8.6.3 on toning).
Unfortunately the treatment is also accompanied by a distinct loss of
density from the mid-tones, presumably due to peptisation of the
pigment: using the same units (100∆D) as for the fade, the loss was 10
from the Herschel specimen after one hour, and 15 after a 25 hour soak
at room temperature; the corresponding losses from the Ware specimen
were somewhat greater: 16 and 21, respectively. These specimens were
densitometered then tested for their resistance to alkali by immersing
them in a buffer solution at pH 9.4 for 10 minutes, in comparison with
control specimens of the major processes that had not been treated.662
The test strips were briefly (5 minutes) rinsed in tap water and then dried
263
and re-measured. The losses (100∆D) are shown in Table 14 for shadow-
tone regions, from which it may be seen that the nickel(II) treatment is
quite efficacious in protecting the Prussian blue from alkaline
degradation at this pH by about an order of magnitude.
Sensitizer type Di Loss (100∆D) Loss %

Smee 0.953 58 61
Herschel 1.355 68 50
Herschel (Ni(II) treated 1 hr) 1.352 6 5
Herschel (Ni(II) treated 24 hr) 1.122 5 5
Lietze 0.952 65 68
Valenta 1.322 94 71
Ware 1.086 65 60
Ware (Ni(II) treated 1 hr) 0.929 12 13
Ware (Ni(II) treated 24 hr) 1.335 4 3
Table 9.5 Effects of alkaline hydrolysis at pH 9.4 for 10 minutes
There remains the question as to whether a cyanotype image that has

been bleached by alkali can be restored (see Appendix II.8). It is likely
that the ‘degree of irreversibility' of an alkaline bleached cyanotype will
be dependent on the history of the object – for instance whether it has
been washed after the bleaching action, which would remove the soluble
ferrocyanide. When 'freshly bleached' there is some evidence that
cyanotypes can be partially restored by dilute acid, such as hydrochloric
or nitric, and added ferrocyanide, but over time the ferric
oxide/hydroxide hydrolysis product becomes insoluble in dilute acid, and
cannot regenerate Prussian blue. This is an area calling for further
research.
Hydrolysed Prussian blue leaves behind a yellow-brown stain of
iron(III) oxide-hydroxide, which may bind quite strongly to cellulose
substrates - paper or fabric. If it is desired to remove this, the treatment
of reductive chelation recommended for removing iron stains from
Platinotypes may be the best option: an effective treatment is immersion
for ca. one hour at room temperature in an aqueous solution 5% w/v in
each of sodium dithionite (Na2S2O4) and tetrasodium EDTA, at pH ~9. If
the sodiium dithionite is unavailable, sodium sulphite (Na2SO3) may be
substituted, with some loss of effectiveness.
264
9.3 Peptization of Prussian blue

The tendency of Prussian blue to peptize causes one of the major
problems of the traditional cyanotype process, namely, a considerable
loss of image substance, which washes out of the paper during wet
processing, thus necessitating heavy over-exposure (as much as two or
three stops, depending on the paper substrate and its sizing) to
compensate for the loss.663 Clearly it is desirable to inhibit the formation
of ‘soluble’ Prussian Blue in favour of an ‘insoluble’ variety. Formulations
for some commercial blueprint papers contain added iron(III) chloride,
which may have this effect. However, if the potassium ions are eliminated
from the sensitizer, then a less easily peptized form results in the first
place. The use of ammonium hexacyanoferrate(III) in the sensitizer goes
some way to solving this problem, with the result that the new cyanotype
process described in Chapter 7 loses very little image substance on wet
processing.
Similar considerations apply to conservation treatments of cyanotypes
using aqueous washing. This procedure is unlikely to diminish much the
susceptibility to light fading (which is substantially reversible, anyway) if
the cyanotypes have been properly processed in the first place, but it is
likely to cause some irreversible loss of image due to peptization.
Quantitative experimental results on this problem are scarce, but in
1989 Ian and Angela Moor reported some densitometered washing
experiments on cyanotypes. The washing time was 15 minutes in all
cases; in tapwater (pH 7.5-8.5) the ‘average loss in image density’ was
18%; in distilled water (pH 6-6.5) the loss was 4%; in deionised water (pH
6.3-6.6) the loss was said to be 0.00%.664
In 1991 Sarah Wagner reported some research at the Library of
Congress on washing cyanotypes of both recent and historical origins:
specimens were bathed in neutral pH tap-water or in de-ionized water at
pH 6 for a period of 1.5 hours; some loss of pigment occurred, and was
especially marked with the cyanotypes of recent manufacture for which:
“dramatic lightening occurred by immersion in tap water... and slight
lightening for samples which were immersed in deionised water”.665
The effects of aqueous washing on cyanotypes of different varieties
was tested in the present work. Sets of four nearly identical test strips of
a given process were immersed in a gentle flow of tap water at 20 ± 1 °C,
the pH of which was monitored periodically and found to be 7.2 ± 0.1.
The strips were removed successively from the washing process after
265
time intervals of 5 minutes, 20 minutes, 1 hour, and 4 hours, dried, and

their densities remeasured. A similar test was performed in a 2 litre static
bath of ‘purified’ water at 20 °C, having a pH of 6.5 ± 0.1.666 The results
are summarised in Table 15, where the ‘Loss’ figures are 100∆D.
Sensitizer Water Initial Loss in Loss in Loss in Loss in

Type quality density 5 mins 20 mins 1 hour 4 hours
Smee Tap 0.90 3.1 6.6 13.3 28.1
Smee Pure 0.93 0.3 4.6 6.2
Herschel Tap 0.95 0.4 2.3 6.4 13.9
Herschel Pure 0.96 0.1 0.4 2.2 3.2
Lietze Tap 0.98 0.1 2.9 7.0 17.5
Lietze Pure 0.98 0.2 0.5 1.7 1.9
Valenta Tap 0.86 1.2 3.9 9.6 30.7
Valenta Pure 0.85 0.7 1.0 2.2
Ware Tap 0.77 1.3 4.8 13.7 29.0
Ware Pure 0.95 0.1 0.8 2.1 2.2
Table 9.6 Effects of aqueous washing. ‘Loss’ is 100∆D
A phenomenon called the ‘edge etch effect’ was conspicuous on some

peptisation step tests and requires explanation. This was the loss of
image substance from the edges of a step, where it is adjacent to the
surrounding regions of maximum density. That the effect is attributable
to the adjacent area of high density was demonstrated by printing the
step tablet up against the boundary of the sensitized area, in which case
there was no loss. The ‘edge etch effect’ is invisible immediately after the
print-out exposure; it only appears during the wet-processing.
Nevertheless, it can be shown not to be due to migration of ions through
the solution during the wet processing; because the effect still occurs
even when the entire adjacent high-density region is cut off after
exposure but before the wet processing. Evidently the loss is due to ions
which have migrated through the paper from the adjacent area during the
exposure. These promote peptization of the Prussian blue as soon as
aqueous treatment begins. If such a test strip is allowed to rest in the
dark at normal humidity for a few hours after exposure and before the
wet processing, the effect disappears, presumably because the
continuing diffusion of the causative ions has by then led to their uniform
distribution throughout the specimen, and there is no longer a local
266
effect. It is significant that some mobile by-product of the cyanotype

photochemistry can promote peptization of the image substance; the
effect is likely to cause an enhanced acutance in images. It was observed
to be much worse in samples prepared from the brown variety of
ammonium ferric citrate compared with the green variety. It was greatly
reduced or absent in papers sized with gelatin - which may act by
inhibiting the mobility of the ions causing the peptization. In general,
solutions of high ionic strength tend to promote peptization of Prussian
blue. Thus, treatment in a solution of sodium hexametaphosphate of
strength 5-10% can be used to remove the blue in the highlights of a
fogged cyanotype.
As explained in Appendix II.8, a 20% solution of potassium oxalate
cleanly decomposes and dissolves Prussian blue, and can be used for
stain-removal, title-writing on cyanotypes, or, at a lower concentration,
for ‘reducing’ the blue density, provided the soluble products are
removed afterwards by a brief wash in water.
9.4 Conservation issues with cyanotypes

9.4.1 Recognition of cyanotype
It must first be established, by a magnification of 10 to 30 times if
necessary, that the object under consideration is a true photograph
having a continuous tone image, and not one of the several
photomechanical printing processes, of which all except Woodburytype
show a tell-tale micro-texture: a dot screen for letterpress halftone, a
grain pattern for photogravure, or reticulation in collotype.667
Cyanotypes may seem the easiest of all photographic processes to
identify by virtue of their colour, but care should be taken to distinguish
them from other types of photograph that can also be made in a similar
blue: namely, carbon prints, blue-toned silver-gelatine prints, and
Woodburytypes, plus the more recent possibilities offered by the ink-jet
printer – some digital cameras even provide a ‘cyanotype’ setting, and it
is easy to generate a cyanotype look for a print file in Photoshop - but
electromechanical prints also have a patterned microstructure.
An essential distinguishing feature is that all the traditional processes
include a binder layer of gelatin holding the pigment, which lies on the
surface of the paper, and confers some lustre or reflectivity; whereas the
cyanotype is a totally matte, single-layer print in which the pigment is
contained within the surface fibres of the paper sheet, which should be
clearly visible under magnification. However, prints made by the
267
(relatively rare) positive-working Pellet process will also have a deposit of

insolubilised gum retaining the pigment, so will show a binder layer in
relief in the shadow regions; in consequence, Pellet prints will rather
resemble blue Woodburytypes or carbon prints and may be more difficult
to distinguish from them.
9.4.2 Exhibition and light sensitivity
The concept of the threshold exposure was introduced in my previous
book on early silver photographs with the aim of providing a useful
parameter for curators and conservators to describe and discuss, in a
semi-quantitative way, the vulnerability of cultural objects to light.668 The
threshold exposure for a monochrome image is that exposure to light
which results in a just noticeable density change in any significant area of
it. The notion of ‘significant area’ is somewhat subjective, but could be
taken to be from about 1 mm2 upwards. The threshold exposure
therefore defines the light exposure at which ‘perceptible damage’ begins
to accumulate. It is measured in the conventional units of exposure - lux
seconds (lx s) or, more conveniently, kilolux hours (klx hr).
The psychometrics of human vision find that an optical density
difference between adjacent areas in the order of 0.01 is a ‘just
noticeable difference’ (JND) under good viewing conditions.669 Differences
less than one JND are generally imperceptible to the unaided human eye
for areas of smooth mid-tone grey, placed side-by-side under bright
lighting, but they can be measured with sufficiently sensitive
instrumentation (densitometers reading to 0.001). The use of optical
density as the measurable unit of change presupposes a monochrome
object – which is usually the case with C19th photographs. However, if a
colour shift in the object is a significant possibility, the criterion for a JND
is more properly expressed according to the measurements of colour
science, and the change of coordinates in a colour space, such as the CIE
L*a*b* system, as measured with a colour meter (chromameter). In this
system, a colour change of ∆E = 1 or 2 is approximately one JND. If there
is no colour shift, ∆E = ∆L*, where the lightness scale L* in the CIELAB
system runs from 0 (black) to 100 (white). A practical JND of ∆E = 1.5 is
widely quoted for coloured objects.
The experimental results for cyanotypes (see §9.1.3 above) have
shown that an exposure of 2 kilolux hours, i.e. half an hour at a ‘daylight’
illuminance of 4 kilolux, brings about a fade in density of ca 0.1, i.e.
about 10 JNDs. This implies that the threshold exposure for cyanotypes is
268
only 200 lux hours, which may seem alarmingly small, being comparable
with the threshold exposure values calculated, and observed, for
photogenic drawings, which are deemed too light-sensitive for exhibition
under any conditions. However, the behaviour of photogenic drawings is
quite different from that of cyanotypes because the fogging of a
photogenic drawing is irreversible, cumulative and permanent, but the
fading of a cyanotype is reversible by air re-oxidation. As has already
been explained in §9.1.6, the failure of the reciprocity law in the case of
cyanotypes means that the duration of exposure at an illuminance of 50
lux needed to fade them by even one JND may be very long indeed, if not
indefinite, and it is quite safe to exhibit them at, or near, this low level of
illumination. Even after a cyanotype has been faded by moderate daylight
exposure, the evidence is that recovery of image density will usually be
effectively complete.
There is one potentially deleterious factor that should be borne in
mind, however, when subjecting cyanotypes to repeated cycles of fading
by light exposure and regain by aerial oxidation, namely: the build-up of
alkalinity in the print. The chemistry of the re-oxidation of Prussian white
by air necessarily entails the generation of hydroxide ion (see Appendix
II.9), which has the capacity to permanently destroy Prussian blue by
hydrolysis (see §9.2). It is therefore desirable that cyanotypes should be
maintained in a mildly acidic condition, pH 4-5, to counter this hydrolytic
tendency, and exhibition protocols should minimise repeated showings.
9.4.3 Safe copying techniques
The safe copying and exposure of photographs, and other objects
potentially sensitive to light, is assisted by the knowledge of the
following two parameters:
(a) the total light exposure inflicted on the object by the copying process,
(b) the object’s sensitivity to light, measured by its threshold exposure.
A comparison of these two figures (expressed in the same units) provides
a risk-assessment of the ‘potential damage’ to the object, i.e. the
notional ‘cost’ of making one copy, and thus can inform curatorial
decision-making in a semi-quantitative manner. These factors will now
be quantified as far as current knowledge allows.
(a) The total dose of light to the object
This is the exposure sustained by the object during the entire process
of acquiring a copy image, or digital image file, by a specified procedure
– whether by analogue or digital photography with appropriate external
269
lighting or photo-flash illumination, or by a scanner, or a photocopier,

with their built-in light sources. Light exposure is the product of light
intensity (illuminance) and duration; to quantify this the preferred
Système Internationale (SI) units are lux seconds (lx s). It is convenient to
build in a factor of one thousand, so kilolux seconds (klx s) will be used
here as the appropriate units for camera exposures, but lux hours (lx hr)
may be more convenient for measuring light exposure arising from
exhibition or study: 1 lx hr = 3.6 klx s.
A number of workers have previously assessed the total exposure for
several copying set-ups and systems, see Table 9.7 below for values and
references. However, it is essential that individual photographers should
measure the total light dose for their own copying procedures.
Typical Light Level (klx) Exposure time Light Dose

Equipment (klx s)
Scanning CCD 2 15 minutes 1800
camera
(Blackwell)670
Flat bed scanner various various 3.6 - 54
(Vitale)671
Photocopier various various 7.2 - 36
(Neevel)672
35 mm SLR 0.6 (2x100W bulbs ~3 seconds 1.8
camera @ 60 cm distance) minimum*
with photo-flash Set f/11 @100 ISO 0.3
magnification 1/8
120 roll film 1.5 ~3 seconds 4.5
camera minimum*
magnification 1/4
5”x4” studio 6 ~3 seconds 18
camera minimum*
magnification 1/2
Exhibition 0.05 10 hours/day 1800/day
Class 1
Table 9.7 Typical light doses for copying compared with exhibition.
270
*Note: 3 seconds is about the minimum total time for the lights to be
switched on and become thermally stable, and for the camera exposure,
which is generally in the order of one second, to be made.
Some of the figures in Table 9.7 remain highly machine-dependent,
and are likely to change with advancing technology, so these guidelines
for scanners and photocopiers, which were typical of the 1990s, may
require updating and re-calculating. For comparability these calculations
all assume a common, nominal camera exposure setting. If the film speed
is not 100 ISO (ASA) or the lens aperture number A differs from f/11, the
figure should be adjusted accordingly. With flash, the light exposure of
the object is just sufficient for the correct exposure of the camera film, so
it depends only on the camera parameters, and not on the model of
flashgun, which is assumed to be set correctly by the photographer.
It will be noted from Table 9.7 that the light dose to the object can be
minimised by employing photo-flash illumination. The flash exposure for
a given distance, as expressed in the Guide Number of the photo-flash
unit, determines the camera parameters that must be set by the
photographer: an effective arithmetic speed, S, in ISO (ASA), and a lens
aperture setting expressed as an f/stop number, A. These two
parameters are all that is needed to calculate the light dose to the object:
illumination, L, in lux multiplied by time, t, in seconds, Lt, a formula for
which has been derived elsewhere:673
Lt = 0.189(1+m)2 A2/S
Where: m is the magnification = image size/object size.

From this it may be seen that, for example, a camera set to a lens
aperture of f/11, so A = 11, and an effective speed of S = 100 ISO, (e.g.
for ‘fine grain’ film or ‘low noise’ digital imaging), the minimum light
exposure of the object Lt is calculated to be 0.36 klx s or 0.1 lx hr, if m =
0.25, i.e. if the camera image is about 1/4 the size of the original object
(e.g. an object ca. 10x8 in. photographed on ‘120’ roll film).
The setting-up exposure: ‘modelling lights’
The light exposure inflicted by the actual photography is only part of
the total dose sustained by the object being copied: the exposure when
setting-up, framing and focussing the image can be even greater. For
instance, even if the setting-up is conducted under a low light level of 50
lux, one minute spent doing this will incur an exposure of 3 klx s, which
in some cases is greater than that of the actual photographic exposure. It
is important that the operating protocols for copying should minimise
271
unnecessary additional exposure of this kind, either by using very low

level modelling lights, or preferably by setting-up on a ‘dummy’ or
surrogate object of the same size, and only substituting the precious
object at the last moment when ready. It is important to assess, and
include in the evaluation, any exposure incurred in this way. On past
occasions, more damage has been done to precious sensitive
photographs by neglectful and unnecessary exposure to the full
illumination at the copying easel, than has been caused by the
photographic exposure itself.
(b) The sensitivity of the object to light
Asstated in §9.4.2, this is conveniently expressed as the threshold
exposure (also in klx s) which causes a Just Noticeable Difference (JND) in
the object. The condition of the object before and after exposure cannot
be directly compared, so either identical reference specimens are
required as ‘controls’, or precise densitometer readings must be taken of
accurately-located image regions before and after exposure. A caveat
here: densitometers generally have very intense light sources (ca. 500
klx) and can cause perceptible damage to very sensitive objects over the
sample area within a short time. The readout from a densitometer is in
optical density, which is defined as log10(opacity), where the opacity is
defined as: incident light intensity/ transmitted or reflected light
intensity, depending on the mode. The optical density has no units – it is
a pure number.
Nearly all treatments of photographic fading published to date tend to
assume that the Law of Reciprocity holds; e.g. the change (which may be
fading or fogging) caused by exposure to a light source of, say, 5000 lux
for one hour, will be the same as the change caused by a source of 50 lux
in 100 hours. While this may be generally true for silver images, it
certainly is not valid for cyanotypes, which recover from fading by aerial
reoxidation.
(c) The ‘cost’ of making one copy
This is expressed in terms of the ‘potential damage’ to the object as
the quotient of light dose (Table 9.7) and threshold exposure (§9.4.2):
c = a/b so c is the number of JNDs inflicted on the object by the
copying procedure.
Curators and conservators are thus assisted in formulating their own
criteria as to what may constitute an acceptable value of c, as a
compromise between the conflicting demands of conservation ethics,
commercial factors, and scholarship. An acceptable value of c will usually
272
be fractional, i.e. smaller than 1 – probably much smaller – so a more

convenient way of expressing the ‘cost’ is:
1/c = b/a where 1/c can be usefully regarded as the number of
times that the object could be copied with the specified method
before it begins to sustain a ‘perceptible change’ i.e. the accumulated
exposure reaches the threshold exposure. This assumes, of course, a
‘worst case scenario’ that the potential damage is arithmetically
cumulative, which may not always be the case; for example the
assumption of exposure reciprocity is not valid for the fading of
cyanotypes, which recover their densities on exposure to air in the dark.
An example calculation: A cyanotype photographed by photo-flash
illumination, using a 35 mm camera set to a speed of 25 ISO (arithmetic)
and an aperture of f/16:
The light dose (a) is calculated from the value 0.3 klx s (Table 9.7) for
100 ISO at f/11 (assuming a magnification of 1/8) modified thus: there
are two ‘stops’ less film speed, and one stop smaller aperture, meaning
three stops more light exposure is needed in all, arithmetically: 2x2x2 =
8 times, so the light exposure = 8 x 0.3 = 2.4 klx s.
Again, we must add the setting-up exposure, say 3 klx s, to give a total
(a) = 5.4 klx s.
The sensitivity (b) of cyanotype under high intensity illumination is ca.
200 lx hr (§9.4.2) which converts to 720 klx s.
So the cost (c) is 5.4/720 = 0.0075 JNDs, which is perfectly ‘safe’. The
cyanotype could be photographed in this way 1/0.0075 = 133 times
before reaching the theoretical threshold of perceptible damage.
9.4.4 Avoidance of alkali
The extreme sensitivity of cyanotypes to alkali suggests that mildly acidic
conditions are preferable for their storage, even if this is not deemed
ideal for the paper substrate. From the evidence cited above, there can be
no doubt that cyanotypes are destroyed by mild alkali (pH 8 to 10) much
more rapidly than cellulose paper is destroyed by correspondingly mild
acid (pH 6 to 4). Confronted with this dilemma, the conservator should be
prepared to subordinate the best interests of preserving the paper
substrate to the best interests of preserving the image upon it. Any sort
of de-acidification treatment, aqueous or non-aqueous, is therefore out
of the question for cyanotypes, because it will certainly damage, and
probably totally destroy, the image. However, Prussian blue that has been
treated with nickel(II) is much more resistant to alkaline hydrolysis, as
273
was shown by Holtzman and confirmed for cyanotypes in the present

work; it opens the possibility that such cyanotypes could be treated with
a de-acidifying bath at about pH 9, such as magnesium bicarbonate,
without serious loss, but this has not yet been tested as a conservation
procedure. The preliminary treatment with nickel(II) would itself cause
some loss from the image, as indicated above, so such a treatment would
have to be carefully justified in the first place. Even simple aqueous
washing should be avoided because of the risk of image loss through
peptisation of the Prussian blue; if washing is considered essential for
preservation, then float-washing of the verso only, or damp-pack
treatment might be considered.
It is fortunate for cyanotypes that, among the likely volatile impurities
in the atmosphere, alkaline gases are much less common in occurrence
than acidic gases. Cyanotypes should never be stored with other types of
process that might emit alkaline vapours, for instance diazo-copies,
which are usually treated with ammonia as part of their processing (and
are also, quite commonly, blue). Care should also be exercised with
proprietory cleaning agents for picture-glass, which are sometimes used
when framing a work for exhibition: some such cleaners can contain
ammonia, so all should be pH tested before use. Having made this
warning, it should be stated that the effect of ammonia gas – which is an
intensification to a violet colour - seems to be quite reversible on re-
exposure to pure air. It is only aqueous ammonia that causes hydrolysis.
For the benefit of those making cyanotypes today, it may be useful to
point out that precautions taken in their preparation could improve their
resistance to light-fading and alkaline hydrolysis. Washing of the print
should be thorough so that there is no trace of residual sensitizer.
Moreover, the new cyanotype method described in Chapter 7
recommends development in dilute acid, which will destroy any unwanted
alkaline buffer that may be present in the paper and ensure that the print
is left in a condition favourable to the Prussian blue, possibly by the
incorporation of acidic hydroxonium cations into its lattice.674 If a shift in
colour towards greenish-blue is not unacceptable to the maker, then
treatment with nickel solution might also be contemplated as a means of
fortifying the resistance of the image to alkali.
9.4.5 Buffered substrates
Buffered papers and mount boards contain chalk (calcium carbonate)
which, with the best of intentions, is included as a reservoir to neutralise
274
any excess acid from the environment or the imposed image, and to
protect the cellulose molecules of paper objects against acid hydrolysis,
which shortens the chain length with consequent embrittlement of the
sheet. The use of buffered materials has become standard conservation
practice in caring for all works of art on paper, and some types of
photograph.675
Increased awareness of the need for conservation of paper documents
and works of art has led to the formulation by the International Standards
Organization of specifications ISO 9706:1994 for Permanent Paper, and
ISO 11108:1996 for Archival paper, whose content is summarised as
follows:676
"Permanent paper (ISO 9706 : 1994)
shall have a pH value between 7.5 and 10.0
shall have an alkali reserve at least corresponding to 0.4 moles acid
[H+] per kilogram of paper (corresponding to at least 2% w/w
calcium carbonate)
shall have tearing resistance of at least 350 mN
shall have a Kappa number less than 5, which means that the paper
may contain only a small amount of easily oxidized material (e.g.
lignin).
Archival paper (ISO 11108 : 1996)
shall be made from cotton, cotton linters, hemp or flax, but may
contain a minor fraction of fully bleached chemical wood pulp
shall have a folding endurance of at least 2.18 (MIT-, Köhler Molin-
or Lhomargy-instruments) or 2.42 (Schopper instrument)
Archival paper shall meet the requirements for permanent paper.
The archival paper is stronger than the permanent paper and can
withstand considerably more handling."
From these specifications, we may calculate the implications of the

presence of this equivalence of 0.4 moles of alkali per kilogram of paper:
If the paper weight is, say, 250 g/m2, then a 1 m2 sheet contains:
0.25 x 0.4 = 0.1 moles of OH- equivalent.
If such a paper is used as the substrate for a cyanotype print by
coating it with cyanotype sensitizer, we can calculate the implications for
the chemistry. A typical coating weight of siderotype sensitizer would be
25 cm3/m2 of a solution ca. 0.5 molar in Fe(III), i.e. a 1 m2 sheet of paper
would absorb 0.0125 moles of Fe(III) on coating. It follows that, per unit
area, the permanent and archival papers contain an 8-fold molar excess
of alkaline reserve (equivalent to [OH]-) available to react with and
275
hydrolyse the Fe(III). There is clearly great potential for damage to the
sensitizer chemistry if only a fraction of the alkali in the sheet can contact
the iron(III) salt, area for area, in the wet state, although the chalk buffer
is quite involatile and unlikely to migrate through dry paper.
This accounts quantitatively for the fact that, over the last few
decades, it has become increasingly difficult for the printer to find papers
which work well for the iron-based processes: the high quality cotton
papers manufactured for the fine-art market of watercolorists, sketchers
and printmakers and the archival document market, all presumably
conform to the above ISO standards. Such papers now always boast that
they are "acid free" – implying that they probably contain an additive of
chalk filler which, unfortunately, will rapidly cause the decomposition of
any iron(III) sensitizer. Most contemporary practitioners have been driven
to the last resort of soaking their chalk-buffered papers in dilute acid in
order to destroy the calcium carbonate before coating them with iron(III)
sensitizer. Unfortunately many practitioners think that oxalic acid is also
suitable for this purpose, but calcium oxalate is just as insoluble as
calcium carbonate. Hydrochloric acid can be used, but recently it has
been found very convenient and effective to employ 10% sulphamic acid
for this purpose.677 Nonetheless, it is a tedious and unnecessary
procedure that tends to degrade the paper surface and strength.
A further caveat when mounting cyanotypes is to avoid applying any
adhesive to the verso of the sheet, within the picture area. Many
adhesives are alkaline and OH– ions can migrate through plain paper,
causing discoloration of the image.
9.4.6 Storage enclosures
In conservation practice, it is now generally advised that cyanotypes
should not be mounted on, or wrapped in buffered materials, in case they
suffer alkaline deterioration. A saturated solution of calcium carbonate is
said to have a pH of 9.4,678 and we have seen that solutions of this pH
can rapidly destroy Prussian blue. To test the extent of the danger from
chalk buffers, processed cyanotype tests of all five sensitizers were
partially coated with a paste of calcium carbonate powder and distilled
water, leaving a ‘control’ area for comparison, and housed in a high
humidity environment (RH ≥ 95%) at room temperature to keep the paste
moist, in a simulation of the ‘worst case scenario’. Within 48 hours, there
was a serious density loss (100∆D ≈ 40 to 60) in the coated areas of all
the sensitizers, compared with the adjacent ‘control’ areas. Calcium
276
carbonate clearly poses a threat to cyanotypes when in direct contact with

the image substance, although it is involatile and has little ability to
migrate through cellulose. In fact, ISO 18902:2007, Imaging materials --
Processed imaging materials -- Albums, framing and storage materials
now recommends the use of buffered papers for storage of all
photographic materials;679 however, cyanotypes are not specifically
discussed or excluded from this protocol.
The dangers of using chalk-buffered enclosures for cyanotypes have
been recently tested by conservation scientists, Held, Burge and
Nishimura, at the Image Permanence Institute.680 In these new
experiments, papers of various pH, (3.1 and 10.3) as well as commercial
buffered (pH 9.3) and non-buffered (pH 6.4) papers were contacted with
test prints prepared using the three cyanotype formulations: Herschel,
Lietze and Valenta (see Table 9.2, §9.1.1). The sandwiches were then
artificially aged in a cabinet at 70 ºC and 86% RH for 15 days. It was
found that all specimens in contact with alkaline and even neutral papers
were seriously faded. Only acidic paper almost fully protected the images.
The Herschel formula was the least stable, followed by the Valenta
formula; prints made using the Lietze formula were the most stable. The
authors conclude:
"The experiments clearly showed that alkaline enclosures should
not be used. However, the data also precludes making a
recommendation for neutral pH enclosures as they also resulted in
significant damage to the test cyanotypes especially the most
sensitive formulation. On the other hand, it cannot be
recommended that cyanotypes be stored in acidic enclosures as
they may deteriorate the paper substrates."
With regard to wrapping materials, it is desirable that cyanotypes

which receive any exposure to light should be allowed access to the air,
particularly when they are returned to dark storage, thus facilitating
density regain by air oxidation which reverses any fading incurred. For
this reason, archival polyester sleeves cannot be recommended as ideal
enclosures for cyanotypes, because they allow the ingress of light but not
of air. A preferable wrapping material is an archival wrapping paper of a
non-buffered type, i.e. Atlantis Silversafe Photostore, in the traditional
‘four flap folder’ style. Being fairly opaque but relatively porous, it will
have the desired properties of attenuating the light but admitting the air.
The volume of air required to accomplish reoxidation is surprisingly
small: for a typical 10x8 inch cyanotype, it is calculated that less than one
277
cc of air would suffice. It has been demonstrated that cyanotypes

contained in anoxic housings (with ‘ageless’ oxygen scavenger) are very
susceptible to light-fading – the very converse of usual recommended
conservation practice! When on exhibition, it is obviously desirable that
the mount for a cyanotype should have a deep window mat, providing a
reservoir of air, which is in communication with the ambient atmosphere.
Let your cyanotypes breathe!
278
Appendix I Chronology of Cyanotype

1706 Prussian blue first prepared by Johann Jacob Diesbach in Berlin
1724 John Woodward publishes method of preparing Prussian blue
1785 Turnbull’s blue being manufactured in Glasgow
1811 Joseph Gay-Lussac isolates and analyses hydrocyanic acid
1818 First Law of photochemistry proposed by Christian von Grotthuss
1822 Leopold Gmelin discovers potassium ferricyanide
1828 John Mercer detects action of light on ferric salts on cloth by test
formation of Prussian blue
1831 Johann Döbereiner notes action of sunlight on ferric oxalate
1839 Announcement of the invention of silver photography
1840 Alfred Smee prepares potassium ferricyanide by electrolytic
oxidation of potassium ferrocyanide
1842 John Herschel discovers effect of light on Smee’s potassium
ferricyanide (23 April) and ammonium ferric citrate (May) to form
Prussian blue. He coins the term ‘cyanotype’ on 16 August.
His paper for Philosophical Transactions accepted 15 June, partly
read to the Royal Society 16 June, and published in September.
1843 Anonymous cyanotype album of ferns printed (March-April)
1843 Anna Atkins begins British Algae: Cyanotype Impressions using
Herschel’s negative-working formulation
1871 Death of Sir John Herschel
1872 Marion & Co launch ‘Ferroprussiate paper’ on the market for
blueprint reprography
1876 First commercial blueprinting machine introduced into the USA
1877 Henri Pellet patents his positive-working cyanotype process
1889 Peter Henry Emerson denounces cyanotype for pictorial purposes
1897 Eduard Valenta prepares green ammonium ferric citrate
1900 Dorel brothers invent gel-lithographic photomechanical process
1950 Blueprint increasingly replaced by diazo copying process
1995 Mike Ware publishes the New Cyanotype process
279
Appendix II Chemistry of Prussian Blue

The chemical investigation of the ‘iron blues’ - for there are several - has
a long and convoluted history. Although Prussian blue has been studied
for over 250 years, it is only in the last two decades that the molecular
structure of one of its forms has been confidently elucidated. The
interpretations put forward in chemistry texts published earlier than the
1970s are all erroneous in varying degrees, and these misconceptions
have been carried over into the technical photographic literature on the
blueprint, none of which now represents the process correctly.681
This appendix attempts to summarise our present understanding of
Prussian blue in its various forms, without detailing the physico-chemical
evidence that underpins the conclusions; those who wish to read more
deeply will find citations of the key references to the primary research
literature.682
Prussian blue is now recognised as a model for a whole class of
transition metal hexacyanoferrates683 - insoluble mixed-valence
compounds which share essentially the same cubic, cyanide-bridged,
open polymeric lattice (see fig. 3.2, §3.2).684 It should be appreciated,
however, that knowledge of the structures of this family of substances is
even now incomplete, owing to three difficulties:
• Most preparations of Prussian blue tend to result in colloidal
precipitates which preclude the obtaining of single crystals for full
3-dimensional X-ray structure determination;
• The presence of disorder due to vacancies in the lattice; and
• The Prussian blue lattice is ‘zeolitic’, in that it offers interstitial
voids to accommodate substantial proportions of impurity ions and
molecules, the incorporation of which complicates its analytical
characterization.
Rather than forming as a single substance, of well-defined
stoicheiometry, the composition of Prussian blue is found to vary
according to the method by which it is prepared. The inherent difficulties
of characterization have been summed up by one principal group of
investigators, thus:
‘In fact, the chemical designation Prussian blue should be regarded
as essentially a generic term for complex materials which may
contain coprecipitated or occluded ions, indefinite amounts of
water, possibly hydrolysed ferrocyanide, variable stoichiometry,
and structural disorder.’685
280
This variability often precludes Prussian blue from lending itself to the
rigorous minuteness of structural analysis that is customarily employed
in modern physico-chemical investigations.
II.1 Methods of preparation

Prussian blue is, in essence, ferric ferrocyanide - or (to use IUPAC
chemical nomenclature henceforth) iron(III) hexacyanoferrate(II). It may be
conveniently precipitated from aqueous media by any of the following
three types of reaction:
1 Iron(III) salts, Fe3+(aq), with hexacyanoferrate(II), [Fe(CN)6]4–.
2 Iron(II) salts, Fe2+(aq), with hexacyanoferrate(III), [Fe(CN)6]3–.
The product of this reaction was earlier thought to be a different
substance, namely ferrous ferricyanide or iron(II) hexacyanoferrate(III),
and was therefore distinguished by the name of Turnbull's Blue, after
the manufacturing chemists of Glasgow. Recent studies, however, have
unambiguously shown it to be essentially the same as Prussian blue.
Nonetheless it is convenient to retain the name as an indication of the
route by which it is prepared.
3 Iron(II) salts, Fe2+(aq), with hexacyanoferrate(II), [Fe(CN)6]4–.
This yields initially a precipitate of the colourless iron(II)
hexacyanoferrate(II), known as Prussian white (or Berlin white,
Williamson’s salt or Everitt’s salt). It is readily oxidised to Prussian
blue, by oxidants such as air, hydrogen peroxide, dichromate,
bromate, chlorate, hypochlorite, etc. When carried out in the presence
of a large excess of ammonium ions, this is the industrial method for
the manufacture, from commonly available materials, of the Prussian
blue pigments of commerce:686
FeSO4 + Na4[Fe(CN)6] + (NH4)2SO4 ® (NH4)2Fe[Fe(CN)6] + Na2SO4
(NH4)2Fe[Fe(CN)6] + H2SO4 + NaClO3 ® (NH4)Fe[Fe(CN)6] + NaCl +
(NH4)2SO4
The composition of Prussian blue depends on the relative proportions
of the reactants used for its preparation: two formulations are usually
defined, the variety resulting when the iron salt is not in excess contains
alkali metal or ammonium cations is conveniently, but inappropriately
termed ‘soluble’ Prussian blue; whereas that obtained with excess iron
salt is called ‘insoluble’ and does not contain any alkali cations.687 Now,
all forms of Prussian blue are highly insoluble in water, with a solubility
281
product of ca. 10–41, but the so-called ‘soluble’ form (Ks = 3 x 10-41),
which contains K+ cations is, in fact, most readily peptized - i.e.
dispersed as a blue colloidal suspension in water (a hydrosol) - thus
giving the illusion of a true solution. This tendency to peptization of
Prussian blue is of great importance to the blueprint process, where its
occurrence during the wet treatment is most undesirable.
Evidence from Mössbauer spectroscopy,688, 689 and photoelectron
spectroscopy,690 leaves no doubt that all forms of Prussian blue and
Turnbull’s blue are, essentially in terms of oxidation states, iron(III)
hexacyanoferrate(II). The octahedrally-coordinated iron(II) is in the
diamagnetic low-spin (t2g6) configuration, as would be expected for the
very high ligand field of C-coordinated cyanide, while the octahedral
iron(III) is in the high spin configuration (t2g3eg2), expected for the weak
ligand field of the nitrogen donor atoms of the cyanide ligands, which is
comparable with that of water in the spectrochemical series. These
configurations account for the measured effective magnetic moment at
room temperature, which corresponds to five unpaired electrons per two
iron atoms.691
Most recently, the compositions, structures, and spectral properties of
several Prussian blue pigments have been determined in samples
obtained by various methods of synthesis, by Louise Samain and her co-
workers. To quote the authors' abstract:
"Several Prussian blue pigments have been prepared by different
methods and characterized by thermogravimetric analyses, high-
energy powder X-ray diffraction, atomic absorption and flame
emission, UV−visible, iron-57 Mössbauer, iron K-edge X-ray
absorption, and Raman spectroscopy. The type of synthesis
influences the hue, tinting strength, and hiding power properties of
the Prussian blue pigments. Two major features appear to be
strongly dependent on the preparative methods, the particle size
and the local disorder. Both a nitrogen atmosphere and an
intermediate aging step of the Berlin white, FeII2[FeII(CN)6], during
the synthesis are required to obtain a highly colored pigment
through the optimization of particle size, minimization in the
perturbations to the FeII−CN−FeIII intervalence electron transfer
pathway, and the minimization of disordered vacancies." 692
Each variety of Prussian blue will now be considered in more detail.
II.2 ‘Soluble’ Prussian blue

This form is obtained when potassium hexacyanoferrate(II) and the
iron(III) salt are reacted in 1:1 molar proportions, or when the
282
hexacyanoferrate(II) is in excess. Its formula is usually written

KFe[Fe(CN)6].H2O, although the water content is variable, and recent
preparations have beeen reported to analyse as KFe[Fe(CN)6].5H2O.693
Because single crystals of this variety have not yet been obtained, its
proposed structure is based only on an early powder X-ray study by
Keggin and Miles.694 These authors found the powder pattern to be
consistent with a face-centred cubic lattice, of probable space group
Fm3m (Oh5): ideally the octahedral [Fe(CN)6]4– units are centred on all the
lattice sites, and all the octahedral holes (each surrounded by six N atoms
of the cyanide groups) are occupied by Fe3+; the centres of the octant
cubes of the face-centred cubic unit cell (i.e. the sites of the formal
‘tetrahedral holes’ - although they are now cubically surrounded by Fe2+
and Fe3+ alternately) were thought to be occupied alternately by K+ and
H2O. The lattice parameter is a = 1016 pm, and the number of formula
units per unit cell, Z = 4. This idealised structure is illustrated in Fig 3.2.
This is a very roomy crystal lattice, having a density of only 1.78 g/cm3,
which is quite low for an inorganic compound. It provides the simplest
and most symmetrical model for discussing the structures of the metal
hexacyanometallate family of compounds, many of which have been
found to be derived from it.
However, the recent investigations of Samain and co-workers have
revisited the structure of this modification, and shown that it is not
ideally regular, but exhibits ~25% lattice vacancies, like the ‘insoluble’
version:
"It can be described with the Pm3m space group, where
approximately one-quarter of the [FeII(CN)6]4− sites are vacant and
where the potassium cation is located at a zeolitic-like position
inside the lattice cavities. The degree of ordering of the [FeII(CN)6]4−
vacancies in all Prussian blues was quantified using atomic pair
distribution analysis, an ordering that is consistent with the iron K-
edge X-ray absorption spectra." 695
Samain et al propose that ‘soluble’ Prussian blue should be

reformulated to take account of the presence of hydroxide ions which are
required for charge neutrality in the defect lattice containing potassium
cations, their formula is approximately: K2FeIII4[FeII(CN)6]3.(OH)2.7H2O.
The infrared spectrum of ‘soluble’ Prussian blue shows a single sharp
peak at 2106 cm–1 in the ν(CN) stretching region, consistent with the
single infrared-active, triply degenerate fundamental, species t1u, for a
regular octahedral Oh moiety [Fe(CN)6]4–. The Raman spectrum of
283
electrodeposited Prussian blue696 is reported to have peaks at 2089,

2123, and 2154 cm–1 which is not consistent with the regular octahedral
structure, for which only two Raman-active fundamentals, a1g + eg , are
predicted. Upon electroreduction these peaks are replaced by features at
2058, 2109, and 2140 cm–1 which are assigned to Prussian white.
However, Samain et al have recently reported Raman spectra both for
'soluble' and ‘insoluble’ Prussian blues that are all very similar and quite
consistent with the octahedral model, showing the intense a1g mode at
2154 cm–1 and the weak eg mode at 2090 cm–1, within ±2 cm–1.
‘Soluble’ Prussian blues of approximate formula M+FeIII[FeII(CN)6]- can
be prepared with other cations, M+, as well as potassium, and many of
these can be ion-exchanged because the the cubic cavity containing the
cation is calculated to have an interior radius of 182 pm,697 compared
with the following radii due to Shannon and Prewitt:698
Li+ Na+ K+ Rb+ Cs+ Tl+ NH4+
90 116 152 166 181 164 151 pm (6-coordinate crystal radii)
153 178 186 202 184 (12-coordinate crystal radii)
Thus, a soluble Prussian blue analogue may be obtained with Na+, K+,
Rb+, and NH4+ as cations in the lattice; Cs+ is irreversibly trapped, and Li+
cannot be incorporated at all because it is effectively much larger than
the free-ion radius due to its firmly-bound primary hydration shell,
Li(OH2)4+. The preparation of Prussian blue using lithium salts yields the
cation-free ‘insoluble’ form of Prussian blue (see below).699 Moreover the
sodium form can be irreversibly ion-exchanged for potassium, which is
preferred.700 When soluble Prussian blue is formed in the presence of
acids, such as HCl, there is evidence that it incorporates some H3O+,
which has a radius similar to NH4+.701
As mentioned above, this misnamed ‘soluble’ form is easily dispersed
in water as nanoparticles (previously termed a colloid) - a tendency which
is believed to be associated with the presence of the cations in the lattice.
Replacement of K+ by NH4+ diminishes the tendency to peptize. ‘Soluble’
Prussian blue may also be rendered less liable to peptization by
treatment with a variety of multiply-charged cations, most notably Fe3+
and Ni2+, which are also believed to be partially exchanged for the K+
ions in the lattice. Rendering the Blue ‘insoluble’ is important, because it
confers a greater degree of resistance to decomposition by alkalies.702
The toning of cyanotypes by Pb2+ or Tl+ salts is very probably due to
cation exchange in the cubic voids of the lattice. The incorporation of
lead is greatly improved in slightly alkaline conditions (§8.6.1), which
284
suggests that it may enter the lattice as the singly-charged cation PbOH+,
which is known to be present in the hydrolytic equilibria for lead(II).
This ability of Prussian blue, which also has a considerable affinity for
the singly-charged cation, Tl+, has been exploited to provide a sensitive
analytical electrode for determining thallium(I) ions in solution.703
Reference has already been made in §3.3 to the effective use of Prussian
blue as an antidote for absorbing toxic or dangerous cations such as
thallium(I) or radiocaesium;704 it has the benefit that no toxic effects of
administering Prussian blue have been found either in animals or
humans.705
II.3 ‘Insoluble’ Prussian blue

This form results from preparations where the iron(III) salt is in large
excess over the hexacyanoferrate(II), i.e. in molar proportions of 4:1 or
more. It does not contain potassium ions in significant stoicheiometric
amounts and, ideally, it has the composition Fe4[Fe(CN)6]3.14H2O,
although the water content is variable, between 14 and 16 molecules.
This form has been successfully grown as single crystals, which has
enabled Ludi and co-workers to carry out a full 3-dimensional X-ray
diffraction study,706 and a neutron diffraction study.707 The structure
determined by these authors can be conveniently described as a defect
version of the ideal face-centred cubic lattice described above, in which
one quarter of the [FeII(CN)6] groups are absent. Ludi et al. were able to
carry out their least squares refinement in space group Pm3m (Oh1),
rather than in the Fm3m (Oh5) of the ideal Keggin and Miles structure; the
fact that the cubic cell, with a = 1017 pm, now appears primitive
suggests at least a partial degree of ordering of the lattice vacancies. The
cyanide sites thus emptied are reoccupied by extra water molecules to
complete the octahedral coordination spheres about the remaining
iron(III) centres, leading to an ‘average’ coordination sphere represented
by [FeIII(NC)4·5(OH2)1·5]. If the FeII(CN)6 vacancies are ordered, one per unit
cell as is thought by Ludi et al, then one quarter of the iron(III) sites will
remain as regular octahedral [FeIII(NC)6]3- and three-quarters will have the
distorted octahedral coordination geometry trans-[FeIII(NC)4(OH2)2]-. If, in
writing the formula, we wish to emphasise the relationship to the ideal
cubic lattice of KFeIII[FeII(CN)6], but with one quarter of the [FeII(CN)6]4-
sites vacant, the stoicheiometry may be appropriately represented as
FeIII[FeII(CN)6]¾.3½H2O.708
285
It is possible that hydrolysis by proton loss can occur at the aquated

iron(III) sites, effectively increasing the charge on the centres to
[FeIII(NC)4(OH)(OH2)]2-, or even [FeIII(NC)4(OH)2]3-. The extent of this
hydrolysis will depend on the pH of the system. It is possible that the
dramatic effect of gaseous ammonia on the colour, which is fully
reversible, may be due to this deprotonation. It has been noted that the
widths of peaks in both infrared spectra and powder X-ray patterns are
significantly narrower for specimens of Prussian blue prepared in acidic
media.709 The presence of different iron(III) centres in this variety of
Prussian blue is believed to be responsible for the broadening and
splitting of the absorption band for ν(CN) stretching vibrations, which are
observable in the infra-red spectrum around 2080 cm–1, compared with
the single sharp peak at 2106 cm–1 in the ‘soluble’ form.710
If each octant cube centre in the unit cell were occupied by one water
molecule, and the ‘missing’ cyanides each replaced by one water, there
should be a total of 8 (‘zeolytic’) + 6 (‘coordinated’) = 14 water
molecules in the unit cell. The neutron diffraction study confirms this, but
with the additional proviso that some of the ‘zeolytic’ water may become
hydrogen-bonded to the ‘coordinated’ water. A proton n.m.r. study of
‘insoluble’ Prussian blue demonstrates that there are, indeed, three
different types of water molecule in the lattice, rather than two, and this
is confirmed by thermogravimetric analysis of the thermal dehydration of
the substance; but the quantitative findings are not fully in accord with
the previous formulation: up to 65°C there is a continuous loss of six
‘zeolytic’ water molecules per unit cell; around 95°C there is a fairly
abrupt loss of four ‘lattice’ waters (presumed to be hydrogen-bonded to
the coordinated waters), and finally at 125°C there is another stepped
loss of four water molecules (presumed to be coordinated in the primary
coordination sphere around the iron(III) centres).711 Prussian blue does
not begin to decompose thermally until the temperature exceeds
200°C.712 Dehydrated Prussian blue is extremely hygroscopic, taking up
water from the atmosphere most readily. This form of Prussian blue can
also take up suitably-sized cations from solution; indeed, there is
evidence that it is somewhat more effective than the ‘soluble’ form as a
detoxifying agent.713
II.4 Turnbull’s blue

The blue precipitate obtained from the reaction of iron(II) salts with
hexacyanoferrate(III) ions is definitely not iron(II) hexacyanoferrate(III), as
286
was originally thought; instead, irreversible electron transfer between the

two iron centres immediately gives an iron(III) hexacyanoferrate(II),714
essentially the same as Prussian blue. However, it will be convenient to
retain the historical name, Turnbull’s blue, when it is important to
emphasise the route by which this modification is prepared.
Wilde and co-workers have obtained infra-red spectra to suggest that
Turnbull’s blue can also exist in ‘soluble’ and ‘insoluble’ forms, with and
without incorporated cation, respectively, but the evidence is indirect.715
Turnbull’s blue is the least well-investigated of the iron blues by
analytical or spectroscopic means, but there is now some evidence
emerging that it may differ significantly in composition from both the
‘soluble’ and ‘insoluble’ forms of Prussian blue previously described.
Recently, further light has been shed on the nature of Turnbull’s blue
by the Mössbauer spectroscopy of Reguera and co-workers,716 who find
that the relative intensities indicate that the two iron centres in the solid
are in the stoicheiometric ratio Fe(III):Fe(II) = 3:2, rather than the 1:1 ratio
of ‘soluble’ Prussian blue or the 4:3 ratio of the ‘insoluble’ variety.
Accordingly, they postulate that an anion must be incorporated into the
structure to restore charge neutrality, suggesting formulae for Turnbull’s
blue of the type FeIII3[FeII(CN)6]2+X– where X– is a singly charged anion
such as OH– or ½(SO42–). In keeping with the model based on a defect
cubic lattice, this formula can be re-written as FeIII[FeII(CN)6]⅔ X⅓, showing
that one third of the hexacyanoferrate(II) ions are absent from the perfect
cubic structure, giving a lattice that resembles ‘insoluble’ Prussian blue,
but with an even greater proportion of vacancies, probably re-occupied
by foreign anions and water. The average coordination geometry of the
iron(III) centres is [FeIII(NC)4(OH2)2], but this is only a mean of a statistical
distribution over several coordination species. In support of this
conjecture, the X-ray diffraction powder patterns for Turnbull’s blue and
‘insoluble’ Prussian blue match rather closely, both having peaks much
broader than those for ‘soluble’ Prussian blue, due to their disordered
lattices. There is additional evidence for the incorporation of anions in
Prussian blue from X-ray fluorescence element analysis, and infrared
spectra.717 The whole family of Prussian blue related compounds is
compared in Table II.1, where fr denotes the fraction of iron(III) sites
occupied in the ideal cubic lattice.
287
Substance Formula fr Oxidation States

Soluble Prussian blue KFe[Fe(CN)6] 1 KFeIII[FeII(CN)6]
Insoluble Prussian blue Fe4[Fe(CN)6]3 ¾ FeIII[FeII(CN)6]¾
Turnbull’s blue Fe3[Fe(CN)6]2 X ⅔ FeIII[FeII(CN)6]⅔ X⅓
Prussian white K2Fe[Fe(CN)6] 1 K2FeII[FeII(CN)6]
Berlin green [Fe(CN)6]⅔ [FeIII(CN)6]⅔
[Fe(CN)6]⅓10/3– [FeII(CN)6]⅓10/3–
Prussian yellow Fe[Fe(CN)6] 1 FeIII[FeIII(CN)6]
Table II.1 Varieties of Prussian blue and related compounds
II.5 Electronic spectrum of Prussian blue

The substance is a prototype for a whole class of ‘mixed oxidation state’
compounds, which are all characterised by intense colours. The
absorption of light by Prussian blue is due to a strong broad band
centred around 700 nm in the red region of the visible spectrum.718
Different authors vary somewhat in the value they assign to the band
maximum, λmax, which may well be a reflection of the variability of the
composition of the substance. Indeed, Mortimer and Rosseinsky have
proposed that the value of λmax can be used to characterise the type of
Prussian blue, with the ‘soluble’ variety peaking at λmax = 690 nm
(14,500 cm–1) and the ‘insoluble’ form at λmax = 730 nm (13,700 cm-1).719
This shift in λmax could be attributed to the differing ligand field splittings
in the Fe(III) centres, where N-bonded cyanide ligands are replaced partly
by water in the latter compound. An alternative view has been put
forward that the apparent shift may be due to partial oxidation of the
hexacyanoferrate(II) centres (see below).
Values for the molar decadic extinction coefficient720, e, of the
‘soluble’ Prussian blue also range rather widely with different
investigators; from 0.98 x 104 dm3mol–1cm–1 to 3 x 104 dm3mol–1cm–1.721
A value of 1.55 x 104 dm3mol–1cm–1, measured for thin films of Prussian
blue,722 will be taken in this work as a mean value for the 'soluble' form.
The interpretation of the electronic spectrum considers the visible
bands to be due to charge transfer: the 700 nm absorption corresponds
to the energy change arising from the transfer of an electron from the
low-spin Fe(II) to the high-spin Fe(III) centre.723,724 This transition
corresponds to the change in one-electron configurations:
{FeIII(t2g3eg2) FeII(t2g6)} ® {FeII(t2g4eg2) FeIII(t2g5)}*
288
where the star denotes an excited state. The transition is allowed because
it involves transfer of an electron from a t2g orbital on one centre to a
similar orbital on the other, with no change in net spin.
There is also a weaker absorption band at 400 nm (25,000 cm–1) in
Prussian blue, corresponding to the transition arising from the
configuration change:
{FeIII(t2g3eg2) FeII(t2g6)} ® {FeII(t2g3eg3) FeIII(t2g5)}*
Here the electron is transferred from a t2g orbital on one centre to an
eg orbital on the other, which is forbidden by local orthogonality. The
difference in energy between the two bands can be seen to be equal to
-1
the ligand field splitting (10,500 cm ) for an FeIIN6 moiety, which is
further evidence for the correct assignment of the iron oxidation states.
More precisely expressed,725 the permitted absorption corresponds to a
transition from the totally symmetric ground state, 6A1g, to two near-
degenerate excited states of 6T1u symmetry.
II.6 Iron(III) aquapentacyanoferrate(II)

It has long been known that the monoaquapentacyanoferrate(II) complex
anion can be obtained from hexacyanoferrate(II) by a photoaquation:726
hν + [Fe(CN)6]4– + H2O ® [Fe(CN)5(OH2)]3– + CN–
and that this complex anion is also capable of yielding an ‘iron blue’
with iron(III). Only recently, however, has this corresponding ferric
aquapentacyanoferrate(II) (FAPCF) been isolated in pure form, as
FeIII[FeII(CN)5(OH2)], and characterised by X-ray diffraction of the
powder.727 The substance exhibits a moderate contraction, but no
significant distortion of the cubic lattice. The visible absorption spectrum
differs from that of Prussian blue in that the 680 nm band suffers a
bathochromic shift to around 780 nm, due to the diminished average
ligand field splitting at the aquated iron(II) centres. Visually, this
corresponds to a more greenish blue. The electrochemical behaviour
resembles that of Prussian blue, showing reversible oxidative and
reductive waves at similar potentials. The infrared spectrum of FAPCF is
reported to show a strong very broad band at 2060 cm-1.
The complex aquated anion [FeIII(CN)5(OH2)]2– is likewise formed by
hydrolysis of hexacyanoferrate(III), promoted by light, acids, or by
‘overoxidation’ of the hexacyanoferrate(III):
[Fe(CN)6]3– + Cl2 + H2O ® [Fe(CN)5(OH2)]2– + Cl– + ClCN
289
It is probably present in ‘old’ solutions of potassium

hexacyanoferrate(III), which discolour from yellow to olive-brown. An
early investigation of this problem also noted:
‘Some crystals of potassium ferricyanide have a brownish-green
tinge instead of the bright ruby-red colour...due to the presence of
a small quantity of the violet aquopentacyanoferrate(III).’728
The potassium salts are isomorphous and can co-crystallise as the

greenish-brown salt 3K3[Fe(CN)6].K2[Fe(CN)5(OH2)]. The presence of the
aquated complex anion may be responsible for some of the colour
differences observed between cyanotypes, especially the greener tones of
older 19th century specimens made with impure hexacyanoferrate(III),
which was often prepared by chlorine oxidation.
II.7 Investigation of cyanotype Prussian blue

The process of forming a traditional negative-working cyanotype, where
an iron(II) photoproduct encounters hexacyanoferrate(III) in the sensitizer,
corresponds in part to the preparative method for Turnbull’s blue, with
the hexacyanoferrate(III) always in excess; however it should also be born
in mind that some of the image substance (that which has been
‘solarized’) will result from the re-oxidation of Prussian white by air,
which is believed to yield the ‘soluble’ form of Prussian blue. The crucial
question for the present investigation is: which of the forms of Prussian
blue described above most nearly represents the pigment of a cyanotype?
Or are the process and the substance so variable as to make this question
meaningless? In situ analysis and spectroscopic investigation of the
Prussian blue of a cyanotype are made difficult by the relatively small
amounts of the substance present and the predominance of the cellulose
substrate. Infrared spectroscopy usually provides a diagnostic
‘fingerprint’ for a substance; although the various types of Prussian blue
are not very well differentiated by their infrared spectra, it is possible to
distinguish the ‘pure, undiluted’ forms by the stretching vibrations of the
cyanide groups, n(CN), as indicated in Table II.2.
In the case of the simple, non-defect, face-centred cubic lattice
assumed for ‘soluble’ Prussian blue, Prussian white and Prussian brown,
the site symmetry for the Fe(CN)6 moities is regular octahedral, point
group Oh. Vibrational theory leads us to expect a single feature in the
n(CN) region of the infrared: the triply degenerate infrared active t1u
normal mode. For the lattices having anion vacancies, the consequent
290
lowering of site symmetries will give rise to broadening and splitting of

this degenerate fundamental, as is seen in the spectra of ‘insoluble’
Prussian blue and Turnbull’s blue which have envelopes containing
unresolved bands.
The infrared spectra of cyanotypes, prepared by various formulae,
were obtained using a Matson ‘Genesis’ Fourier-transform infrared
instrument with diffuse reflectance attachment, at the Chemistry
Department of Manchester University.729 Results were found to be
essentially the same for all preparations involving the two-component
sensitizers, regardless of their composition. All these cyanotype spectra
showed a single n(CN) absorption band peaking at 2090 cm–1 which was
broad and had indications of a shoulder at ca. 2060 cm–1; this pattern
does not correspond exactly with any of the ‘control’ spectra.
Substance Formula ν(CN)/cm–1 Fe(II) Sites

Soluble Prussian KFeIII[FeII(CN)6].H2O 2106 vs sp Oh
blue
Insoluble Prussian FeIII[FeII(CN)6]¾.3½H2O 2106 m Oh (¼)
blue 2080 vs D4h (¾)
Turnbull’s blue FeIII[FeII(CN)6]⅔X⅓ .H2O 2112 vs Oh (⅓)
X = anion- 2052 m D4h ?(⅔)
Prussian white K2FeII[FeII(CN)6] 2085 vs sp Oh
Berlin green K10/3[FeIII(CN)6]⅔[FeII(CN)6]⅓ 2074
2169
Prussian brown FeIII[FeIII(CN)6] 2175 vs Oh (Fe(III)
Cyanotype blue FeIII[FeII(CN)6]X.nH2O 2090 vs bd
(Herschel) 2063 sh
Proto-Cyanotype FeIII[FeII(CN)5(OH2)] ? 2060 vs bd C4v
blue (Smee)
Abbreviations: vs, very strong; m, medium; sp, sharp; bd, broad
Table II.2 Infrared spectra ν(CN) of Prussian blues and derivatives
The rather inconclusive inference to be made from these spectra is

that ‘cyanotype Prussian blue’ is not identical with any one of the
‘established’ varieties, but it more nearly resembles ‘insoluble’ and
Turnbull’s blue than ‘soluble’ Prussian blue. It does not have the regular
crystalline structure, but is defective, possibly amorphous, and most of
the iron(III) centres are likely to be aquated, and some partially
291
hydrolysed. Prussian blue nanoparticles synthesised by the same

photochemical method as used in new cyanotypes have been synthesised
in emulsions and studied by electron microscopy, electron diffraction and
energy-dispersive x-ray analysis.730 The unit cell parameter was found to
be 1013 pm and the TEM showed regular cubic Prussian blue
nanoparticles, 12 nm in size. The visible absorption band was broad 680
nm and the infrared spectrum showed a peak at 2069 cm-1.
The only cyanotypes that produced a significantly different infrared
spectrum were those prepared from the single-component sensitizer of
potassium hexacyanoferrate(III) alone: the peak here is at 2060 cm–1. This
lower frequency is an indication of an increase in the t2g ® p* back-
bonding from iron(II) d orbitals to the antibonding molecular orbitals of
the cyanide, which would be consistent with the complex anion being
aquated [Fe(CN)5(OH2)]2–, as might be expected from the known
photochemistry of hexacyanoferrate(III). Further evidence supporting this
supposition comes from the visible spectra.
UV-Visible absorption spectra (220 - 880 nm) of these varieties of
cyanotype were obtained using a Shimadzu UV-260 instrument in
reflectance mode at the Chemistry Department of Manchester
University.731 There is little difference between the specimens prepared
from the various two-component sensitizers, all show a broad absorption
peaking at ca. 700 nm. But, once again the cyanotypes prepared from
potassium hexacyanoferrate(III) alone are significantly different, in that
the band maximum is shifted to 780 nm. Taken together with the
infrared spectra, this evidence suggests that the form of Prussian blue
produced by the ‘proto-cyanotype’ process is the variety with the
aquated anion, FAPCF, FeIII[FeII(CN)5(OH2)].
II.8 Hydrolysis of Prussian blue

Prussian blue is stable in the presence of acids of moderate strength,
down to pH 1 or 2, below which there is some risk of release of toxic
HCN, but it is very readily decomposed by alkalies at pH 10 and above,
which hydrolyse the iron(III), and take the hexacyanoferrate(II) into
solution, thus:
Fe[Fe(CN)6]– + 3OH– ® ‘Fe(OH)3’ + [Fe(CN)6]4–
‘Ferric hydroxide’, which is more accurately described as a
precipitated hydrous iron(III) oxide gel, only retains the capability of
being dissolved by dilute acids while freshly precipitated; immediate
292
acidification can therefore reverse this hydrolysis reaction, regenerating

Prussian blue. However, ‘ferric hydroxide’ slowly transforms, over a
period of a few days, to more highly polymerized crystalline forms, such
as goethite, a-FeO(OH), which cannot be readily dissolved in dilute
acids.732 With time, therefore, the alkaline destruction of Prussian blue
becomes permanent and irreversible.
This is an appropriate point to note that the ‘bleaching’ of Prussian
blue by a concentrated solution of an oxalate is neither a hydrolysis nor a
reduction, but a destruction of the Prussian blue lattice by the small
oxalate ion penetrating it and complexing the iron(III); the products are
completely soluble:
Fe[Fe(CN)6]– + 3C2O42– ® [Fe(C2O4)3]3– + [Fe(CN)6]4–
A 20% solution of potassium oxalate cleanly decolorises and dissolves
Prussian blue, and can be used for stain-removal, writing or drawing on
cyanotypes, or, at a lower concentration, for ‘reducing’ the blue density,
provided the soluble products are removed afterwards by a brief wash in
water.
II.9 Prussian white

The colourless, insoluble reduction product of Prussian blue is iron(II)
hexacyanoferrate(II), otherwise known as Prussian white, Berlin white,
Everitt’s salt or Williamson’s salt. If the two types of iron centre are equal
in number, as is generally believed, the formula must incorporate two
cationic charges for electroneutrality: M2FeII[FeII(CN)6], where M is a singly
charged cation, usually K+ or NH4+.
The substance is readily oxidised by air, in the presence of water
molecules, to Prussian blue:
4Fe[Fe(CN)6]2– + O2 + 2H2O ® 4Fe[Fe(CN)6]– + 4OH–
Many other oxidants will also effect this transformation, e.g. chlorate,
bromate, dichromate, permanganate, and hydrogen peroxide:
2Fe[Fe(CN)6]2– + H2O2 ® 2Fe[Fe(CN)6]– + 2OH–
There is analytical evidence that the Prussian blue formed by
oxidation of Prussian white is of the ‘soluble’ - i.e. cation-containing -
variety. It should be especially noted that all these oxidations generate
hydroxyl ions, tending to make the system more alkaline, and this is
supported by some analytical evidence for the formation of hydroxide in
the system.733 If hydrolysis of the resulting Prussian blue is to be avoided,
293
such re-oxidations are best carried out under acidic conditions. The
electrochemical reduction of Prussian blue by cyclic voltammetry (see
below) shows that the process is fully reversible provided an adequate
supply of suitable cations is available to diffuse in and out of the lattice
to maintain charge neutrality.
The structure of Prussian white has not been fully determined, due to
lack of single crystals, but Keggin and Miles reported observing a cubic
X-ray diffraction pattern for the powder, similar to that of Prussian blue
itself, and with a slightly larger lattice parameter, for which a value of
1036 pm has been proposed.734 It is generally supposed, therefore, that
the structure is regular face-centred cubic, like that proposed for
‘soluble’ Prussian blue, but with all the octant cube centres, which have a
slightly larger radius of 192 pm, occupied by cations, usually K+ or NH4+,
although H3O+ is also a possibility. The absence of colour in the
substance is obviously due to the lack of any facile transition involving
inter-valence electron charge transfer.
II.10 Berlin green

Oxidation of Prussian blue by nitric acid results in ‘Berlin green’ or
‘Prussian green’, which can also be obtained by heating iron(III) chloride
and potassium hexacyanoferrate(III) in the dark, or by the action of
chlorine on a boiling solution of potassium hexacyanoferrate(II).735 It was
originally formulated as iron(III) hexacyanoferrate(III), Fe[Fe(CN)6], but this
presumption is now known to be incorrect. Keggin and Miles found its X-
ray powder diffraction pattern to be cubic, with similar unit cell
dimensions to Prussian blue, and presumed that the lattice was the same
as that proposed for ‘soluble’ Prussian blue and Prussian white, but with
no cations at all in the octant cube centres.
However, this green substance turns out to be an intermediate, which
always contains some iron(II); from electrochemical studies the complex
anion has been formulated as:736
[FeIII(CN)6]⅔[FeII(CN)6]⅓10/3–
although the extent of oxidation of Prussian blue has been shown to
be cation dependent, and alternative formulations have been
proposed.737
It was first pointed out by de Wet and Rolle that a mixture of iron(III)
chloride and hexacyanoferrate(III) is evidently powerfully oxidising,
because the chloride anion is oxidised with the evolution of chlorine738:
294
FeCl3 + [Fe(CN)6]3– ® Berlin Green + Cl22

which, with excess ferric chloride, leads to the formation of Prussian
blue and more chlorine. For this reason, any inclusion of ferric ions in the
developer for a cyanotype leads to blue fogging of the background, due
to reaction with excess hexacyanoferrate(III).
II.11 Prussian brown or yellow

The brown mixture of iron(III) and hexacyanoferrate(III) also rapidly turns
blue when spotted onto cellulose filter paper or any other oxidisable
substrate.739 Ferric ion in the presence of hexacyanoferrate(III) has a very
high oxidising power, much stronger than either ion separately, because
the insolubility of the reduction product, Prussian blue, drives the
reaction and makes the redox potential about +1.5V, and so it forms in
the presence of almost any oxidisable substance. This has provided the
basis for a qualitative analytical spot-test740 - a "classic reaction for the
detection of H2O2",741 but it must be a very non-specific unless H2O2 is
the only oxidisable species present.
It has been suggested that in this system bound cyanide ions
themselves may be oxidised, even in the solid state:
Fe3+ + [Fe(CN)6]3- + 2H2O ® Fe2+ + [Fe(CN)5]3– + NH4+ + CO2
accounting for the fact that solid iron(III) hexacyanoferrate(III) cannot
be isolated from the aqueous system, and the product is always green.
Recently, the sophisticated techniques of cyclic scanning voltammetry
(see below) have enabled the preparation of pure Prussian yellow,
presumed to correspond to FeIII[FeIII(CN)6], on a platinum electrode
surface when the potential reaches ca. +1.5 V on the Hydrogen Scale.742
This potential is high enough to bring about the oxidation of water,
which provides another reason why the substance cannot be isolated in
pure form from aqueous solution:
4Fe[Fe(CN)6] + 6H2O ® 4(H3O)Fe[Fe(CN)6] + O2
II.12 Electrochemical studies of redox behaviour

The standard redox potentials for iron(III)/iron(II) couples only have a
limited utility when applied to the Prussian blue family of substances,
owing to the large influence that insolubility has on the positions of
equilibria. For instance, it was once argued that the formulation of
295
Prussian blue itself as iron(III) hexacyanoferrate(II) was contrary to the

standard redox potentials:
E0(Fe3+/Fe2+) = +0.771 V
E0([Fe(CN)6]3–/[Fe(CN)6]4–) = +0.356 V
which would imply that Fe3+ will oxidise hexacyanoferrate(II), and that
the stable form of the compound ought to be iron(II)
hexacyanoferrate(III). Clearly the iron(III)/iron(II) redox potentials are
considerably raised from their standard values by the formation of
insoluble reaction products such as Berlin green and Prussian blue, so
that a mixture of Fe3+ and [Fe(CN)6]3– becomes capable of oxidising
chloride ions:
E0(Cl2/Cl–) = +1.358 V
or cellulose:
E0(RCHO, 2H+/RCH2OH) = +0.18 V
The magnitude of this insolubility effect on the iron(III)/iron(II) redox
couples is highlighted by the fact that the reverse reaction occurs under
‘standard’ conditions: hexacyanoferrate(III) may be prepared from
hexacyanoferrate(II) by oxidation with chlorine gas:
Cl2 + [Fe(CN)6]4– ® [Fe(CN)6]3– + Cl–
Owing to the high insolubility of Prussian blue (solubility product, Ks ≈
10–41) modifying the standard potentials,743 the electrochemistry has to
be studied by cyclic voltammetry of the solid coated on an inert electrode
surface dipped into suitable electrolytes. It was first shown by Neff in
1978 that a platinum electrode can be coated with Prussian blue.744 Since
then, thin films of the substance, prepared by reductive electrodeposition
from an electrolyte solution containing iron(III) and hexacyanoferrate(III),
have been extensively studied, because of the potential commercial
applications for electrochromic display devices.745 The thin films so
deposited have coating weights comparable with those obtained in
cyanotypes. The precise nature of the Prussian blue obtained
electrochemically has been disputed, but it is now thought that the
‘insoluble’ variety is deposited initially for kinetic reasons connected with
faster flocculation, but the process of reductive cycling of the electrode
through a voltage range in a potassium ion electrolyte leads to its
conversion to the ‘soluble’ form, incorporating potassium ions, which is
the more thermodynamically stable,746 although there is evidence from
infra-red spectroscopy that this conversion may not be complete.747 The
voltammograms show anodic waves, occurring approximately at the
296
following values, expressed on a scale relative to the Standard Hydrogen

Electrode:
E(FeIII[FeII(CN)6]-/FeII[FeII(CN)6]2-)= +0.46 V (White → Blue)
E(Berlin Green / FeIII[FeII(CN)6]-) = +1.17 V (Blue → Green)
E(Fe [Fe (CN)6] / Berlin Green) ≈ +1.5 V
III III (Green → Yellow)
The first two waves are fully reversible in an appropriate electrolyte,
for instance the White ® Blue transition has been cycled as many as 107
times in a single sample;748 but the last oxidation wave to Prussian yellow
is irreversible. Although these potentials are not strictly comparable with
standard redox potentials, they do convey a qualitative idea of the redox
chemistry, and explain why Prussian blue is the most stable member of
this family of substances. The anodic oxidation wave occurring at a
potential of ca. +0.46 V on the Standard Hydrogen Electrode scale
implies that Prussian blue, in its ground-state, is not significantly
oxidising in general chemical terms; it also explains why Prussian white is
readily oxidised by air because E0(O2, 4H+/2H2O) = +1.229 V, and why,
by the same token, Prussian yellow, with a potential of +1.5 V, can
oxidise water to oxygen gas, as was noted above.
It has recently been shown that microcrystalline solid particles of
Prussian blue or Turnbull’s blue, prepared by conventional aqueous
precipitation, can be mechanically attached to suitable electrodes and
that they also display very similar redox properties to the thin films of
Prussian blue generated electrochemically, so the observations
summarised above also seem generally valid for the ‘bulk material’.749
297
Fig. II A cyanotype print featured on the cover of the American Chemcal

Society’s Journal of Chemical Education for May 2008, which contains an
article by the author on Prussian blue.
298
Appendix III Photochemistry of Cyanotype

The phenomenology of the cyanotype process has not yet received a full
chemical explanation, nor has its mechanism been completely elucidated.
Published accounts agree in outline that photo-induced reduction of
iron(III) to iron(II) is the primary reaction, and the iron(II) then reacts with
hexacyanoferrate(III) to give ‘Turnbull’s blue’; but this blue was long
believed, mistakenly, to be iron(II) hexacyanoferrate(III).750,751,752,753,754 It
has not previously been pointed out that the photochemically-produced
iron(II) is still coordinated by carboxylates, and must be divested of its
ligands before it can be incorporated into the Turnbull’s blue lattice.
Furthermore, the extensive ‘solarization’, or tonal reversal which is
observed when most cyanotype sensitizers are well-exposed, has
received explanations which are neither in agreement with one another,
nor photochemically plausible. None of the existing texts gives any
account of the more recent commercial blueprint sensitizers, called Type
B and Type C, which operate on chemical principles somewhat different
to the traditional Type A.
It is the author’s conviction that the solarization of cyanotypes
observed during the printing exposure has a commonality with the
photochemical fading of processed cyanotypes. Both phenomena become
explicable on the basis that Prussian blue itself is a highly photoactive
substance - but only when in intimate contact with an electron
donor. The eventual purpose of this Appendix is to offer some
background to, and evidence for, this assertion, which will help remove
apparent contradictions in the anecdotal evidence regarding the stability
of Prussian blue and the fading and reversal of blueprints and
cyanotypes. However, the appropriate point to begin this appendix is with
a detailed account of the photochemistry underpinning the cyanotype
sensitizers, as far as it is currently understood.
III.1 Light-sensitive iron(III) carboxylates

Photosensitivity has been noted in a number of salts or, more properly,
complexes of iron(III) with the carboxylic acids which are listed in the
Table III.1; many of these occur naturally in the plant kingdom, most
notably oxalic, citric, and tartaric acids. On irradiating the iron(III) salts or
complexes of these acids with light having some blue or ultra-violet
content, electron charge is transferred from ligand to metal: the iron(III)
299
is reduced to iron(II) and the ligand subsequently undergoes oxidative

decarboxylation.
oxalic acid COOH.COOH

glyoxylic acid CHO.COOH
formic acid HCOOH
pyruvic acid CH3CO.COOH
lactic acid CH3CH(OH)COOH
phenyllactic acid CH3C(C6H5)(OH)COOH
citric acid CH2COOH.C(OH)COOH.CH2COOH
isocitric acid CH(OH)COOH.CHCOOH.CH2COOH
tartaric acid CH(OH)COOH.CH(OH)COOH
malic acid CH(OH)COOH.CH2COOH
glycolic acid CH2(OH)COOH
mandelic acid C6H5CH(OH)COOH
hydroxybutyric acid CH3CH2CH(OH)COOH
malonic acid HOOCCH2COOH
succinic acid HOOCCH2.CH2COOH
fumaric acid HOOCCH=CHCOOH
aconitic acid HOOCCH2C(COOH)=CHCOOH
Table III.1 Carboxylic acids having photosensitive iron(III) salts
It appears that one structural feature of such ligands that enhances

photosensitivity, but is not essential for it in all cases, is the presence of
an a-hydroxy group, which enables the oxidative decarboxylation to
proceed without leaving behind an energetically unfavourable carbonium
ion. Thus ferric acetate, by contrast, shows no evidence of photoactivity
at all, and the succinate has only low sensitivity.
A general mechanism for the photochemistry of the a-
hydroxycarboxylato-iron(III) salts was first proposed by Balzani and
Carrassiti:755
hn + [FeIII(RCH(OH)COO)]2+ ® [FeII(RCH(OH)COO·)]2+
FeIII + [FeII(RCH(OH)COO·)]2+ ® 2FeII + RCHO + CO2 + H+
In a number of cases for the acids listed above, the expected ketonic
oxidation products have been identified. The primary photochemical
reaction is assumed to be an intramolecular redox process in the
complex caused by electron transfer from a coordinated anion to iron(III);
300
the radical formed in the primary process can then reduce an unexcited
complex, leading to the formation of the final products with a theoretical
quantum yield (in terms of Fe2+) of f = 2.0, but f is generally observed to
be lower than this owing to competition from the reverse dark reaction.
Fig. III.1 Quantum yield of Fe(C2O4)32-
III.2 Quantitative aspects of the photochemistry

The yield of photoproduct for an iron-based process has been calculated
for typical parameters in a platinum/palladium sensitizer.756 To obtain a
full scale image, exposure times in the order of two minutes were
predicted for a UVA light source delivering a power flux of 50 W/m2 at a
wavelength of 365 nm. This calculation is born out in practice. A similar
approximate calculation for cyanotype, where the formation of the
Prussian blue image is not so efficient, now follows.
The exposure, It, needed to bring about photochemical change in m

moles of substance is given approximately by:
It = mNhc/Alff
301
Where:
I = irradiance in W/m2
t = exposure time in s
m = number of moles of substance transformed by light
N = Avogadro’s number = 6.023 x 1023 mol-1
h = Planck’s constant = 6.626 x 10-34 J s
c = speed of light = 2.998 x 108 m/s
A = area of normal surface irradiated in m2
l = wavelength of radiation in m (assumed monochromatic)
f = quantum efficiency of the photoactive species at wavelength l
f = fraction of the incident radiation absorbed by the photoactive
substance at wavelength l (f is assumed, for simplicity, to be
independent of the exposure - an assumption that will only be good for
small changes).757
Fig. III.2 UV-Visible absorption spectrum of Fe(C2O4)32-

For a two-component system, in which eP and CP denote the molar
absorptivity and molar concentration of the photoactive component, P,
and eF and CF denote the same for the inert filter, F, the fraction of light
absorbed by P in the sensitized layer is given by:
f = (1-10–D )eP CP d /D
302
where the absorbance or optical density D of the layer, assuming the

Beer-Lambert law holds,758 is given by:
D = (eP CP + eF CF ) d
where d is the layer thickness in cm, but cancels out in the expression
for f. In cyanotype sensitized layers, f is strongly wavelength-dependent
owing to the intense absorption spectrum of the hexacyanoferrate(III)
anion, which acts as an ‘internal filter’. Knowing the absorption spectra of
the two components, figs. III.2 & III.3 P (trisoxalatoferrate) and F
(hexacyanoferrate) and the coating weight of the sensitizer, f can be
calculated at various wavelengths. At 365 nm, for a typical cyanotype
coating, f = 0.65, but this falls off rapidly towards longer wavelengths as
the absorption of hexacyanoferrate increases and that of
trisoxalatoferrate decreases, so that at 420 nm, f = 0.03.
Fig. III.3 UV-Visible absorption spectrum of Fe(CN)63-

If the waveband of useful available actinic light is taken as 320-410
nm, a median value is l = 365 nm, (corresponding with the atomic
mercury line at the maximum emission of most UVA lamps).759
Then, inserting values for the physical constants:
Nhc/l = 3.28 x 105 J/mol
303
We may estimate an ‘order of magnitude’ figure for the limiting

maximum sensitivity to light of any photochemical system without
amplification, if we assume optimum conditions for the photochemistry
by setting f = f = 1, giving:
It = 3.28 x 105 m/A J/m2
If it is further assumed that all the photoproduct is converted into
image pigment, and that a just-perceptible image (Dmin = 0.02) is
provided by a coating weight of pigment m/A = 10–4 mol/m2,760 we arrive
at a minimum exposure:
It = 32.8 J/m2
To obtain a tone of mid-density, would require about ten to twenty
times this much exposure;761 to obtain the maximum optical density,
Dmax, corresponding to an image coating weight of about m/A = 10-2
mol/m2,762 would require about 3000 J/m2. These figures for the
sensitivity at the limiting assumption of f = f = 1 may be compared with
the calculations of Brinckman et al, who arrived at a typical figure of only
82 J/m2 for an image of maximum density.763
In the case of cyanotype, the molar coating weight, W, of Prussian blue
needed to provide a maximum optical reflectance density (Dmax ~ 1.55)
may be calculated from its molar extinction coefficient, or absorptivity,
e:764
W = 5Dmax/e
inserting a value of e for 'soluble' Prussian blue (Appendix II.5):
e = 1.55 x 104 dm3/mol/cm

the image coating weight W for a typical Dmax ~1.55 is then:
W = 5 x 10-4 mol/m2
but we cannot identify this with the coating weight, m/A, of the
photoproduct because the reaction forming Prussian blue from the iron(II)
oxalato complex may be less than 100% efficient. (We will assume the
RMM for 'soluble' Prussian blue, requiring one mole of Fe(II) photoproduct
per mole of SPB.
For a UVA source delivering 50 W/m2, the time required in practice for
a full exposure of such a cyanotype sensitizer is found to be about two
minutes, corresponding to an exposure It = 6000 J/m2. This would be
304
sufficient to produce an approximate photoproduct coating weight of

m/A = 10–2 mol/m2. (Expressed per mole of Fe(II)).
Clearly the performance of the sensitizer falls well short of the ideal in
converting iron(II) photoproduct into pigment. Some of the light may be
absorbed by the self-masking effect of the Prussian blue product as it is
formed. It is also instructive to compare the coating weight of image
pigment with the original coating weight of the sensitizer before
exposure and processing. For a typical coating, experience shows that an
appropriately-sized paper will absorb, per square meter, ca. 25 cm3 of
sensitizer solution, which has a concentration, typically, of 0.3 M, giving
a sensitizer molar coating weight of 7.5 x 10–3 mol/m2. Comparison with
the value found for the coating weight, W, of Prussian blue actually
formed shows that, even for the maximum densities, only about 6% of the
sensitizer is converted into image substance.?
III.3 The Jones-Condit Equation

The standard camera-image equation connecting subject luminance with
image-forming illumination at the focal plane is:
E = L(1-f/u)2(cosq)4V T F/4A2
Where:
E is the illumination at the focal plane in lux
L is the subject luminance in apostilbs
f is the lens focal length in m
u is the subject distance in m
q is the ‘off-axis’ angle (assumed to be 12°)
V is the vignetting factor (assumed to be 1.0)
T is the lens transmittance (assumed to be 0.9)
F is the flare factor (assumed to be 1.03)
A is the lens aperture f/d
Taking the case where the subject is focussed at infinity, and inserting
approximate values for the other parameters we get:
E = 0.212 L/A2 lux
Which is the usual numerical form of the Jones-Condit equation.
Now, converting from illuminance, E, to actinic irradiance, I, and
taking the subject to be mid-grey, reflectivity 0.18, and irradiated at an
average actinic solar value of 38 W/m2 we get:
I = 1.45/A2 W/m2
305
Inserting the value of It from Appendix I, we find the exposure time, t,

to produce a mid-density image is given by:
t = 258 A2 seconds
t = 4 A2 minutes
as a convenient approximation for the ‘order of magnitude’
calculations.
III.4 Photochemistry of trisoxalatoferrate(III)

Of the several photosensitive carboxylato-complexes of iron(III), the
trisoxalatoferrate(III) anion is well-characterised by the determination of
its structure, both in the solid and in aqueous solution.765 As the
potassium or ammonium salt, its photochemistry has been by far the
most extensively investigated,766 and it provides the basis for a well-
known chemical actinometer.767 The stoicheiometry of the overall
photochemistry is:
hn + 2[Fe(C2O4)3]3– ® 2[Fe(C2O4)2]2– + C2O42– + 2CO2
The measured quantum yield per mole of iron(II) formed in aqueous
solution is approximately 1.2 for radiation of wavelengths between 250
and 420 nm, falling slightly to 0.9 at 500 nm, but very sharply
thereafter.768 For the photolysis of the potassium salt in the crystalline
solid state, however, the quantum yield is significantly lower,769 and the
iron(II) photoproduct does not appear to be the simple oxalate, FeC2O4,
but may be a bridged dimer, K6[FeII2(C2O4)5].770
A mechanism for the photolysis in aqueous solution was first
suggested by Hatchard and Parker; it proceeds from the initial formation
of a radical anion by electron transfer from one oxalate ligand to the
iron(III), reducing it to iron(II):771
hn + [FeIII(C2O4)3]3– ® [FeII(C2O4)2(·C2O4)]3-
The ligand radical will be in dissociative equilibrium with the complex:
[FeII(C2O4)2(·C2O4)]3- = [FeII(C2O4)2]2– + ·C2O4–
Reaction of the radical anion, ·C2O4–, or even its partially coordinated
complex, with a further molecule of the original iron(III) complex enables
the transfer of the second electron from the ligand, which is then lost as
two molecules of carbon dioxide, leaving another mole of the iron(II)
oxalato complex:
[FeIII(C2O4)3]3– + ·C2O4– ® [FeII(C2O4)2]2– + C2O42– + 2CO2
306
The quantum yield for ferrioxalate photoreduction in this process

should ideally be f = 2; but the competing reverse dark reaction tends to
diminish this. The quantum yield is found to be largely independent of
pH. Under acidic conditions, ca. pH 2, the colour of a solution of
trisoxalatoferrate(III) changes from emerald green to light yellow,
corresponding to the formation, as the predominant species, of the
hydrated complex ion [FeIII(C2O4)2(H2O)2]– which may be isolated as a
crystalline solid with suitable large cations; it is still light-sensitive with
approximately the same quantum yield.772
The iron(II) oxalato-complex photoproduct is a moderate reducing
agent as can be seen from its redox potential:
E0([Fe(C2O4)3]3–/[Fe(C2O4)2]2–) = +0.02 V
and it is capable of liberating ‘noble’ metals from their complexes,
kinetic factors permitting, thus:
[MX4]2– + 2[FeII(C2O4)2]2– + 2H2O ® M + 4X– + 2[FeIII(C2O4)2(H2O)2]–
That the noble metal complexes or salts are reducible thermodynamically
is evident from their large positive redox potentials:
E(PtCl42–/Pt) = +0.73 V
E(PdCl42–/Pd) = +0.62 V
E(Ag+/Ag) = +0.80 V
E(Hg /Hg) = +0.85 V
2+
E(AuCl4–/Au) = +1.00 V
Photographic images have been printed by this means in platinum,
palladium, silver, mercury and gold. In principle, complexes of
ruthenium, rhodium, rhenium, osmium, and iridium are also reducible,
but their use for imaging is inhibited by kinetic factors in some cases,
and by sheer expense in others.
III.5 Composition of ammonium iron(III) citrate

This substance has been available since the 1840s in a variety of ill-
characterised forms, see §4.3. The pharmacopoeias listed it traditionally
as Ferri et Ammoniae Citras and recognised that the ‘...want of chemical
compactness, the loose state in which the iron is combined, precludes
their recognition as well-defined chemical compounds.’773 As a
convenient source of biologically assimilable iron, it has been used
extensively in iron-deficiency therapy.774 It is usually obtained as glassy
amorphous flakes - the ‘scale salts’ of iron - which can have an iron
307
content ranging from 14% to 28%, the colour varying correspondingly

from a light green to a deep red-brown as the iron:citrate ratio increases.
All cyanotypes prior to 1897 were made with varieties of the ‘brown’
scale salt, for which Valenta gave the analysis shown in Table III.2. The
so-called ‘green’ salt was first prepared in 1897 by Valenta,775 who gave
the analysis as conforming to the composition shown in Table III.2, in
which cit4- represents the fully deprotonated citrate anion C6H4O74–, and
Hcit3- denotes the tribasic anion C6H5O73–. L.P. Clerc, in a later publication
(1937), gives somewhat different formulations of the two varieties,776 but
in a subsequent edition of his manual (1954) he declines to offer any
formulae at all, implying just how “ill-characterised” these salts must be.
Most present-day catalogues of the chemical supply houses offer a
product described as the ‘green’ salt which contains rather more than
14% iron - commonly up to about 18% - and this product frequently does
not appear green at all, but light brown. However, it appears to perform
well enough as a sensitizer, especially if excess citric acid is added to
redress the ratio in the composition. Many present-day varieties of the
‘brown’ salt likewise offered in the chemical catalogues are stated to
contain much more iron than 19% - even up to 28% in one instance. The
variety used in the experimental work described in §9 had an iron
content of 19% like that of the 19th century product.
Formula FW Fe % FW Mole
w/w per Fe Fe:cit
Brown form (Valenta)

4Fe(Hcit).3(NH4)3(Hcit).3Fe(OH)3 2030 19.26 290 7:7
Green form (Valenta)

5Fe(Hcit).2(NH4)3(Hcit).NH4(H3cit).2H2O 1956 14.27 391 5:8
Brown form (Clerc)

Fe(Hcit).(NH4)(H3cit).Fe(OH)3.3H2O 615 18.16 308 2:2
Green form (Clerc)

2Fe(Hcit).(NH4)3(Hcit) 733 15.29 367 2:3
Table III.2 Analyses of ammonium ferric citrates

cit4- = C6H4O74– Hcit3- = C6H5O73– H3cit- = C6H7O7–
308
The only structural information available on these ill-characterised

salts is an investigation by electron paramagnetic resonance
spectroscopy, which has revealed that the ‘green’ and ‘brown’ varieties
differ profoundly.777 The spectrum of the green salt shows specific
resonances close to the values found in ferric citrate, indicating a definite
complex or mixture of oligomeric complexes, with a specific
stoichiometry. On the other hand, the brown salt exhibits a broad
featureless resonance of the same character as that seen in hydrated
ferric oxide. It is reasonable to assume that the brown salt is the more
highly polymerized species, consisting of large aggregates of hydrated
iron(III) oxide gel, bound with citrate ligands only on the exterior.
Solution studies on the alkaline hydrolysis of ferric citrate have shown
that the simple anions present at low pH begin to polymerise in the
region of pH 8 to 9, and finally result in a polymer consisting of
nanometer-sized particles containing around 1200 iron atoms linked by
hydroxy and oxy bridges.778 It seems likely that these extensively
hydrolysed forms of the substance will have a much greater tendency to
leave residues of hydrated ferric oxide in the paper, with consequences
both for the colour and the stability of the resulting cyanotypes.
III.6 Photochemistry of citratoferrate(III)

Studies of the speciation of the iron(III)/citrate system in aqueous
solution have shown that a photo-inactive monomer predominates
between pH 0.5 and 1.5, but a photo-active dimer is formed above pH
2.779 The quantum yield at 365 nm has a maximum value of 0.45 at pH 4,
but then tends to drop with increasing pH, possibly owing to hydrolysis
of the iron(III) and further oligomerization. Previous studies have shown
that the rates of attainment of equilibrium become slow above pH 3.780
The existence of dimeric iron(III)/citrate species in solution has
recently received confirmation from the successful isolation of two
crystalline dinuclear citratoferrate(III) complexes with suitable cations,
and the determination of their structures by X-Ray diffraction.781 One
dimeric anion is [Fe2cit2(OH2)2]2–, which has a centrosymmetric structure,
and the other [Fe2(Hcit)3]3– , where cit4– represents the fully deprotonated
citrate anion, C6H4O74–. In both complex anions the iron(III) centres are
octahedrally coordinated, and they are oxy-bridged by the two or three
available alkoxide groups of the ligands, respectively. The carboxylate
groups coordinate in a monodentate fashion. The ability to crystallise
either anion selectively from an aqueous solution containing excess citric
309
acid suggests that these two species are present in equilibrium. It is likely
that they play an important role in the complex biochemistry of iron
transport within living cells.
The initial photochemical reaction product of citratoferrate(III) has
been identified as acetone dicarboxylic acid, but the nature of the iron(II)
photoproduct remains unknown:782
hn + [Fe2(C6H4O7)2(OH2)2]2– ® 2[FeII ]+ CO(CH2COOH)2 + CO2
The acetone dicarboxylic acid may suffer further decarboxylation to yield
acetone finally:
CO(CH2COOH)2 ® CO(CH3)2 + 2CO2
The quantum yield of 0.45 at 365 nm drops to a value of 0.28 at 436
nm. Photolysis by sunlight of iron(III) polycarboxylates occurring naturally
in surface waters has important effects on the complex speciation of iron,
with consequences for its availability in numerous biogeochemical
processes.783
III.7 Photochemistry of hexacyanoferrate(III)

It has long been known that the photochemical decomposition of the
hexacyanoferrate(III) anion is complicated, with several possible reaction
products, including Prussian blue as first observed by Herschel. The
reaction scheme is summarised in the monograph by Balzani and
Carassiti;784 although the reaction paths are not yet fully mapped out, it
is well-established that the first step is a photoaquation with the
formation of aquapentacyanoferrate(III):
hn + [FeIII(CN)6]3– + 2H2O ® [FeIII(CN)5(OH2)]2– + HCN + OH–
Due to the weakness of hydrocyanic acid, the pH tends to rise during
irradiation. At high pH, a protolytic reaction yields the corresponding
hydroxo complex anion:
[FeIII(CN)5(OH2)]2– + OH– ® [FeIII(CN)5(OH)]3– + H2O
A stepwise continuance of this reaction can ultimately produce a
precipitate of ferric hydroxide. The aquapentacyanoferrate(III) ion has
also been shown to undergo a facile reduction to
aquapentacyanoferrate(II), which can ultimately yield iron(II) by
hydrolysis, so the ingredients are available to form Prussian blue, or its
aquated analogue, iron(III) aquapentacyanoferrate(II), FeIII[FeII(CN)5H2O].
310
It has also been shown that the free cyanide ion released in the
photoaquation is photo-oxidised by iron(III) species:785
hn + 2[Fe(CN)6]3– + CN– + H2O ® 2[Fe(CN)6]4– + OCN– + 2H+
The resulting hexacyanoferrate(II) is then available to form Prussian blue
with any liberated iron(III). Quantum yields are very low for all these
photolysis reactions of hexacyanoferrate(III); reported values for the
aquation and reduction reactions are in the order of ϕ = 0.01, which
explains why Herschel’s first method, the ‘proto-cyanotype’, using
hexacyanoferrate(III) by itself, is very ‘slow’ compared with the method
employing ferricitrate, where the quantum yield is around ϕ = 0.4.
III.8 Photochemistry of the blueprint process: 1

In the usual negative-working process, the sensitizer consists of a
mixture of potassium hexacyanoferrate(III) with a photosensitive iron(III)
salt - ammonium iron(III) citrate in Herschel’s original recipe, which is
replaced by ammonium trisoxalatoferrate(III) in later commercial
blueprint papers. Upon exposure to ultraviolet light, the iron(II) salt is
formed photochemically in situ as described above:
hn + 2[FeIII(C2O4)3]3– ® 2[FeII(C2O4)2]2– + C2O42– + 2CO2
In the case of the citrate system the precise identity of the iron(II)
photoproduct is not known, but it is likely to be a citrato- complex.
Prussian blue is then supposed to be formed via the Turnbull’s blue
route, presumably by an overall reaction such as:786
[FeII(C2O4)2]2– + [FeIII(CN)6]3– ® FeIII[FeII(CN)6]– + 2C2O42–
However, this complex reaction cannot occur in one concerted step,
because the carboxylate ligands must be stripped off the iron(II) before it
can be incorporated in the Prussian blue lattice.787 There is a simple piece
of experimental evidence to show that this overall reaction does not take
place readily: if a solution of trisoxalatoferrate(III) is irradiated with UV
light then added to a solution of hexacyanoferrate(III), little or no
Prussian blue is formed. It is more probable that the initial reaction is
direct reduction of the hexacyanoferrate(III):
[FeII(C2O4)2]2– + [FeIII(CN)6]3– ® [FeIII(C2O4)2]– + [FeII(CN)6]4–
which is a simple electron transfer, favoured by the relevant redox
potentials:
E([FeIII(CN)6]3–/[FeII(CN)6]4–) = +0.356 V
311
E([FeIII(C2O4)3]3–/[FeII(C2O4)2]2–) = +0.02 V
The production of more iron(II) during the on-going exposure then
provides the material to form the pigment after the attendant ligands are
removed. Iron(II) does not bind oxalate very strongly, as may be seen
from the overall dissociation constant:
[FeII(C2O4)2]2– = Fe2+ + 2C2O42– b2 = 4 x 10–7
Removal of oxalate will be promoted by the presence of hydrogen
ions, which compete with the metal to form undissociated oxalic acid, for
which the overall equilibrium constant is:
C2O42– + 2H+ = H2C2O4 Ka = 3.24 x 105
Accordingly, the equilibrium:
[FeII (C2O4)2]2– + 4H+ = Fe2+(aq) + 2H2C2O4
lies to the right hand side, with K = Ka2β2 = 4.2 x 104.
The development of the image from such a sensitizer is strongly
promoted by initial treatment in a bath of dilute acid, rather than water,
as may be seen from the D/logH curves in Figs 7.1 and 7.2 in §7.2.8.
Dissociation of the iron(II) oxalato complex thus releases ferrous ions,
which will be competed for by both hexacyanoferrate(III) and
hexacyanoferrate(II), to form both Prussian blue and Prussian white:
Fe2+(aq) + [FeIII(CN)6]3– ® FeIII[FeII(CN)6]–
Fe2+(aq) + [FeII(CN)6]4– ® FeII[FeII(CN)6]2–
The relative proportions of the two products will depend on the

hexacyanoferrate concentrations: in regions of very low exposure,
hexacyanoferrate(III) will predominate, producing Prussian blue; in
regions of heavy exposure the main species present will be
hexacyanoferrate(II), and Prussian white will result. Thus, we are led to an
explanation of the phenomenon of cyanotype ‘solarization’ i.e. the
reversal of the tonal scale in the shadow areas, which fade to a pale
bluish-grey as the exposure proceeds.
Previously, two different explanations have been put forward for this
effect. One, due to Suzuki maintains that it is the direct photoreduction
of hexacyanoferrate(III) on prolonged exposure that provides the
hexacyanoferrate(II) which reacts with the iron(II) oxalato photoproduct to
give Prussian white:788,789
hn + [FeIII(CN)6]3– ® [FeII(CN)6]4–
312
[FeII(C2O4)2]2– + [FeII(CN)6]4– ® FeII[FeII(CN)6]2– + 2C2O42–

The low value of the quantum yield for hexacyanoferrate(III)
photoreduction, mentioned previously, would seem to disallow this
mechanism as a significant contributor to blueprint ‘solarization’.
Moreover, there is no explanation of how this addition of a small
proportion of white pigment to the image can be responsible for the
extensive solarization that is actually observed, in which Prussian blue
already formed is seen to be reduced. The other explanation, offered by
Murray, is more plausible: the heavy exposure in the shadow areas forms
excess iron(II) oxalato complex, which reduces the Prussian blue already-
formed to Prussian white:790
[FeII(C2O4)2]2– + FeIII[FeII(CN)6]– ® [FeIII(C2O4)2]– + FeII[FeII(CN)6]2–
This explanation is also supported by the values of the relevant redox
potentials:
E([Fe(C2O4)3]3–/[Fe(C2O4)2]2–) = +0.02 V
E(FeIII[FeII(CN)6]–/FeII[FeII(CN)6]2–) = +0.46 V
It should be noted that, in order to bring about this reduction, the
iron(II) complex does not have to be divested of its ligands. Furthermore,
it has been shown that ‘soluble’ Prussian blue has a high ionic mobility
and can perform electrochemistry even in the absence of a liquid
electrolyte.791
The partial correctness of Murray’s explanation was confirmed
experimentally by irradiating a solution of pure ammonium
trisoxalatoferrate(III) under a UV lamp for ten minutes, then immersing a
washed and processed cyanotype step test in the solution for five
minutes. The test strip was partially bleached, in comparison with a
control strip that was immersed in an identical, but unirradiated, solution
for the same length of time, and suffered little or no loss. Evidently the
photoproduct is capable of reducing Prussian blue. After washing and
drying this test, it slowly re-oxidised in the air to Prussian blue,
demonstrating that the bleaching reaction had indeed been a reduction to
Prussian white and not a hydrolysis or peptization. An identical
experiment was also made using a solution of ammonium iron(III) citrate,
with the same result. This mechanism may therefore contribute to the
solarization.
However, a further experiment demonstrated that the simple thermal
reaction between the iron(II) photoproducts and Prussian blue does not
313
fully account for the phenomenon. Solutions identical to those in the

previous experiment were irradiated with a test strip of processed
Prussian blue immersed in situ and exposed to the UV lamp. In these
circumstances the reduction to Prussian white was observed to be much
faster and more complete even in the regions of high density. It may be
inferred that the Prussian blue itself becomes photochemically activated,
making it easier to reduce:
hn + FeIII[FeII(CN)6]– ® {FeII[FeIII(CN)6]–}*
[FeII(C2O4)2]2– + {FeII[FeIII(CN)6]–}* ® [FeIII(C2O4)2]– + FeII[FeII(CN)6]2–
where the formulation {FeII[FeIII(CN)6]–}* represents an electronically
excited charge-transfer state of Prussian blue - essentially, iron(II)
hexacyanoferrate(III).
It has been shown that Prussian white can reduce molecular oxygen to
water in aqueous acidic electrolytes.792 The Prussian white in the
‘solarized’ regions of a cyanotype is therefore re-oxidised to Prussian
blue by the oxygen of the air during processing and drying, so the
shadow values slowly regain their density over a period of several hours,
because the oxygen molecule is small enough to diffuse through the
pores in the lattice:
4FeII[FeII(CN)6]2– + O2 + 2H2O ® 4FeIII[FeII(CN)6]– + 4OH–
Alternatively, for a rapid result, re-oxidation may be accomplished by
a bath of dilute (0.3%) hydrogen peroxide:
2FeII[FeII(CN)6]2– + H2O2 ® 2FeIII[FeII(CN)6]– + 2OH–

In the past, a bath of potassium dichromate (ca. 5% to 10%) was
commonly employed as the oxidant:
6FeII[FeII(CN)6]2– + Cr2O72– + 7H+ ® 6FeIII[FeII(CN)6]– + 2Cr3+ + 6OH–
but its high toxicity now discourages this practice.
It should be noted that each of these oxidations consumes protons, so
generating hydroxyl ions, which may build up sufficiently to cause some
hydrolysis of the Prussian blue. The oxidations should therefore be
carried out for preference under acidic conditions.
III.9 Photochemistry of the blueprint process: 2

In the mid-1930s a number of patents were granted claiming
improvements in blueprint sensitizers which operated on a rather
different principle from the traditional mixtures. The photosensitive
314
component was still, almost universally, ammonium trisoxalatoferrate(III),

but the hexacyanoferrate(III) was replaced, partially or wholly, by
potassium hexacyanoferrate(II).793 It might be expected that mixing these
two compounds would instantly lead to the precipitation of Prussian blue,
but this is not so: the stability constants of the iron(III) oxalato complex
are so high, that there is very little free ferric iron present, especially if
the solution is neutral or slightly alkaline, or if the equilibrium is
suppressed by the presence of excess oxalate ligand. The overall
dissociation constants for the complex species are:794
[Fe(C2O4)3]3– = Fe3+ + 3C2O42– b3 = 4 x 10–21
[Fe(C2O4)2]– = Fe3+ + 2C2O42– b2 = 6.3 x 10–17
[Fe(C2O4)]+ = Fe3+ + C2O42– b1 = 4 x 10–10
These very small values governing the concentration of free ferric iron
explain why only a trace of Prussian blue is formed when the sensitizer
components are mixed. The dissociation of the iron(III) citrato complexes
is slightly greater (β = 6.3 x 10–14), but in either case only sufficient
colloidal Prussian blue is formed at equilibrium to impart a blue colour to
the sensitizer solution, and the equilibrium lies well to the left hand side
in the reaction:
[FeIII(C2O4)3]3– + [FeII(CN)6]4– -/-> Prussian blue
Upon irradiation, such a sensitizer will form Prussian white directly
when the iron(II) photoproduct reacts with hexacyanoferrate(II) according
to the overall equation:
hn + [FeIII(C2O4)3]3– + [FeII(CN)6]4– ® FeII[FeII(CN)6]2– + 2C2O42– + 2CO2
The dissociation constants listed above show that the displacement of
oxalate ligands from iron(II) is much easier than from iron(III), as might
be expected. The Prussian white image is subsequently oxidised to
Prussian blue during wet processing. This procedure for making
blueprints offers the possibility of achieving a higher Dmax than the
traditional method, while maintaining high printing speeds, for two
reasons:
1. Hexacyanoferrate(II) has a much lower optical absorbance in the
blue and near ultraviolet regions of the spectrum, so it offers less of an
internal filter effect than hexacyanoferrate(III), which absorbs strongly, as
described above. Hexacyanoferrate(II) may therefore be used in higher,
even equimolar, concentrations without greatly diminishing the speed.
315
2. The colourless Prussian white photoproduct offers less self-

masking to the incident light, which is altogether more efficiently used
for the photochemistry, and not wastefully absorbed by the photo-
inactive components in the system.
Sensitized papers using hexacyanoferrate(II) are referred to in the
industry as ‘blue layers’, or ‘latitude papers’ to distinguish them from the
‘yellow layers’ of the traditional hexacyanoferrate(III) papers (see Chapter
4). Because the unexposed paper has a blue background that is bleached
by light, the exposure initially gives a positive-working image, which is
gradually transformed into a negative-working image by the wet
processing and action of air. The greatest problem with ‘blue layer’
papers is that the background fog of Prussian blue pigment must be
completely washed out of the highlights before the bulk of the Prussian
white is oxidized. This ‘wash-out’ is assisted by dispersing agents such
as sodium hexametaphosphate, (NaPO3)6.795 It appears that this whole
procedure is feasible because Prussian blue is readily peptized whereas
Prussian white is not; the latter does not easily form a colloid, but
immediately flocculates when precipitated.796 To suppress excessive
Prussian blue formation in the sensitizer, extra quantities of alkali-metal
oxalates, citrates, or phosphates can be added, but in moderation,
otherwise they impair the quality of colour and reduce the speed.797
These two different types of blueprint sensitizer have been termed
Type A (‘yellow layer’) and Type B (‘blue layer’). Mixtures, Type AB, were
also recommended in the patent literature.798 There is additionally a Type
C sensitizer, also ‘blue layer’ in appearance, but differing in its
constituents: it uses a mixture of iron(II) oxalate or the bisoxalatoiron(II)
complex with hexacyanoferrate(III).799 Again it may seem surprising that
the equilibria do not allow the instant formation of Prussian blue in
copious amounts from these two components, but excess oxalate is
added to suppress the free ferrous iron concentration. Although the
mixture does become quite strongly coloured, the amount of colloidal
Prussian blue formed is still quite small. Redox potential data (see the
values quoted above) suggest that substantial electron transfer should
take place between these two components to give a mixture of an iron(III)
oxalato complex (which will be photosensitive) and hexacyanoferrate(II):
[FeII(C2O4)2]2– + [FeIII(CN)6]3– ® [FeIII(C2O4)2]– + [FeII(CN)6]4–
316
so that one ends up with a light-sensitive mixture essentially resembling

Type B. The printing characteristics of Types B and C are indeed similar,
and a mixture of the two, Type BC, has also been recommended.
III.10 Photochemical reduction of Prussian blue

In spite of the well-known tendency of blueprints to fade in light, it is
important to acknowledge that there is good evidence for pure Prussian
blue being photochemically inert to a high degree. Thus, the Colour
Index lists its lightfastness as ‘excellent’.800 Recent studies of an aqueous
suspension of pure ‘insoluble’ Prussian blue by Harriman and co-workers
found that irradiation with visible light causes no perceptible
photodecomposition, which would be marked by loss of optical density
and/or the evolution of oxygen, due to the oxidation of water.801
The early reports by Desmortiers and Chevreul that prolonged
exposure to sunlight causes bleaching of Prussian blue and the formation
of some cyanogen, and a shift of colour to a greener hue on re-oxidation,
suggests that very high exposures (ca. 50 Mlux hr) to light with some
ultraviolet content (the sun) may bring about oxidation of some of the
cyanide ligands, with formation of the aquopentacyano complex, possibly
by reactions such as:
hn + 2FeIII[FeII(CN)6]– + 2H2O ® 2FeII[FeII(CN)5(OH2)]– + (CN)2
However, Harriman et al further observed that the introduction of
trisbipyridylruthenium(II) cation into the system of the aqueous Prussian
blue sol renders it photochemically active and enables rapid
photoreduction of the Prussian blue to Prussian white, and the
concomitant oxidation of water. This is a complex system because the
trisbipyridylruthenium(II) cation is itself highly photoactive, but one of the
reactions considered is:
hn + [bipy3Ru]2+ + FeIII[FeII(CN)6]– ® [bipy3Ru]3+ + FeII[FeII(CN)6]2–
in which photoexcitation of Prussian blue as well as the ruthenium
complex may play a part. In another recent experiment, methanol was the
chosen electron donor, but the electron transfer was mediated by the
presence of titanium dioxide, which is photoactive in generating
electron/hole pairs under irradiation; its presence facilitated rapid
photobleaching to Prussian white under visible light.802 These two
experiments were carried out in rather complex systems, differing from
the cyanotype environment, but they do show that the photoreduction of
317
Prussian blue is critically dependent on the presence of other electron-

donating species.
Absorption of light at 400 or 700 nm by the ligand-to-metal charge
transfer transitions in the Prussian blue (see Appendix II.5) leads to an
excited state which can be described approximately as iron(II)
hexacyanoferrate(III), {FeII[FeIII(CN)6]–}*, and which is a more powerful
oxidising agent than Prussian blue in its ground state. Such enhancement
of redox activity in electronically excited states by the absorption of
visible light has been extensively studied in some inorganic complexes
such as the trisbipyridylruthenium(II) cation,803 but there has been little
report of the phenomenon in Prussian blue, whose excited electronic
states have not been fully characterised.
If there is an oxidisable species accessible to the photoexcited
Prussian blue, electron transfer could occur to bring about bleaching of
the blue to Prussian white and oxidation of the electron donor. However,
this mechanism is critically dependent on the half-life of the
electronically-excited state of Prussian blue, which must be long enough
to permit chemical reaction to occur more rapidly than thermal relaxation
to the grond state. In view of the insolubility of Prussian blue, this
reaction would have to proceed in the solid state, and it could only be
facile if the Prussian blue and its electron donor were in close contact.
This becomes possible if the reducing agent is incorporated into the
zeolitic cavities or the anion vacancies of the Prussian blue lattice. Surface
adsorption onto nanoparticle Prussian blue is also a possibility, because
of the high surface area and the fact that Prussian blue has been shown
to be a photoconductive material. In view of the increased negative
charge in the Prussian white lattice, neutrality must be maintained by the
migration of cations into the lattice on reduction, as described in
Appendix II; so a supply of appropriately-sized cations is also necessary
for photo-reduction to take place.
The reaction is easily demonstrated experimentally, and effective
reducing agents can be readily identified, as follows. The potential
reducing agent is applied in solution (concentration ca 5-10%), under
subdued lighting, to half of a small area of paper coated with Prussian
blue, which may be prepared either by exposing and processing
cyanotype paper, or by chemical metathesis and precipitation. The two
areas are compared, still under subdued illumination, for any signs of a
‘dark’ thermal reduction reaction. If there is none, the dry sheet is then
irradiated with visible light, for an exposure of about one kilolux hour
318
(i.e. 15 minutes exposure to a 4 kilolux artificial daylight source,

provided by four standard daylight fluorescent tubes). Fading of the
coated area versus the uncoated control area should be apparent in this
time. Substances that have been found to promote such
photochemically-induced bleaching of Prussian blue include alkali metal
or ammonium citrates and oxalates, and the iron(II) oxalato complex as
demonstrated above. This reaction would appear to offer a possible
mechanism to explain the fading of cyanotypes, by reactions such as:
hn + FeIII[FeII(CN)6]– ® {FeII[FeIII(CN)6]–}*
2{FeII[FeIII(CN)6]–}* + C2O42– ® 2FeII[FeII(CN)6]2– + 2CO2
where the oxalate may be replaced by other oxidisable species, e.g.
citrate. The implications of this type of photochemical sensitization of
Prussian blue for the conservation of cyanotypes are discussed in §9.1.
319
Appendix IV Units of Concentration

The concentration of a solution can be specified in a variety of different
ways, the choice of which may depend on whether the user is a chemist
or a lay person. Those who have worked with or researched early
photographic processes will be aware of this diversity and the occasional
difficulty in making comparisons. Reconciling or interconverting these
different modes of expressing concentration is not altogether trivial, and
sometimes even impossible with complete exactness. Fortunately, high
precision is not usually a prerequisite in this area. The usual measures of
concentration that need be considered are as follows.
IV.1 Concentration as weight percent volume (% w/v)

If a solution is said to have a concentration of C % w/v, this implies that C
grams (g) of the solute are contained in 100 cubic centimeters (cc) of the
solution. This measure of concentration only has meaning if the units are
appropriately chosen - usually as grams per decilitre.
(1 decilitre = 0.1 litre = 100 millilitres or 100 cubic centimeters)804
The convenience of this mode of expression is obvious, because
solutions are most easily dispensed by liquid volume, and the measured
volume in decilitres multiplied by the concentration in these units
provides an exact statement of the weight in grams of solute that is
present in the volume dispensed.
IV.2 Concentration as molarity

This is the ‘chemist’s unit’ of concentration. A molarity of M, implies that
the solution contains M gram molecular weights (or ‘moles’ or ‘formula
weights’) of the solute in one litre (= 1000 cc = 1 cubic decimeter) of the
solution.
A knowledge of molarities enables the chemist to predict exactly the
equivalent combining volumes of solutions of reacting substances.
Molarity is the basis for all analytical chemistry in solution, especially
titrations.805 Although its use is not necessary in everyday photographic
practice, it is desirable to be able to calculate concentrations as
molarities when devising new formulations, or reviewing old ones in
which the level of chemical understanding may not have been accurate.
It is easy to convert the expression for concentration as C % w/v into the
molarity, M, by multiplying by 10 and dividing by the gram molecular
weight, R, (also called the relative molecular mass, or formula weight):806
320
M = 10 C/R moles/litre
R is found by summing the atomic weights of the component elements in
the molecular formula.
IV.3 Concentration as weight percent weight (% w/w)

If a solution is specified as having a concentration of Cw % w/w, then Cw
grams of the solute are contained in 100 grams of the solution (and
likewise for any other unit of weight). Since it is much less convenient to
measure and deliver samples of solutions by weight, rather than by liquid
volume (a balance is much more costly than a measuring cylinder or
pipette), this is a less practical mode of expression. The only advantage is
its independence of the units of weight used, provided they are both the
same; i.e. no units need to be specified.
IV.4 Concentration as volume percent volume (% v/v)

If a solution is specified as having a concentration of Cv % v/v, then a
volume Cv cc of the solute is contained in 100 cc of the solution. This is
obviously a useful measure when the solute is commonly obtained in
liquid form.
IV.5 Early formulations and recipes

Most of the older formulations and recipes for making-up historic
photographic preparations usually prescribe a procedure of the form:
‘X parts (by weight) of solute are dissolved in V parts (by volume) of
water (or other solvent).’
While it is easy to specify and carry out the making-up of a solution in
this way, it also creates a quantitative difficulty, because the dissolution
of the solute in the solvent changes the total volume, so that the solution
finally obtained no longer has a volume equal to V, but one that has been
increased by an unspecified and unpredictable amount that depends on
the individual chemical identity of the solute. Only for very dilute
solutions, say 1% or less, is it a reasonable approximation to assume that
V remains unchanged, and that the concentration may therefore be taken
as % w/v, once the units have been converted to g/100cc.
The concentration of a solution made up in such a way can be
calculated exactly as Cw % w/w by:
Cw = 100 X/(X + V d)
321
where d is the density (or specific gravity) of the solvent in appropriate

units. If the solvent is water at room temperature, the assumption that d
= 1 g/cc is generally accurate enough, so the expression may be
simplified to:
Cw = 100 X/(X + V)
provided that X is expressed in g and V in cc.
Unfortunately, if a solution is made up in this way it becomes impossible
to relate its concentration accurately to the more useful, rational, and
chemically significant units of C % w/v, unless further information - the
density - is available (see below).
IV.6 Modern formulations and recipes

Most practising photographers have some experience in making up
solutions for darkroom work,807 and will be aware that the instructions in
modern formularies now tend to take the form:
‘Dissolve X g of the substance in water (an approximate volume may be
indicated as a guide) and make up the solution with water, with thorough
mixing, to a final volume of V cc.’
It is obvious that this way of making up solutions can be accurately
expressed by a concentration of C % w/v simply defined by:
C = 100 X/V g/decilitre
This method is greatly to be preferred, especially if quantitative chemical
reasoning is to be employed, and it is also the way of working that is
natural for any trained chemist.
IV.7 Interconversion of Cw and C

The information needed to relate values of Cw and C is the specific gravity
or density, D, of the solution in appropriate units:
D = C/Cw g/cc
In general, information regarding the density, D, of a solution of any
particular strength is not readily available (although it has been tabulated
for aqueous solutions of a limited number of common chemicals in the
CRC Handbook,808 for example)
However, tables of solubility in the Merck Handbook809 do provide
values of C, Cw, and D for saturated solutions of quite a large number of
substances, at room temperature. By making an assumption concerning
the effect of the solute in modifying the final volume of solution, it is
322
possible to derive a factor, called f, for any substance, which may then be
used to make an approximate conversion between C and Cw for any value
of the concentration. This assumption is that the ‘effective volume’
occupied by unit mass of dissolved solute is independent of the
concentration. Given this, it can be shown that:
1/C = 1/Cw - f
and this formula can be used to find a value for C, given Cw, where the
incremental factor f is conveniently defined by either:
f = (Dsat - 1)/Csat
or:
f = 1/Cwsat - 1/Csat
in which Dsat , Cwsat , and Csat are the density and concentrations of the
saturated solution, at the same temperature. The factor, f, for a variety of
chemicals important to alternative photography, has been calculated from
the Merck Tables as shown in Table 21
Substance Dsat Csat Cwsat f Vg

g/cc % w/v % w/w 11-100f
Ammonium dichromate 1.18 33 27.9 0.00545 0.455
Potassium dichromate 1.092 14.2 13.0 0.00648 0.352
Ammonium ferric citrate 1.43 97 67.7 0.00443 0.557
Ammonium ferric oxalate 1.26 65 51.5 0.00400 0.600
Potassium ferricyanide 1.187 38.1 32.1 0.00491 0.509
Potassium ferrocyanide 1.173 28.2 24.0 0.00613 0.387
Silver nitrate 2.29 164 71.5 0.00787 0.213
Citric acid 1.311 88.6 67.5 0.00351 0.649
Trisodium citrate 1.272 61.2 48.1 0.00445 0.555
Sodium chloride 1.198 31.7 26.5 0.00625 0.375
Potassium bromide 1.380 56 40.6 0.00679 0.321
Potassium iodide 1.721 103.2 59.8 0.00699 0.301
Potassium oxalate 1.20 34 28.3 0.00592 0.408
Table IV.1 Solubility data for relevant chemicals
Vg is the effective volume in cc ‘occupied’ by the solute per gram of its

weight dissolved, and can be useful for estimating the final volume of the
solution, knowing the volume of solvent taken.
323
IV.8 Conversion of units in obsolete formulae

When working with ‘historic’ formulations, it is advisable to convert all
units of weight to grams, and all volumes to cubic centimeters. To assist
this, the following conversions may be useful.
Apothecaries’ measure was used by early scientists for weighing out

solids and making up formulae:
1 grain = 0.0648 g
1 scruple = 20 grains = 1.296 g
1 drachm = 3 scruples (60 grains) = 3.888 g
1 ounce = 8 drachms (480 grains) = 31.104 g
1 pound = 12 ounces (5760 grains) = 373.242 g
Troy weight has the same basis as Apothecaries measure, the grain, and
was always used for weighing precious materials such as gold and silver
metal:
1 grain Troy = 1 grain Apothecaries = 0.0648 g
1 pennyweight = 24 grains = 1.555 g
1 ounce = 20 pennyweight (480 grains)= 31.104 g
Avoirdupois measure was the common system for weighing and selling
solids; it was adopted by the British Pharmacopoeia in 1864:
1 grain = 1 grain Troy = 0.0648 g
1 ounce = 437.5 grains = 28.35 g
Texts do not always make clear which system their ‘ounces’

(abbreviated ‘oz.’) refer to, but they are more likely to be avoirdupois.
British fluid measure was used for making up volumes of solutions:

1 minim (ca. 1 drop) = 0.0592 cc
1 fluid drachm = 60 minims = 3.552 cc
1 fluid ounce = 8 fluid drachms = 28.413 cc
1 gill = 5 fluid ounces = 142.065 cc
1 pint = 20 fluid ounces = 568.261 cc
1 quart = 2 pints = 1136.522 cc
1 gallon = 8 pints = 4546.087 cc
324
1 fluid ounce of water weighs approximately 1 ounce avoirdupois

A concentration of 1 grain/fluid ounce = 2.28 g/l = 0.228 % w/v
US volume measurement differs from that used in Britain:

1 minim (USA) = 0.0616 cc
1 fluid dram (USA) = 60 minims = 3.697 cc
1 fluid ounce (USA) = 8 fluid drams = 29.574 cc
1 pint (USA) = 16 fluid ounces (USA) = 473.176 cc
1 gallon (USA) = 8 pints (USA) = 3785.412 cc
Specific gravity was commonly used to indicate the strength of some

concentrated solutions, because these could be readily determined with a
hydrometer or areometer. Unfortunately the scale often employed was
the now long-obsolete (and ambiguous):
Beaumé hydrometer scale.
For liquids denser than water, the number on the Beaumé scale, 0, 1, 2,
3, etc., denotes the specific gravity of an aqueous solution containing 0,
1, 2, 3, etc., % w/w of sodium chloride (common salt). If B represents the
Beaumé number, and D is the specific gravity (or density in g/cc) as
above, the relationships between them are:810
D = 1/(1 - B/m)
or, conversely:
B = m(1 - 1/D)
where the parameter, m, has been assigned various values between
144 and 147 on slightly differing definitions of the scale, but may be
taken approximately as m = 145 for nineteenth century measurements at
20°C.
Twaddell’s Hydrometer Scale.
If T represents the reading on the Twaddell scale, the specific gravity D is
given by:
D = 1 + 0.005 T
325
Notes and References
Notes and references

1 Gernsheim, H, and Gernsheim, A, The History of Photography from the
Camera Obscura to the Beginning of the Modern Era, (New York:
McGraw-Hill, 1969), p13;
Newhall, B, The History of Photography, 1839 to the Present Day,
(New York: Museum of Modern Art, 1949);
Eder, J M, History of Photography, Epstean, E (trans), (New York:
Dover Publications Inc, 1978).
The essential idea of photosensitivity in silver halides can be traced
back to Schulze in 1727.
2 Taylor, R, and Schaaf, L J, ‘The Talbot Collection at Bradford’, in Weaver,
M (ed.), Henry Fox Talbot: Selected Texts and Bibliography (Oxford:

Clio Press, 1992), p131
3 Herschel, J F W, Notebook vol III (3 March 1840) p427, Science Museum
Library, London.
4 Our awareness of this aspect of early photography has been greatly
enhanced by Dr. Larry Schaaf’s scholarly contributions to its

historical study.
5 Smith, Graham, Disciples of Light: Photographs in the Brewster Album,
(Malibu: The J. Paul Getty Museum, 1990), p69, 77.

6 Letter, Brewster to Talbot, 18 November 1843, Science Museum Library,
London
7 Talbot, W H F, ‘On Photography without the Use of Silver’, The British
Journal of Photography, 9 (9 December 1864), pp495-6

8 Official Descriptive and Illustrated Catalogue of the Great Exhibition of
the Works of Industry of all Nations, Vol 1 (London: Spicer Brothers,

1851), p441
I am indebted to Roger Taylor for this reference.
9 Emerson, P H, Naturalistic Photography for Students of the Art, (London:
Sampson Low, Marston, Searle and Rivington, 1889)

10 Child Bayley, R, The Complete Photographer, (London: Methuen, 1932),
(10th edn.), p392

11 Wall, E J, ‘The Iron Salts. I.’ American Photography, 16, 11 (November
1922), p682
12 Sutcliffe, Francis Madingly, The Photogram, (1901), pii
I thank Philip Jackson for this quotation

13 See his archive at the Museum of the History of Science, Oxford:
http://www.mhs.ox.ac.uk/collections/imu-search-
page/results/?query=Teasdale&thumbnails=on&querytype=basic
14 Roberts, P, The Royal Photographic Society Collection, (Bath: The Royal
Photographic Society, 1994)

I am grateful to Pam Roberts, Curator of the RPS Collection, for
information.
326
15 I thank Sara Stevenson, Chief Curator of the Scottish National

Photography Collection at the Scottish National Portrait Gallery, for
information on the collection. Dr. Stevenson’s numerous fine
publications on this collection are listed by Buchanan, W, ‘Snapshot
of a Curator of Photography: Sara Stevenson’, Studies in
Photography, Julie Lawson (ed.), (Edinburgh: Scottish Society for the
History of Photography, 1996)
16 Coe, B, and Haworth-Booth, M, A Guide to Early Photographic
Processes, (London: Victoria & Albert Museum, 1983)
17 Coe, B, Collings, T J, and Gill, A T, ‘The Recognition of Photographic
Processes’, (Royal Photographic Society Historical Group, 1976),
republished as a supplement to The Photohistorian, 119 (Bath: Royal
Photographic Society, November 1997); Collings, T J, Archival Care
of Still Photographs, (Sheffield: Society of Archivists Information
Leaflet, nd)
18 Hope, A, The Amateur Photographer’s Handbook (Chicago: The John
Wilkinson Company, 1891), p76
I am grateful to Judy Seigel for providing this reference.
19 Tennant J A, (ed), ‘The “Blue-print” and its Variations’, The Photo-
miniature, 1, 10 (January 1900), p481
20 Newhall, N, A Portfolio of Sixteen Photographs by Alvin Langson
Coburn, (Rochester NY: George Eastman House, 1962), p3
21 Robertson, Frances, “Blueprints as master plans: Photo-mechanical
reproduction, industrial duplication and the machine aesthetic”,
Studies in Photography, Scottish Society for the History of
Photography, 2013; see also idem., Blueprint, Roger Farnham and
Harry Magee (eds.), Glasgow Print Studio, 2013.
22 Priseman, C, ‘Iron Printing Processes’, The British Journal of
Photography, (19 July 1918), p324.
1s. 6d. in Sterling was the same as 7.5 p (new pence, or £0.075).
23 Tennant, J A, (ed), ‘The Blue Print Process’, The Photo-miniature, 8, 81
(1906), pp432-7
24 Tritton, F J, ‘Photography in the Drawing Office’, The Photographic
Journal, (September 1943), p329;
Lathrop, A K, ‘The Provenance and Preservation of Architectural
Records’, The American Archivist, 43, 3 (Summer 1980), pp325-338
25 I am grateful to Reginald Heron for this information.
26 Ware, M, ‘A New Blueprint for Cyanotypes’, Ag+ Photographic, 7
(1995), pp74-80
27 There is a physico-chemical explanation for this, founded on the
energetics of the transitions between molecular electronic states
responsible for the absorption of visible light. Most stable
substances have a relatively large energy gap between their highest
occupied molecular orbitals (HOMOs) and their lowest unoccupied
molecular orbitals (LUMOs), corresponding to absorption of blue and
UV light.
28 Broughton, W B, (ed.) Colour and Life, (London: The Institute of Biology,
1964)
327
29 For readers in the USA, the blueberry is, of course, the exception that
proves the rule.
30 The intensity of the Rayleigh scattering is inversely proportional to the
fourth power of the wavelength. Thus, blue light with its smaller
wavelength is scattered most by the atmosphere. This is also
responsible for the yellowness of direct sunlight; our sun is actually
a blue-white star. The effect is most marked at sunset, when the
path length of light through the atmosphere is greatest.
31 Lynch, David K and Livingston, William, Color and Light in Nature,
(Cambridge: University Press, 1995); see also: Chemistry in Britain,
36, 12 (December 2000), pp37-39; ibid., 37, 3 (March 2001), p18
32 de Saussure, H-B, ‘Description d'un cyanomètre ou d'un appareil
destiné à mesurer la transparence de l'air’, Mem. Acad. Roy. Turin, 4,
(1788-89), pp409
33 Sella, Andrea, ‘Classic Kit: Saussure’s Cyanometer’, Chemistry World,
7:10, (October 2010)
34 Gen. A. J. Pleasanton, The Influence of the Blue Ray of the Sunlight and
of the Blue Colour of the Sky in developing animal and vegetable life,
in arresting disease and in restoring health in acute and chronic
disorders to human and domestic animals, as illustrated by the
experiments of Gen. A. J. Pleasonton, and others, between the years
1861 and 1876. Addresed to the Philadelphia Society for Promoting
Agriculture, (Philadelphia: Claxton, Remsen, & Haffelfinger,
publishers, 1877).
35 ‘Structural’ coloration is due to optical effects such as diffraction,
scattering or interference, in which a selective reflection of some
wavelengths of light occurs through a re-direction of the energy
without loss by absorption.
36 New Scientist, 162, 2192 (26 June 1999), pp34-37
37 Nassau, K, ‘The Fifteen Causes of Color: The Physics and Chemistry of
Color’, Color Research and Application, 12, 1 (February 1987), pp4-
26; idem, The Physics and Chemistry of Color, (New York: Wiley,
1983)
38 It might be noted that neither the Indigofera plant nor the woad plant
(Isatis) are blue: they contain the dye in a colourless or leuco
molecular form, which only acquires colour upon chemical oxidation.
39 The chromatic shift is explained by its modified chemical composition
as dibromo-indigo.
40 Michael, R H and McGovern, P E, Archeomaterials, 1 (1987), p135.
41 Eigen, Edward, “On Purple and the Genesis of Photography, or the
Natural History of an Exposure”, Carol Armstrong and Catherine de
Zegher (eds.), Ocean Flowers: Impressions from Nature, (New York:
The Drawing Center & Princeton University Press, 2004), pp270-288
42 Pastoureau, Michel, Blue: The History of a Colour’ Marcus I. Cruse
(trans.), (Princeton: University Press, 2001)
43 Jung, C G, Psychology and Alchemy, Hull, R F C (trans.), (London:
Routledge, 1980), p213
44 Jacobi, J de, The Psychology of C G Jung, (London: 1951)
328
45 Ferguson, G, Signs and Symbols in Christian Art, (Oxford University

Press, 1954), p151
46 The Wilton Diptych, in the National Gallery, London, is a superb
example.
47 Swami Harshananda, Hindu Gods and Goddesses, (Madras: Sri
Ramakrishna Math, 1987)
I am grateful to Patricia Kumar for this reference.
48 Cirlot, J E, A Dictionary of Symbols, Sage, J (trans.), (London: Routledge,
1988), pp52-6
49 Lüscher, M, The Lüscher Colour Test, Scott, I A (trans), (London: Pan
Books, 1987), pp15-16
50 Kandinsky, W, Concerning the Spiritual in Art, (New York: Wittenborn
Art Books, Inc., 1947), p58
51 Orna, M V. Archaeological Chemistry; Fitzhugh, E, Artists Pigments,
OUP 1997; Butler and Gash, Chemistry in Britain, 952, 1993.
52 Lapis lazuli is probably the same blue gemstone referred to as ‘Kyanos’
by writers of classical antiquity such as Dioscorides, Theophrastus,
and Pliny. See: Stillman, J M, The Story of Early Chemistry, (New York:
Appleton and Company, 1924), pp19, 44. The mineral has been
mined in the mountains of the Hindu Kush for 6000 years.
Chemically it is a sodium/calcium aluminosilicate,
(Na,Ca)8(AlSiO4)6(S,SO4,Cl)1-2 but the blue colour is imparted by the
presence of the unusual trisulphide radical anion, S3–.
53 Irene Brückle, "The Historical Manufacture of Blue-Coloured Paper",
The Paper Conservator, 17, 1993, 20-31. Smalt, a cobalt potassium
silicate glass, has the nominal formula, K2CoSi3O8.
54 Egyptian blue is calcium copper silicate, CaCuSi4O10 see: New Scientist,
165, 2222 (22 February 2000), 44-45; Paul Brack, “Egyptian blue:
more than just a colour”, Chemistry World, 12: 5, (May 2015) 43;
Olivier Morizot, Eric Audureau, Jean-Yves Briend, Gaetan Hagel, and
Florence Boulc’h, “Introducing the Human Element in Chemistry by
Synthesizing Blue: Pigments and Creating Cyanotypes in a First-Year
Chemistry Course”, Journal of Chemical Education, 92, (2015) 74-
78.
55 Thénard's blue is a cobalt(II) aluminate CoAl2O4; it is also known as
Gahn's ultramarine. Cerulean blue, first marketed in 1860, is a
cobalt(II) stannate, CoSnO3.
56 Wall, A H, A Manual of Artistic Colouring, as applied to Photographs: A
Practical Guide to Artists and Photographers (London: Thomas Piper,
1861), pp46-7
57 Jens Bartoll, “The early use of Prussian blue in paintings”, Proceedings
oj the 9th International Conference on NDT of Art, Jerusalem, Israel,
25-30 May 2008
http://www.ndt.net/article/art2008/papers/029bartoll.pdf
58 In the eighteenth century the price of ultramarine was as high as five
guineas an ounce. Following the discovery by Guimet in 1828 of a
method for preparing it synthetically from materials as common as
clay, sulphur, and soda, the price fell to about a shilling per pound.
329
59 http://chazhuttonsfsm.tumblr.com/post/3333051019/edouard-
adam-the-search-for-the-perfect-blue
60 Theroux, Alexander, “Rhapsody in Blue”, Art & Antiques, (April 1990),
pp70-77
61 Gage, John, Into the Blue, Exhibition catalogue “Blue”, The New Art
Gallery, Walsall, 2000

62 Wood, J, ‘The Art of the Cyanotype and the Vandalous Dreams of John
Metoyer’, in The Photographic Arts, (Iowa: University of Iowa Press,

1997) pp32-44
63 The English language is full of metaphors involving ‘blue’ in this
context. It is a moot point whether the language determines the

visual connotation or derives from it.
64 Komar, V, and Melamid, A, Painting by Numbers: Komar and Melamid’s
Scientific Guide to Art, edited by J Wypijewski, (New York: Farrar,

Straus & Giroux, 1997)
See also reviews of this book in The Guardian, (Friday, 16 January,
1998); The New York Times, (4 January, 1998); Bookforum, (Winter
1997), p37.
65 https://www.facebook.com/pages/A-Way-to-Blue/257096124426022
66 https://today.yougov.com/news/2015/05/12/why-blue-worlds-
favorite-color/
67 The scientific system for the measurement of colour is based in the
recommendations of the Commission Internationale de l’Eclairage

(CIE) and is expressed as three coordinates, L*a*b*, in a defined
colour space. However, this system conveys little or nothing to the
general reader, and for a visual reference one of the preferred
industry standards is the Munsell system of hue, value and chroma,
which is embodied in a colour atlas, which is however a rather
expensive reference work. Munsell Book of Color, Munsell Color,
Macbeth Division of Kollmorgen Corporation (Baltimore: 1976)
68 Kornerup, A, and Wanscher, J H, Methuen Handbook of Colour,
(London: Eyre Methuen Ltd., 1978)

69 Lyons, John, ‘Colour in Language’, in Lamb, Trevor, and Bourriau,
Janine, (eds.), Colour: Art & Science, (Cambridge: the University

Press, 1995), pp195-224
70 Zajonc, A, Catching the Light, (London: Bantam Press, 1993), pp13-15
71 Regrettably, Bismarck was not referring to the manufacture of a
pigment at the time. (The fact that the only dyestuff to be named
after him should be brown, can be considered an ‘irony’.)
72 Kirby, J, ‘Fading and Colour Change of Prussian Blue: Occurrences and
Early Reports’, National Gallery Technical Bulletin, 14 (1993), pp63-
70
73 Roscoe, H E, and Schorlemmer, C, A Treatise on Chemistry (London:
Macmillan and Co., 1881), pp104, 655-6; Coleby, L J M, ‘A History of

Prussian Blue’, Annals of Science, 4 (1939), pp206-211
74 A decisive experiment in 1828 was F Wöhler’s simple conversion by
heat of the inorganic salt, ammonium cyanate NH4NCO, into its
330
isomer, the organic animal natural product, urea (NH2)2CO. See:

Poggendorff’s Annalen, 12 (1828), p253.
75 The name “Prussian Blue” has been reappropriated in the early 21st
century by a young duo of female pop singers – whose fascist lyrics
prove to be far from ‘politically correct’. Internet searchers beware!
76 There was a school of rabbinical thought that identified the dyestuff
called tekhelet in the Bible, which was prescribed for the woollen
yarn of priestly apparel, as Prussian blue, but this has been
disqualified; see: Ziderman, I I, ‘Blue Thread of the Tzitzit: Was the
Ancient Dye a Prussian Blue or Tyrian Purple?’ Journal of the Society
of Dyers and Colourists, 97 (1981), pp362-4
77 Egyptian blue must take the title of the earliest synthetic pigment, but
its formula was lost for many centuries, see §1.8.
78 Bartoll, Jens and Jackisch, Bärbel, "Prussian Blue: A Chronology of the
Early Years," Zeitschrift für Kunsttechnologie und Konservierung, 24,
No. 1, 2010
79 Kraft, Alexander, 'On the Discovery and History of Prussian Blue',
Bulletin for the History of Chemistry, vol. 33, no. 2, 2008, pp. 61-
67; idem, 'On Two Letters from Caspar Neumann to John Woodward
Revealing the Secret Method for Preparation of Prussian Blue', ibid.,
vol. 34, no. 2, 2009, pp. 134-140; idem, ' "Notitia Cœrulei
Berolinensis Nuper Inventi" on the 300th Anniversary of the First
Publication on Prussian Blue', ibid., vol. 36, no. 1, 2011, pp. 3-9.
80 ‘Notitia Coerulei Berolinensis nuper inventi’, Miscellanea Berolinensia
ad incrementum scientiarum, 1 (1710), pp377-8.
An English translation of the Latin original by Jo Kirby and Raymond
White may be found in the Appendix to Kirby, J, ‘Fading and Colour
Change of Prussian Blue: Occurrences and Early Reports’, National
Gallery Technical Bulletin, 14 (1993), pp63-70.
81 Woodward, J, ‘Praeparatio Caerulei Prussiaci ex Germania Missa ad
Johannem Woodward’, Philosophical Transactions of the Royal
Society, 33, 381 (January-February 1724), pp15-17.
An English translation of Woodward’s Latin paper can be found in:
Powell, H M, ‘The Beginnings of Co-ordination Chemistry’,
Proceedings of the Chemical Society, (March 1959), pp73-5
82 Stahl, G E, Experimenta, Observationes, Animadversiones, (Berlin:
1731) p281
83 Johann Konrad Dippel was born in Castle Frankenstein in 1673. Having
failed to make gold, he launched an ‘animal oil’ upon the public in
1700. It was hailed as a panacea, presumably on the widely-held
premise that anything so obnoxious must be beneficial.
http://en.wikipedia.org/wiki/Johann_Conrad_Dippel
84 A mixture of alkyl cyanides and nitrogenous organic bases, such as
pyrrole, quinoline and pyridine, obtained by the destructive
distillation of blood, bones and offal. It was also used by Niépce as a
solvent for bitumen.
85 Campbell, A, ‘Keep on taking the oil’, Chemistry in Britain, 33 6 (June
1997), p54.
331
86 McKenzie, C, One Thousand Experiments in Chemistry; with

Illustrations of Natural Phenomena and Practical Observations on the
Manufacturing and Chemical Processes Pursued in the Useful Arts
(London: 1821 and 1825), p161; for an alternative account see: The
Mirror of Literature, Amusement and Instruction, 33, 951 (1 June
1839) p347
87 The French name for the pigment was Bleu de Prusse. An alternative
German name was Preussiches Ultra-Marin.

88 There is a dangerous inconsistency here in the early nomenclature:
although ‘prussic acid’ is unambiguously hydrogen cyanide (or

hydrocyanic acid); ‘prussiate of potash’ does not generally signify
potassium cyanide (also highly toxic), but potassium ferrocyanide
(which is relatively innocuous). This possibly lethal ambiguity is
resolved by some authors, who refer to the latter as ‘yellow prussiate
of potash’ or ‘ferroprussiate of potash’. The German name for this
salt was Blutlaugensaltz – ‘blood residue salt’ – acknowledging one
method of its preparation.
89 Macquer, P-J, ‘Examen chymique du bleu de Prusse’, Mémoire de
mathématique et de physique de l’Académie Royale des Sciences,

(1756), pp60-77
90 Scheele, Traite de l’Air et du feu
91 Nitrogen had meanwhile been identified by Daniel Rutherford in 1772
as ‘dephlogisticated air’.
92 Gay-Lussac, Annales de Chimie, 77 (18??), p128; 95 (18??), p136
93 In Homeric Greek there is evidence that kyanos did not in fact mean
‘blue’ but simply ‘dark’; see Chapter 1.

94 Prussian blue can also claim the more arcane chemical distinction of
being the first coordination compound, or ‘complex’ of a metal to be

synthesised. See Powell, loc cit.
95 The bond here between the atoms of carbon and nitrogen is very
strong - a ‘triple bond’ in chemists’ parlance - so the two bound

atoms tend to act as a single entity, the cyanide radical.
96 Hunt, R, (ed), Ure’s Dictionary of Arts, Manufactures and Mines
(London: Longman, 1860), p545

97 Their ‘Pyroligneous acid’ works were at 100 Camlachie Street, see
Royal Commission on the Ancient and Historical Monuments of

Scotland, John Hume Collection
98 The New Statistical Account of Scotland (1845), Volume VI – Lanark,
City of Glasgow. I thank Dr. Sara Stevenson for this reference.

99 Brückle, I, ‘The Historical Manufacture of Blue-coloured Paper’, The
Paper Conservator, 17 (1993), pp20-31; Bower, Peter, The Quarterly,

31, (August 1999), pp11-17
100 I am indebted to Dr. Roger Taylor for this information.
101 Eder, J M, History of Photography, Epstean, E (trans.) (New York: Dover
Publications Inc., 1978), pp56, 101

102 http://en.wikipedia.org/wiki/Alexey_Bestuzhev-Ryumin
transliterations from the Russian vary.
332
103 Thompson, C J S, The Lure and Romance of Alchemy, (London: G G

Harrap, 1932)
104 The yellow solution of ferric chloride is rendered colourless by
sunlight, corresponding to the reduction of iron(III) to iron(II); some

of the organic component is correspondingly oxidised.
105 Thompson, C J S, The Lure and Romance of Alchemy, New York: Bell
Publishing Company, 1990, p23

106 http://fr.wikipedia.org/wiki/Élixir_d%27or
107 Wittstein, G C, Practical Pharmaceutical Chemistry, (Darby, S, trans. of
2nd German edn.), (London: John Churchill, 1853) p132.

108 Eder, J M, History of Photography, Epstean, E., (trans.), (New York:
Dover Publications Inc., 1978), p159

109 Lathrop, A K, ‘The Provenance and Preservation of Architectural
Records’, The American Archivist, 43, 3 (1980), p331

110 Eder, op. cit., p107
111 Parnell, E, The Life and Labours of John Mercer, (London: Longmans,
Green & Co., 1886). I am indebted to Owain Davies for bringing this
quotation to my attention, and to the late Harry Milligan for first
acquainting me in 1984 with Mercer’s researches. Although ferric
nitrate is not an organic salt of iron, it was doubtless the cellulose of
the cloth that provided the oxidisable material.
112 His later contributions are summarised in Parnell, op cit, and in Dr
Stella Butler’s lecture, Colouring Material: John Mercer (1791-1866)

and his Photographs on Cloth, delivered to the Historical Section of
the Royal Society of Chemistry, in June 1987 at the Maritime
Museum, Liverpool.
113 Döbereiner, J F, ‘Chemische Thätigkeit des Lichts und Erzeugung des
(neutralen) Humboldtits auf chemischem Wege’, Jahrbuche für

Chemie und Physik von Schweigger-Seidel, 62 (1831) pp86-90; this
paper was also summarised in Pharmaceutiches Centralblatt, 2
(August 1831), p383-5; See also Eder, op cit, p177.
114 Talbot, W H F, Notebook M, Folio 34, (before March 1835). ‘Paper
wetted with common mur. acid, & then with Pruss. Potash, turns blue
owing to the presence of iron in the acid.’
I am indebted to Michael Gray, curator of the Fox Talbot Museum,
Lacock, for this transcription from Talbot’s notes.
115 Schaaf, L J, Out of the Shadows: Herschel, Talbot and the Invention of
Photography (New Haven & London: Yale University Press, 1992)

116
MS 478, Archive Collection, Science Museum Library, London.
117
Gernsheim Photography Collection, HRHRC.
118 I am grateful to the Harry Ransom Humanities Research Center,
University of Texas at Austin for supplying photocopies and for

permission to quote from this material.
119 Sir John F.W. Herschel, ‘On the Chemical Action of the Rays of the
Solar Spectrum on Preparations of Silver and other Substances, both

metallic and non-metallic, and on some Photographic Processes’,
Philosophical Transactions of the Royal Society, vol. 131, 1840, pp.
1-59.
333
120 In the Museum of the History of Science, Oxford, and at the Harry
Ransom Center, University of Texas at Austin.
121 Somerville, Mary, 'On the Action of the Rays of the Spectrum on
Vegetable Juices. Extract of a Letter from Mrs. M. Somerville to Sir
J.F.W. Herschel, Bart. dated Rome, September 20, 1845.
Communicated by Sir J. Herschel', Philosophical Transactions of the
Royal Society, vol. 136, 1846, pp. 111-120.
122 Fabbri, Malin, Anthotypes (Stockholm: Alternativephotography.com,
2012).
123 Herschel, Sir J F W, MS Notebook, vol. III, Science Museum Library,
London
124 The sun was, of course, the only practicable light source available to
the early pioneers in photography, and all experiments were at the
mercy of the vagaries of the weather.
125 Letter, Herschel to Talbot, 21 April 1842. NMPFT 1937-4900. See
Schaaf, L J (ed.), Selected correspondence of William Henry Fox
Talbot 1823-1874, (London: Science Museum and National Museum
of Photography, Film & Television, 1994), p36, and Schaaf, op. cit.,
p126
126 ‘Letter from Sir John F W Herschel, dated July 31, 1841’, Report of the
Meeting of the British Association for the Advancement of Science,
Transactions of the Sections (Plymouth: 1841), p40
127 Herschel, Sir J F W, ‘On the Action of the Rays of the Solar Spectrum on
Vegetable Colours, and on some new Photographic Processes’,
Philosophical Transactions of the Royal Society, (1842), pp181-215
128 Schaaf, op cit, p127
129 Smee, A, Elements of Electro-metallurgy, see also: Smee, A, ‘The
Principle, Construction and Use of Smee’s Battery’ (transcribed from
a paper read at the Society of Arts, 1 June 1840), Appendix V in
Smee-Odling, E M, Memoir of the Late Alfred Smee, FRS, (London:
George Bell and Sons, 1878)
130 Gore, G, ‘Electro-Metallurgy’ in The Circle of the Sciences, VII,
Practical Chemistry, (London: Griffin, Bohn, and Co., 1856), pp4, 7-
100
131 See note 15. The ‘sesqui-’ prefix in the first of these names arose
through a misconception at the time of the chemical formulation of
this salt.
132 The name, ‘potassium ferricyanide’ does not conform to the most
recent international (IUPAC) recommendations for chemical
nomenclature, but it probably has wider currency than the approved
form ‘potassium hexacyanoferrate(III)’. It also trips more easily off
the tongue. It is regrettable that the important difference between
this substance and potassium ferrocyanide should reside entirely in
a single vowel. The IUPAC nomenclature is no better, because the
latter becomes potassium hexacyanoferrate(II). Numerous errors
have resulted from misreading this fine distinction. Caveat lector!
133 Smee, A, ‘On the Ferrosesquicyanuret of Potassium’, Philosophical
Magazine, series 3, 17, 109 (September 1840), pp193-201
334
134 Gmelin, L, Schweigger’s Journal, 34 (1822), p325

135 Potassium ferrocyanide was then prepared, as a first step in the
manufacture of the important pigment Prussian blue, by an
obnoxious process of heating potash with animal residues such as
blood and offal or parings of hoof and horn; hence its original
German name Blutlaugensaltz (‘blood-residue salt’).
136 ‘F3/2CP’ is the personal abbreviation that Herschel commonly used in
his notes for ‘ferrosesquicyanuret of potash’ as potassium
ferricyanide was then known.
137 Herschel, Sir J F W, Memoranda, Manuscript Library, Harry Ransom
Humanities Research Center, in the University of Texas at Austin
138 Herschel first coined the word 'cyanotype' in his Memoranda for 16
August 1842; and used it in print in a postscript to his 1842 paper
139 Herschel, Sir J F W, Diary 1833 to 1864, Transcript in the Library of the
Royal Society, MS583-585.
140 The pathways of photochemical decomposition of potassium
ferricyanide are complex, and have not been fully mapped out even
today - see Appendix III.
141 Herschel, Sir J F W, Memoranda, HRC, ref 109
142 Letter, Smee to Herschel, 25 April 1842. Royal Society Library.
143 Letter, Smee to Herschel, 10 May 1842. Royal Society Library.
144 Smee, A, ‘Photogenic Drawing’, Literary Gazette (18 May 1839),
pp314-16
145 Letter, Smee to Herschel, Royal Society Library
146 Christison, R, A Dispensatory, or Commentary on the Pharmacopoeias
of Great Britain (Edinburgh: Adam and Charles Black, 1848), p975
147 The adjective 'chalybeate' derives from the tribe of Chalybes,
metallurgists of Anatolia, living on the east coast of the Black Sea,
who are credited with the first extraction of metallic iron from its
ores, some time around 1500 BCE.
148 Taniguchi, M, Imamura, H, Shirota, T,, Okamatsu, H, Fujii, Y, Toba, M,
and Hashimoto, F, ‘Improvement in Iron-Deficiency Anemia through
Therapy with Ferric Ammonium Citrate and Vitamin-C and the
Effects of Aerobic Exercise’, Journal of Nutritional Science and
Vitaminology, 37, 2 (1991), pp161-171
149 Herschel’s personal diary entry for 17 June 1846 records that he was
prescribed ACI for his stomach complaints. An irony indeed!
150 Smee’s biography by his daughter, which is otherwise fulsome in
cataloguing his achievements, (which were indeed considerable)
makes no mention of this correspondence and collaboration with Sir
John Herschel. See: Elizabeth Mary Smee Odling, Memoir of the Late
Alfred Smee, FRS, (London: George Bell and Sons, 1878)
151 Draft of a letter, Herschel to Smee, 15 June 1842. Royal Society
Library.
152 Although Herschel also thought of platinum as an ideal image
substance, he was unable to make it work with the chemicals
available. Much later, in 1873, in the hands of William Willis, a
successful platinum printing process was achieved. See: Mike Ware,
335
‘The Eighth Metal: the Rise of the Platinotype Process’, in Julie

Lawson, Ray McKenzie and A.D. Morrison-Low (eds.), Photography
1900: The Edinburgh Symposium (Edinburgh: National Museums of
Scotland and National Galleries of Scotland, 1992), pp. 98-111.
153 The word is proposed by Herschel himself in a footnote to The
London, Edinburgh and Dublin Philosophical Magazine and Journal
of Science, series 3, vol. 21, 1842, p. 227. This neologism is also
attributed to Herschel by the editor, Sir David Brewster, in 'Drawing
by the Agency of Light', The Edinburgh Review, vol. 76, no. 154,
January 1842, p. 339:
"Hence Sir J. Herschel considers the name siderotype, taken from the iron
employed in one of the solutions, as preferable …"
The same article is reprinted in The Penny Mechanic and Chemist,
no. 146 3rd series, 29 July 1843, pp. 236-8. Brewster again uses the
word in The North British Review, vol. 7, no. 14, 1847, p. 474. The
word "siderotype" entered the scientific canon in Henry Watts, A
Dictionary of Chemistry, vol. 5 (London: Longmans, Green & Co.,
1868), p. 240.
154 Herschel’s chrysotype (Greek: crusoV = 'gold') furnished permanent,
deep red images made of nanoparticle gold (the ‘Purple of Cassius’
used by ceramicists and stained glass artists), but it failed to gain
acceptance into the photographic repertoire – partly owing to its
obvious expense, but partly also on account of chemical difficulties
over controlling the colour, tonality and contrast of the image. Only
in the last two decades of the 20th century have the photochemical
problems of chrysotype been resolved, see:
Mike Ware, Gold in Photography: the history and art of Chrysotype
(Brighton: ffotoffilm publishing, 2006). See also: Mike Ware,
‘Herschel’s Chrysotype: A Golden Legend Re-told’, History of
Photography, 30, 1, (Spring 2006), pp. 1-24.
155 The later platinum & palladium printing processes due to William
Willis used ferric oxalate.
156 Towards the end of the nineteenth century, Herschel's argentotype
process suddenly fathered a whole family of derivative iron-silver
processes: Van Dyke Brown; Kallitype (Greek: kalli = beautiful);
Sepiaprint; and Brownprint are the self-descriptive names for these
essentially similar processes. Unfortunately they did not enjoy a high
reputation for permanence, unless toned with gold or platinum, but
some are still practised today by the devotees of the hand-made
photograph, see: Christopher James, The Book of Alternative
Photographic Processes, 2nd edn. (New York: Delmar Cengage
Learning, 2008).
157 See, for example: Eder, J M, Ausführliches Handbuch der
Photographie, Part 13, ‘Die Lichtpausverfahren, die Platinotypie und
verschiedene Copirverfahren ohne Silbersalze’, (Halle: Wilhelm
Knapp, 1899), p203-4; Pizzighelli, G, and Hübl, Baron A, Platinotype,
Iselin, J F, (trans.), (London: Harrison and Sons, 1886), p35; Abney,
Capt. W de W, and Clark, L, Platinotype, its Preparation and
336
Manipulation, (London: Sampson Low, Marston & Co. 1895), p156;

Chapman Jones, The Science and Practice of Photography, (London:
Iliffe and Sons, 1904)
158 The application of some modern chemistry, 150 years later, has
enabled me to devise a latter-day version of the chrysotype process

that is less profligate in its consumption of gold, and is capable of
providing images as fine and permanent even as those obtained by
platinotype, but of a more remarkable range of colours. See: Ware, M
J, ‘Prints of Gold: The Chrysotype Process Re-invented’, Scottish
Photography Bulletin, No 1, (1991), pp6-8; idem, ‘Photographic
Printing in Colloidal Gold’, Journal of Photographic Science, 42, 5
(1994), pp157-161
159 Ware, Mike, ‘On Proto-photography and the Shroud of Turin’, History
of Photography, vol. 21, no. 4, Winter 1997, pp. 261-9. See also:
http://www.usask.ca/lists/alt-photo-process-l/200609/msg00420.html
160 Herschel, Sir John F W, ‘On the Action of the Rays of the Solar
Spectrum on Vegetable Colours, and on some new Photographic

Processes’, Philosophical Transactions of the Royal Society, 133,
(1842), pp. 181-221.
161 Now in the photographic collection of the Harry Ransom Center of the
University of Texas at Austin. The items are listed in the Center's

Photographic Collections Database, with the engraving title and year
of publication:
http://norman.hrc.utexas.edu/photoPublic/fullDisplay.cfm?CollID=73
162 Herschel wrote four papers on photographic processes, of which three
were published in the Philosophical Transactions of the Royal

Society, in 1840, 1842 and 1843. The first paper, written in 1839,
was withdrawn before publication, and was long thought lost until
the MS copy was discovered by Schaaf. See: Schaaf, L J, ‘Sir John
Herschel’s 1839 Royal Society paper on photography’, History of
Photography, 3, 1 (1979), pp47-60.
For ease of future reference, Herschel numbered his ‘articles’
throughout these three published papers: the 1840 paper comprises
§1 to §148; the 1842 paper comprises §149 to §216, and its
postscript §217 to §230; the 1843 paper comprises §231 to §241.

164 The other solution referred to here by Herschel, is the subject of his
preceding articles §202 to §205, viz Smee’s potassium ferricyanide.

165 Herschel, Sir J F W, loc. Cit.
166 The first use of ammonium ferric citrate (coated, by itself, on prepared
paper number 724) is recorded in the Memoranda on 22 July, but the

entry also carries Herschel’s parenthetical remark ‘(omitted before)’;
moreover, a later note on 26 August states that paper 724 was
‘prepared two months ago ±’. It seems likely that Herschel forgot to
allocate a prepared paper number to it at the time.
167 Today we would say ‘reducing’ behaviour
337
168 After the Greek for “dark” kelaino = κελαινο

169 Herschel, op. cit., p214
170 Herschel made strenuous efforts in trying to secure a viable mercury-
printing process, which he had christened ‘kelainotype’ or
‘celaenotype’. At Talbot’s suggestion this was temporarily re-named
‘amphitype’, although that same title was later re-adopted for other
processes. But ultimately Herschel was unsuccessful, and the
mercury process came to nothing, probably because the metal of the
image is sufficiently volatile to evaporate completely within a few
days. The amphitypes that can be identified among the 43 indexed
Herschel specimens at HRHRC are completely blank.
171 Herschel, Sir J F W, ‘On certain improvements on Photographic
Processes described in a former Communication, and on the
Parathermic Rays of the Solar Spectrum’, Philosophical Transactions
of the Royal Society, (1843)
172 Herschel, op. ci.t, p210
173 Nadeau, L, Encyclopedia of Printing, Photographic, and
Photomechanical Processes, (Fredericton: Atelier Luis Nadeau, 1989),
p80.
174 Letter, Smee to Herschel, 29 September 1842. Royal Society Library.
175 Fisher, George Thomas, Photogenic Manipulation, (London: George
Knight and Sons, 1843), p36
176 Thornthwaite, W H, Photogenic Manipulation, (London: Edward Palmer,
1843), p43
177 Hunt, Robert, A Manual of Photography, 5th edn., (London: Richard
Griffin and Company, 1857), p140.
178 “A Practical Chemist and Photographist”, A Practical Manual of
Photography, (London: E. Mackenzie, 1845), p58
179 Bingham, Robert J, Photogenic Manipulation, 9th edn., (London: George
Knight and Sons, 1852), p66
180 Sutton, Thomas, and Dawson, George, A Dictionary of Photography,
(London: Sampson Low, Son & Marston, 1867), p79-80
181 Snelling, Henry H, History and Practice of Photography, (New York:
1845), p
182 Humphrey, S D, A System of Photography, (Albany: C. van Benthuysen,
1849), p121
183 Towler, John, The Silver Sunbeam, (New York: Joseph H. Ladd, 1864),
p273
184 Report on the Proceedings of the Royal Society, The Athenæum, 771,
(6 August 1842), 714; remarkably, this error was perpetuated in
1859: The Photographic News, 2, 37, (20 May, 1859), p126
185 Herschel, J F W, ‘Sir John Herschel ‘on certain photographic effects’,
The Athenæum, 773, (20 August 1842), 748
186 Maunder, M, 'Testing Herschel', The PhotoHistorian, 149, (January,
2007), pp 14-21
187 Maunder, M, 'Herschel's Genius', Ag 41, (Autumn, 2005), p56
188 Mrhar, Peter, Cyanotype: Historical and Alternative Photography,
2013.
338
http://www.petermrhar.com/alternative
see also: http://www.f295.org/main/showthread.php?16962-
Cyanotype-10-times-faster
http://jorj.org/blog/?p=2541
189 Döbereiner, J W, ‘Chemische Thätigkeit des Lichts und Erzeugung des
(neutralen) Humboldtits auf Chemischem Wege’, Schweigger’s
Jahrbuch, 62 (1831), p86, pp90-96.
An abstract of this paper was published in Pharmaceutisches Central
Blatt, 2 (August 1831), pp383-5. It was subsequently translated into
French as ‘Influence chimique de la lumière et formation de la
humboldtite neutre par un moyen photomètrique’ [sic], Journal de
Pharmacie et des Sciences Accessoires, 18, 3 (March, 1832), pp117-
123
190 Herschel, Sir J F W, The Photographic News, 8, 278 (12 February
1864), p82
191 Hunt, R, Researches on Light, (London: Longman, Brown, Green and
Longmans, 1844), p147
192 Most importantly, it formed the basis for Willis’s Platinotype (1873)
and Nicol’s Kallitype (1889) processes.
193 The circumstances of this discovery and its future significance have
been described by Schaaf. The substance is now more correctly
described as calcium hexahydroxyplatinate(IV).
194 Herschel saw a copy of Döbereiner’s paper while he was visiting
Hamburg; it stimulated him to write on 12 June 1832 to Daubeny,
who agreed to read Herschel’s paper on the photosensitivity of
platinum(IV) chloride in solution to the Oxford meeting of the British
Association for the Advancement of Science on 22 June 1832. This
was subsequently published as: Herschel, Sir J F W, ‘On the Action of
Light in determining the Precipitation of Muriate of Platinum by
Lime-water’, London, Edinburgh, and Dublin Philosophical
Magazine, 1 (July, 1832), pp58-60
195 Herschel’s paper, op. cit., was translated and published in German:
Herschel, John F W, ‘Ueber die Wirkung des Lichts als Ursache der
Präcipitation des salzauren Platins durch Kalkwasser’, Annalen der
Pharmacie, 3, 3 (1832), pp337-341.
This transcript carries a long editorial footnote (signed only ‘Br.’)
referring to Döbereiner’s work, part of which can be translated from
the German thus: ‘Following this communication from Döbereiner we
further note that we have repeated Herschel’s experiment and found
it to be confirmed...Thus a compound of platinum oxide with
calcium oxide has been formed through the action of light. We have
not made any analysis of this compound, however, in order not to
anticipate Herr Herschel, and content ourselves by confirming the
facts.’
196 Döbereiner, J W, ‘Chemische Eigenschaften und physische Natur des
auf nassem Wege reducirten Platins’, Annalen der Pharmacie, 14
(1835), pp10-21
339
197 Letter, Herschel to Talbot, 31 March 1843. National Museum of

Photography, Film & Television.
198 This application was finally accomplished in 1898 in a tri-colour
printing method due to Brewerton.
199 The prepared paper numbers in the Memoranda referring to the
standard negative-working cyanotype process are 780, 784, and
793 - three out of a total of twenty-six prepared papers devoted to
cyanotype processes of all kinds.
200 ‘Photographic Society Committee on Positive Printing’, Journal of the
Photographic Society, 2, (21 May 1855), p160
201 Delamotte, P H, Diamond, H W, Hardwich, T F, Malone, T A, Percy, J,
Pollock, H, and Shadbolt, G, Journal of the Photographic Society,
2:36, (21 November 1855), pp251-2
202 Ware, M, Mechanisms of Image Deterioration in Early Photographs,
(London: Science Museum and National Museum of Photography,
Film & Television, 1994)
203 The process is not unlike the tanning of leather, and uses the same
chemical agents - dichromates. The pigmented gelatin is rendered
insoluble where the light falls; the unexposed regions remain
soluble and are washed away with water to develop the image.
204 Ware, M J, ‘Photographic Printing in Colloidal Gold’, Journal of
Photographic Science, 42, 5 (1994), pp157-161
205 Ware, M, ‘The Eighth Metal: The Rise of the Platinotype Process’, in
Photography 1900: The Edinburgh Symposium, (National Museums
of Scotland and National Galleries of Scotland, 1994), pp99-111.
206 Liesegang, P, ‘Positive Printing without Silver’, The Photographic News,
283 (5 February 1864), pp66-7
207 ‘The Late Sir John F W Herschel FRS : his Discoveries in Photography’,
British Journal of Photography, 18, 576 (19 May 1871), pp229-231
208 Marion and Co., Practical Guide to Photography, (London: Marion and
Co., 1887), p215-7
209 Tronquoy, C, ‘Reproductions photographiques par le Papier au Ferro-
Prussiate de M. Marion’, Annales du Génie civil, (Paris: Eugène
Lacroix, 1872). An English translation of this paper was published as
‘Photographic Printing by means of Marion’s Ferro-prussiate Paper’,
The Photographic News, 16, 739 (1 November 1872) pp526-527;
continued in 740 (8 November 1872) p536
210 ‘The Cyanotype Process’,The Photographic News, 21 (22 June 1877),
p289; ibid (24 August 1877), p401
211 Whitaker, C, ‘Apparatus for Printing by the Blue Process’, The
Photographic News, (8 June 1883) pp358-9
212 Marion and Co., Practical Guide to Photography, (London: Marion and
Co., 1885), p283
213 In modern nomenclature: iron(III) hexacyanoferrate(II)
214 When iron (symbol Fe, from the Latin Ferrum) is chemically combined
with other elements, its atoms acquire a positive electric charge by
transferring two or three of their (negatively charged) electrons onto
atoms of other elements. (It is a law of nature that electric charge
340
has to balance. Electrons are the fundamental particles of negative

electricity, and form the outer cloud of every atom, surrounding a
positively-charged nucleus). So each iron atom ends up in one of
two states:
‘ferrous’ iron - also called iron(II), denoted by Fe2+ for short,
or ‘ferric’ iron - also called iron(III), denoted by Fe3+.
The superscripted number, 2+ or 3+, is called the oxidation state of
the iron, and signifies the positive electric charge that the iron atom
has acquired in the reaction. Electrically charged atoms are given the
general name of ions: cations are positive, anions are negative.
215 As an example of the ‘mixed valence’ iron minerals that occur
naturally, one of the best-known is the ‘lodestone’, magnetite,
Fe3O4, which is so intensely coloured that it often appears black.
Special magnetic properties are often a characteristic of mixed-
valence compounds.
216 It is a consequence of the quantum theory of Max Planck, that the
smaller the energy needed to promote such a transition, then the
longer is the wavelength of light absorbed. Prussian blue absorbs
long-wavelength light in the red region of the spectrum (around 700
nm).
217 Bader, A, ‘Out of the blue’, Chemistry in Britain, 33, 11 (November
1997), pp24-6.
http://digicoll.library.wisc.edu/cgi-bin/WI/WI-
idx?type=HTML&rgn=div1&byte=1658105294&q1=Prussian%20blue
Dr Bader offered a £1000 reward for evidence leading to the
identification of the author of his painting, which was awarded to
Dee Cook in 2001 for naming him as Thomas Phillips; see Chemistry
in Britain, 37, 7, (July 2001) p99.
218 Brande, W T, A Manual of Chemistry, (London: John W Parker, 1845,
5th edn), pp754-7. Brande’s famous text was the standard
chemistry reference book for Talbot and Herschel, and was first
published in 1819. Faraday’s copy is in the Wellcome Institute
Library.
219 Ware, Mike, ‘Prussian Blue: Artists’ Pigment and Chemists’ Sponge’,
Journal of Chemical Education, 85, 5, (May 2008), pp612-621
220 After Thomas Everitt (1805-1845), English chemistry teacher and
founder member of the Chemical Society of London. Everitt, T., “On
the Reaction which takes place when Ferrocyanuret of Potassium is
distilled with dilute Sulphuric Acid”, The London and Edinburgh
Philosophical Magazine and Journal of Science, 6, series 3, (February
1835), pp97-103. See: Kraft, Alexander, “On the History of Prussian
Blue: Thomas Everitt (1805-1845) and Everitt’s Salt”, Bulletin for the
History of Chemistry, vol. 3?, no. ??, 20??, pp.??
221 Williamson, Alexander W., “On the Blue Compounds of Cyanogen and
Iron”, Chemical Society Memoranda, 3, (1846), pp125-140
222 Maunder, M, ‘Herschel's Genius and Testing Herschel’, The
PhotoHistorian, 149, (January 2007), pp 14-21:
"What you get is the first fallacy - it's not Prussian blue but Turnbull's blue.
Something I learnt in the 1950s for my A-Levels. Thus, anyone who says you get
341
Prussian blue is wrong. This piece of evidence was more than enough for me to
stick to scientific method and protocol and not use modern historians'
documents."
223 Chadwick, B M, and Sharpe, A G, Advances in Inorganic Chemistry and
Radiochemistry, 8 (1966), p119; Sharpe, A G, The Chemistry of
Cyano Complexes of the Transition Metals, (Academic Press, 1976),
pp121-126
224 Some idea of the difficulties that this topic presented to nineteenth-
century chemists can be obtained from: Williamson, A W, ‘On the
Blue Compounds of Cyanogen and Iron’, Chemical Society Memoirs,
3 (1846), pp125-140; and Williams, H E, Cyanogen Compounds,
(London: Edward Arnold, 1948), pp191-200, 226-228
225 Formula Fe4[Fe(CN)6]3.15H2O.
226 Formula KFe[Fe(CN)6].H2O.
227 Buser, H J, Schwarzenbach, D, Petter, W, and Ludi, A, ‘The Crystal
Structure of Prussian Blue: Fe4[Fe(CN)6]3.xH2O’, Inorganic Chemistry,
16 (1977), p2704 and references cited therein.
228 The etymology of peptise, from the Greek ‘to digest’, is appropriate.
229 The technical definition of a colloid is determined by the size of its
particles, which lie within a range intermediate between molecular
dimensions and the crystals of a normal solid. The usual criterion is
that the particles should be smaller than the wavelengths of visible
light, so that they cannot be discerned under a normal optical
microscope. The name ‘nanoparticles’ is now preferred to colloid
because the latter no longer necessarily implies (as its Greek root
might suggest) that the material is ‘gluey’ - although substances
such as gum and gelatin do indeed constitute another class of
colloid, different in physical properties from colloids such as
Prussian blue.
230 Having the formula NH4Fe[Fe(CN)6].H2O.
231 Holtzman, H, ‘Alkali Resistance of the Iron Blues’, Industrial and
Engineering Chemistry, 37 (1945), p855
232 The cations of caesium and thallium, Cs+ and Tl+, are both singly-
charged, and of similar size to potassium, K+.
233 Pearce, J, ‘Studies of any Toxicological Effects of Prussian-Blue
Compounds in Mammals - A Review’, Food and Chemical
Toxicology, 32, 6 (1994), pp577-582
234 Brewer, K, New Scientist, 138 (1993), p10
235 International Atomic Energy Authority.
http://f40.iaea.org/worldatom/Press/Focus/Chernobyl-
15/caesium.pdf
236 http://www.utexas.edu/courses/sami/dieda/socio/chernobyl.htm
237 BBC News. http://news.bbc.co.uk/1/hi/world/europe/1086547.stm
University of Texas, Sociology Department.
http://www.utexas.edu/courses/sami/dieda/socio/chernobyl.htm
238 Giese, W, ‘The use of ammonium-ferric(III)-cyano-ferrate(II) in pigs
fed with radiocesium contaminated whey powder’, Wiener
Tierarztliche Monatsschrift, 82, 10 (1995), pp310-315
342
239 Center for Disease Control.

http://www.bt.cdc.gov/radiation/prussianblue.asp
240 Emsley, John, Molecules at an Exhibition, Oxford: OUP, 1998, 222-4,
and The Elements of Murder: the History of Poison, 2005. Thallium
poisoning features in the crime novels of Agatha Christie (The Pale
Horse, 1961) and Ngaio Marsh (Final Curtain, 1947).
241 Kravzov, J, Rios, C, Altagracia, M, Monroynoyola, A, and Lopez, F,
‘Relationship between Physicochemical Properties of Prussian Blue
and its Efficacy as Antidote against Thallium Poisoning’, Journal of
Applied Toxicology, 13, 3 (1993), pp213-216
242 Hoffman, R S, Toxicolog. Rev. 22, (2003), pp29-40
243 Mortimer, R J, and Rosseinsky, D R, “Electrochemical Polychromicity in
Iron Hexacyanoferrate Films and a New Film Form of Ferric
Ferricyanide”, Journal of Electroanalytical Chemistry, 151 (1983),
pp133-147; Mortimer, R J, “Electrochromic Materials”, Chemical
Society Reviews, 26, 3, (1997), pp147-156
244 Oxidation is the process of making atoms or molecules more positive
(or less negative) in the electrical sense. So when iron forms
compounds, it is oxidized. Reduction is the converse: making atoms
or molecules less positive (or more negative), e.g. oxygen gas, which
consists of molecules containing two linked atoms of oxygen, O2, is
reduced to form oxides, which contain the O2- anion. Both processes
involve the transfer of electrons. Oxidation is the removal of
electrons, reduction is their addition. When iron(III) is transformed to
iron(II), it is reduced.
245 These organic acids, originally referred to as ‘vegetable acids’ by
Herschel, are all familiar ‘household chemicals’ - even the poisonous
oxalic acid, which occurs (in small amounts) in the leaves of rhubarb,
spinach and sorrel.
246 For more details of the chemistry of ‘precious metal’ printing see
Ware, M J, ‘Noble Metals for Common Images’, in Photochemistry
and Polymeric Systems, Kelly, J M, McArdle, C B, and Maunder, M J de
F, (eds.), Special Publication No. 125 (Cambridge: Royal Society of
Chemistry, 1993), pp250-265
247 For introductory texts on photochemistry at the undergraduate level
see: Cox, A, and Kemp,T J, Introductory Photochemistry, (London:
McGraw-Hill, 1971); Wayne, R P, Photochemistry, (London:
Butterworths, 1970). The comprehensive research manual is: Calvert,
J G, and Pitts, J N Jr, Photochemistry, (New York: John Wiley & Sons
Inc, 1966)
248 For biographical background on the scientists, see: Roger Jones and
Mike Ware, What’s Who?, (Leicester: Troubador Publishing Ltd, 2010)
249 The quantum yield (defined as number of molecules changed divided
by number of photons absorbed) is then said to be equal to one.
Departures from this value significantly greater than one can
apparently arise when the light triggers a chain reaction, but this is
not the case in iron-based photographic processes. The quantum
yield can quite commonly appear to be less than 1.0, however, due
343
to competing reactions which diminish the efficiency of the

photochemical process.
250 A ‘print-out’ material acquires its image by the action of light alone;
there is no chemical amplification of the effect by development with

added chemicals.
251 The ISO (arithmetic) speed is defined as 0.8/exposure in lux-seconds
to give a just-perceptible image. A minimal exposure of one second

to the sun’s illumination (on average, say, 80,000 lux) gives a speed
of about 1/100,000 = 10–5 ISO.
252 Power in Watts is the rate of delivering energy: 1 Watt = 1 Joule per
second.
253 Review of 5th Exhibition of the 'Photographic Society',The Athenæum,
1582, (20 February 1858), p246.

254 Hydrogen cyanide is recognisable by a smell of ‘bitter almonds’. It is
highly toxic.
255 A mixture of ferric and ferricyanide ions is powerful enough to oxidise
the cellulose of paper, forming Prussian blue. Further explanation

for the chemistry can be found in Appendix II.
256 Hunt, R, Researches on Light, (London: Longman, Brown, Green, and
Longmans, 1844), p142;

idem, A Manual of Photography, (London: John Joseph Griffin & Co.,
1853), p52
257 Bird, G, The Elements of Natural Philosophy, (London: John Churchill,
1854), 4th edn., p576

258 Wilson, J L, ‘The Cyanotype’, in Pritchard, M, (ed.) Technology and Art:
the Birth and Early Years of Photography, Proceedings of the Royal

Photographic Society Historical Group conference, September 1989
(Bath: Royal Photographic Society, 1990), p19
259 ‘Employment of Ferricyanide of Potassium for Photographic Purposes’,
The Photographic News, (15 March 1872), p124

260 Sutton, Thomas, and Dawson, George, (eds.), A Dictionary of
Photography, (London: Sampson Low, Son, & Marston, 2nd edn.,

1867), p79
261 Jim Patterson, 'Ferricyanide Carbon'.
https://docs.google.com/document/d/1_zsPv2lWkomdc5T7tg6XVEza_Q
OKFfeFfPordvBbzlY/pub
262 Herschel Memoranda, HRHRC, Austin, Texas.
263 This represents yet another enduring contribution by Herschel to the
vocabulary of photography, joining his better-known usage of the

terms ‘negative’, ‘positive’ and, indeed, the invention of the word
‘photography’ itself. Herschel did not, unfortunately, use the word
‘solarising’ in his publications, so he has not been accorded the
credit for coining it. Prof. John W. Draper also used the word in 1840
to describe the overexposure of daguerreotypes: Philosophical
Magazine, 17, (1840), p217
264 Herschel Memoranda, HRHRC, Austin, Texas
344
266 Prof. Alexander Herschel’s account of his father’s recipe (according to
the “London Photo News”) is cited by Ernst Lietze, Modern
Heliographic Processes, (New York: D. Van Nostrand, 1888), p56
267 This problem is fully discussed in: Ware, M, Mechanisms of Image
Deterioration in Early Photographs, (London: Science Museum and
National Museum of Photography, Film & Television, 1994)
268 Hunt, R, Researches on Light, (London: Longman, Brown, Green, and
Longmans, 1844) pp139-144. The ‘hydrargyrocyanotype’ is one of
the seven cyanotype processes that Hunt lists as ‘the most
interesting of this class’. It is remarkable that he does not list, in its
own right, Herschel’s standard negative-working method (a), except
as a preliminary to this process. The making of negative images may
have seemed less important to him.
269 Mercury(I) nitrate, Hg2(NO3)2.
270 ‘Corrosive sublimate’ is the alchemical name for mercuric chloride, or
mercury(II) chloride, HgCl2; it is an extremely toxic salt: ingestion of
only 0.05 g can be fatal.
271 Herschel, Sir J F W, ‘On Certain Improvements on Photographic
Processes described in a Former Communication and on the
Parathermic Rays of the Solar Spectrum’, Philosophical Transactions
of the Royal Society, (1843)
272 Maunder, M, 'Herschel's Genius', Ag 41, (Autumn 2005), p62, and
'Testing Herschel', The PhotoHistorian, 149 (January, 2007) pp14-21
273 Herschel, Sir John, ‘On Certain Photographic Effects’, Letter dated 10
August 1842 to The Athenæum, 773, (20 August 1842), p748. I am
indebted to Roger Taylor for bringing this letter to my attention.
274 Pellet, H, British Patent no. 4632, (6 December 1877)
275 Pellet, H, French Patent, (23 November, 1877)
276 Henri Pellet, assigned to Clarisse Zoe Joltrain, US Patent no. 241713
(17 May 1881)
277 Eder, J M, ‘Modern Cyanotype Printing’, The Photographic News, 25,
1181 (22 April 1881), p186
278 Pellet, H, US Patent no 241713, (17 May 1881), as quoted in Lietze, E,
Modern Heliographic Processes, (New York: van Nostrand, 1888)
pp65-69
279 Lietze, E, Modern Heliographic Processes, (New York: van Nostrand,
1888), p69
p360
281 Duchochois, P C, Photographic Reproduction Processes, (London:
Hampton Judd, 1892) p33: ‘We have used Pellet’s paper imported
from Paris, which it was impossible to work satisfactorily …’
282 Tritton, F J, ‘Photography in the Drawing Office’, The Photographic
Journal, (September 1943), p329
345
283 See Eder, loc cit; Lietze, loc cit; Duchochois, loc cit; Brown, op cit,
pp53-55
284 Obtained by diluting the concentrated acid 10 times: always adding
100 cc of the acid to 1 litre of water, not vice versa.
285 Waterhouse, Col. J, ‘The Positive Cyanotype Process’, The
Photographic News, 32, 1566 (7 September 1888), pp562-566;
idem., General Report on the Operations of the Survey of India
Department during 1887-1888, (Calcutta: Government Printing
Office, 1889), pp18-21
286 Fisch, A, La Photocopie, ou Procédés de reproductions Industrielles
par la lumière, (Paris: Michelet, 1886)
287 Duchochois, op cit, p29
288 This is a solution of 45° Baumé = SG 1.453 = 50% FeCl3 w/w - see
Appendix IV.
289 I am grateful to John Falconer, Keeper of the Oriental and India Office
Collections of the British Library for providing information
concerning Waterhouse.
290 Price, Lois Alcott, Line, Shade and Shadow: The Fabriction and
Preservation of Architectural Drawings, (Wilmington DE: Oak Knoll
Press, 2010)
291 Valenta, E, Photographisches Correspondenz (1897), p74
292 Eder, J M, Jahrbuch für Photographie, (1898), p448
293 J Merk in Darmstadt, and Dr. Theodor Schuchardt, Chemische Fabrik,
in Goerlitz.
294 It is an oversimplification to suggest that there are just two varieties
of this substance. In fact, ammonium ferric citrate is an ‘ill-
characterised’ compound whose composition depends on the
method of preparation, and it may vary in iron content from 14% to
28%, corresponding to a colour shift from green to brown.
295 The main bibliographic sources of the cyanotype formulations in this
survey were:
Brothers, A, A Manual of Photography, (London: Charles Griffin and
Co, 1899)
Brown, G E, Ferric and Heliographic Processes, (London: Dawbarn
and Ward, 1902)
Burbank, Rev W H, Photographic Printing Methods, (New York: Scovill
and Adams Co, 1891)
Burton, W K, Practical Guide to Photographic and Photo-mechanical
Printing Processes, (London: Marion & Co, 1892)
Clerc, L P, Photography, Theory and Practice, (London: Sir Isaac
Pitman and Sons, 1954)
Duchochois, P C, Photographic Reproduction Processes, (London:
Hampton Judd, 1892)
Eder, J M, Die Lichtpausverfahren, from Ausführliches Handbuch der
Photographie, (Halle: Wilhelm Knapp, 1899)
Jones B E, (ed.), Cassell's Cyclopaedia of Photography, (London:
Cassell and Co, 1911)
Kosar, J, Light Sensitive Systems, (New York: John Wiley and Sons,
346
1965)
Lietze, E, Modern Heliographic Processes, (New York: Van Nostrand,
1888)
Neblette, C B, Photography: its Materials and Processes, (London:
Macmillan and Co, 1952)
Tennant, J A, (ed.), The Photominiature, 1, (10) (London: Dawbarn
and Ward, 1900)
Wall, E J, Dictionary of Photography, (London: Hazell, Watson and
Viney, 1895)
Wall, E J, and Jordan, F I, Photographic Facts and Formulas, (Boston:
American Photographic Publishing Co, 1947)
Ward, H S, Figures, Facts and Formulae of Photography, (London:
Dawbarn and Ward, 1903)
Watts, W A, The Photographic Reference Book, (London: Iliffe and
Son, 1896)
296 A concentration of X % w/v (read as ‘X weight per cent volume’) means
that a weight, X grams (g), of the solute is contained in a volume,

100 cubic centimeters (cc), of the solution. For further details of the
various means of expressing concentration see Appendix IV.
297 Whitaker, C, ‘Apparatus for Printing by the Blue Process’, The
Photographic News, (8 June 1883) pp358-9

298 Hunt, R, ‘Chromo-Cyanotype, a new Photographic Process’,
Philosophical Magazine, (June, 1844), pp435-9

299 Hunt, Robert, A Manual of Photography, (London: Richard Griffin and
Company, 1854) 4th edn.

300 Analytica Chimica Acta, 318, 1 (1995), pp71-76
301 ‘Early Photographs by Robert Hunt’, The Year-Book of Photography
and Photographic News Almanac, (1890), pp193-4

302 Letter, Mercer to Playfair, December 1847. Quoted in:
Parnell, E A, The Life and Labours of John Mercer, (London:

Longmans, Green & Co., 1886), p222
303 Mercer, J, ‘On Chromatic Photographs’, Report of the Meeting of the
British Association for the Advancement of Science, Transactions of

the Sections, (Leeds: 1858), p57. An account of cyanotype on cloth
was, however, pre-published by R. Smith, Journal of the Society of
Arts, (4 February 1853), p121
304 But see, for an account of mordanting by iron: Mercer, John, The
British Journal of Photography, 42 (1895), p557

305 Parnell, E A, The Life and Labours of John Mercer, (London: Longmans,
Green & Co., 1886), pp220-230, 328-337

306 The standard histories of photography make no mention of John
Mercer. His contributions to photography were only brought to

public knowledge recently, by Dr. Stella Butler, curator of the North
West Museum of Science and Industry in a lecture to the Historical
Group of the Royal Society of Chemistry, at Liverpool in June 1987.
307http://archivecat.lancashire.gov.uk/CalmView/record.aspx?src=CalmVi
ew.Catalog&id=UDCL/8/19
347
308 Letter, Burnett to Church, dated 1 September 1855. I am indebted to

Sara Stevenson for bringing this letter to my attention.
309 ‘Ferridcyanide’ is an obsolete alternative for ferricyanide. It had the
merit of amplifying the difference between this, and ferrocyanide.

310 Catalogue of the exhibition of Glasgow Photographic Association,
staged for the British Association meeting in Glasgow, September,

1855. I am indebted to Roger Taylor for producing this catalogue
entry from his database.
311 Burnett, C J, “On the application of Uranium and other matters to
Photography”, Photographic Notes, 23 (1857), pp97-101

312 Sodium nitroprusside is now corrrectly called sodium
nitrosylpentacyanoferrate, Na2Fe(NO)(CN)5
313 “Description of Mr. Burnett’s Pictures in the Photographic Exhibition,
Suffolk Street”, Journal of the Photographic Society, (5 February,

1859), p182
314 West, Redfield B., “Photosensitive Composition”, US patent no.
625,527, 23 May 1899.

315 Maddox, R L, ‘A Cyanotype Process’, The British Journal Photographic
Almanac, (1892) pp544-6

316 Lumiére, A, and Lumiére, L, ‘New Photographic Iron Printing Process’,
The British Journal of Photography, 47, 2121 (1900), p820

317 Valenta, E, ‘Ferro-Prussiate Sensitizers’, The British Journal of
Photography, (9 February 1917), p70

318 Halleur, G.C. Hermann, The Art of Photography, (G.L. Strauss, trans.),
(London: John Weale, 1854), p57

319 Editorial, Popular Science Review, 3 (1864), p131; C. Guillou, “Blue
Photographs”, Photographic Notes, 9, 186, (1 January 1864), p10

320 Traill-Taylor, ‘Iron Printing’, The British Journal Photographic
Almanac, and Photographer’s Daily Companion, (1889), pp357-360

321 Valenta, E, ‘Verwendung von Ferrialkalioxalat als Ersatz für
Ferriammoniumzitrat beim Zyanotypverfahren’, Photographische

Korrespondenz, 728 (May 1921), p107
322 The guanidinium cation has the formula C(NH ) +
2 3
323 Barnes, Robert Bowling, ‘Ferric guanidine oxalate and a process of
making it’, U.S. Patent no. 2,265,934, (9 December 1941).

http://www.freepatentsonline.com/2265934.pdf
324 http://thecarbonworks.com/blog/?p=1286
325 Hnatek, A, ‘Uber die Steigerung der Lichtempfindlichkeit von braunem
Ferriammoniumcitrat durch Beimengung verschiedener organischer

Säuren’, Photographische Korrespondenz, 91, 7 (1955), pp111-5
326 King, Terry, ‘Changing the Blueprint’, British Journal of Photography,
151, 7490, (21 July 2004), pp31-33.

327 “Printing with Salts of Iron”, The British Journal of Photography, 21,
718, (6 February, 1874), pp72

328 Sullivan, Richard, "Ferric Oxalate: its Care and Feeding":
http://www.bostick-sullivan.com/techart.php
305 King, Terry, ‘The Cyanotype Rex: a Hands-On Pictures manual:
http://www.hands-on-pictures.com/Tutorials/Rex.html
348
See also: http://www.apug.org/forums/forum42/25029-another-

theoretical-question-ferrogallate-process.html
330 King, Terry, ‘Changing the Blueprint’, op. cit.
331 Maunder, Michael, ‘Ironing Wrinkles’, Ag 40, (Summer 2005), pp78-83
332 King, Terry, “Chrysotype Rex and Cyanotype Rex: A Revolution in Iron
Based Processes”, View Camera, (November/December, 2005),

pp40-43
http://www.viewcamera.com/pdf/2005/king.pdf
333 Ware, Mike, “Contact versus In-camera Exposure Times”,
http://www.usask.ca/lists/alt-photo-process-l/200609/msg00420.html
334 Maunder, Michael, “Herschel’s genius”, Ag 41, (Autumn 2005), pp54-

63,
335https://www.facebook.com/groups/68142897748/permalink/101514
54186012749/?comment_id=10151454267582749&offset=0&total
_comments=6
336 Brown, George E., Ferric & Heliographic Processes: a Handbook for
Photographers, Draughtsmen and Sun Printers, (London: Dawbarn

and Ward, 1902), pp10-11. Brown attributes this discovery to
Alleyne Reynolds of the Sheffield Photographic Society in 1889.
337 Townsend, N, and Ware, M J, unpublished observations, 1990
338 Miles, George Wellington, “Light-sensitive Composition”, United States
Patent no. 1,518,997, (9 December 1924).

339 Burnett, C J, Photographic Notes, (1857), p99
340 http://unblinkingeye.com/Articles/Uranium/uranium.html
341 http://www.youtube.com/watch?v=AeDK6byP2tk&feature=youtu.be
342 http://periodictable.com/Elements/092/
343 Neblette, C B, Photography: Its Principles and Practice, (London:
Chapman & Hall, 1942), 4th edn., pp657-8

344 Ware, Mike, Platinomicon: The History, Science and Art of
Photographic Printing in Platinum and Palladium, 2013.

345 McCorkle, James, "On the Uranium Toning of Platinum Paper", Camera
Notes, 6, 2, (1902), p90; John A, Tennant, (ed.), "Platinotype

Processes", The Photo-Miniature, 1, 7, (October 1899), pp319-355;
"Platinotype Modifications", The Photo-Miniature, 4, 40, (July 1902),
pp153-193, 605-9; "Platinum Printing", The Photo-Miniature, 10,
115, (May 1911), pp305-342.
346 Benecke, R, ‘Transparencies in Prussian Blue’, The Year-Book of
Photography and Photographic News Almanac, (1890), p194

347 Raess, H F, ‘How to Make Blue Lantern Slides and Transparencies’, The
International Annual of Anthony’s Photographic Bulletin and

American Process Year-Book, 10, (1898), pp175-177
348 Brown, G E, Ferric and Heliographic Processes, (London: Dawbarn and
Ward, 1902), pp20-24

349 http://www.alternativephotography.com/carrier-clear-coating-for-
printing-platinumpaladiumgold-and-pigment-on-glass/
350 http://www.apug.org/forums/archive/index.php/t-93695.html
349
351 Dunn, J.F., & Wakefield, G.L., Exposure Manual, (Watford: Fountain
Press, 1981) p212; Jacobson, R.E., The Manual of Photography,
(London: Focal Press, 1981) p97; James,Theory of the Photographic
Process.
352 Ware, Mike, ‘On Proto-photography and the Shroud of Turin’, History
of Photography, 21, 4, (Winter 1997), pp261-9. See also:

http://www.usask.ca/lists/alt-photo-process-
l/200609/msg00420.html
353 John Beaver, ‘Cyanonegative Photography’ 2002:
http://photo.net/users/godoggo/Cyano/cyano.html
354 http://www.alternativephotography.com/making-cyanotypes-in-
camera/
355http://www.alternativephotography.com/wp/processes/cyanotype/ma
king-a-cyanotype-pinhole-in-camera
356https://www.reddit.com/r/Darkroom/comments/4a000o/im_building_
my_own_enlarger_for_making_cyanotypes/
http://imgur.com/a/ed8cE
357 http://www.glithero.com/blueware-tiles
358 http://www.glithero.com/blueware-vases
359 Muspratt, S, Chemistry, Theoretical, Practical and Analytical, as
Applied and Relating to the Arts and Manufactures, (Glasgow:

William Mackenzie, 1860), pp647-9
360 Stieglitz, Mary, “Historic Photographs on Cloth”, History of
Photography, 1, 2, (April 1977), pp105-110

361 Arnow, J, Handbook of Alternative Photographic Processes, (New York:
Van Nostrand Reinhold, 1982)

362 Wilson, W, private communication, Royal College of Art, 1998.
363 Lucas, A D, ‘Health Hazards Associated with the Cyanotype Printing
Process’, Journal of Environmental Pathology, Toxicology and

Oncology, 11, 1 (1992), pp18-20
364 Amateur Photographer Booklet No. 9. Nashua Pillow Co.,
Manufacturers of Silkdown™ Specialities, Pillows, Cushions, and

Photographic Blue Print Covers. 6-18 Pearsons Avenue, Nashua NH,
(1900).
I am indebted to the late Peter Palmquist, curator of the Women in
Photography International Archive, for supplying me with a
photocopy of this catalogue.
365 I thank Hazel McKenzie and Ann Thomas of the NGC for making this
item available to me.

366 Hewitt, Barbara, Blueprints on Fabric, (Colorado: Interweave Press,
1995)
367 http://www.alternativephotography.com/linda-stemers-cyanotype-
business/
http://www.blueprintsonfabric.com/
368 http://www.stonecreeksilk.co.uk/
369 Brown, Ruth, Cyanotypes on Fabric, (Sunk Island: SC Publications,
2006)
350
370 Stieglitz, Mary, “Photography & Fabric: A Long-Term Relationship”,

Surface Design, (Fall, 2006), pp6-7; Malarcher, Patricia, “Picturing
the World in Blue Cyanotype”, ibid., pp8-11
371 Michel, Cathy Corbishley, ‘The History of Cyanotype (Blueprinting) and
its Use in Photography, Industry, Art and Textiles’, The Journal of the
British Quilt Study Group, 17, (2016), pp44-73
372 Tellkampf, A, and Traube, A, German Patent no 201968, (10 August,
1905)
p453 and references cited therein.
374 Dorel, Felix, US patent no 1,386,995 (filed 1918, awarded 1921)
Photography, (19 July 1918), pp324-5
376 Brown, E, ‘Some Notes on the True-to-scale Gelatine Process’, British
Journal of Photography, 86 (8 December 1939), pp719-721
377 Price, Lois Olcott, Line, Shade and Shadow: the Fabrication and
Preservation of Architectural Drawings, (Wilmington DE: Oak Knoll
Press, 2010), pp211-214
378 Spencer, D A, ‘A New Application of the Ferro-Gelatin Process’,
Seventh International Congress of Photography, pp354-8
379 Farnham, Roger, “Gel Lithography”, in Blueprint: Photography and
Engineering, Roger Farnham and Harry Magee (eds.), (Glasgow: Print
Studio and Street Level Photoworks, 2013)
http://www.blueprint2013.com
380 Schömmer, F, Photographische Rundschau und Mitteilungen, 62,
(1925), p46, as reported in American Photography, (1926), p331. I
am grateful to Richard Farber for bringing this process to my
attention.
381 Kissell, Eléonore, and Vigneau, Erin, Architectural Photoreproductions:
a Manual for Identification and Care, (New York: Oak Knoll Press &
New York Botanical Library, 1999); see also Price, Lois Olcott, Line,
Shade and Shadow: the Fabrication and Preservation of Architectural
Drawings, (Wilmington DE: Oak Knoll Press, 2010)
382 Lundgren, Adrienne and Mintie, Katherine, email communication 22
June 2016. Schindler Borthers, Drawing and Process Papers, 26
Beekman Street, New York
383 Crowley, C A, and Goodyear, G H, US Patent no. 2,350,991 (13 June
1944)
384 Kwech, R E, ‘Commercial Blue-print Papers’, The British Journal of
Photography, 71 (15 August 1924), pp491-3; this is a reprint of an
article first published in American Photography.
385 Willcock, K, British Journal of Photography, 82 (19 July 1935), p458
386 Neblette, op cit, pp697-700
387 Kosar, J, Light Sensitive Systems: Chemistry and Application of Non
Silver Halide Photographic Processes, (New York: John Wiley and
Sons, 1965), pp28-37
351

Photography, (19 July 1918), p324
389 Reichert, J S, US Patent no 1,998,883 (23 April 1935)
390 Kwech, R E, US Patent no 1,877,246 (13 September 1932);
Crowley, C A, and Goodyear, G H, US Patent no 2,093,421 (21
September 1937)
391 Holden, J, British Patent no 418,369 (24 October 1934)
392 Reynolds, A P, US Patent no 2,093,738 (21 September, 1937)
393 Jahoda, E, US Patent no 2,517,111 (1 August 1950)
394 Andrews Reproduction Paper Coating, (New York: Andrews Paper &
Chemical Co. Inc., 1964), p20. I am indebted to Reg Heron for this
reference.
395 Heron, R, ‘Blueprint into Blackprint’, Afterimage, 1, 9 (December
1973), pp4-7, 12
396 I am indebted to Reginald Heron for information on this paper and the
loan of specimen prints.
397 Andrews Reproduction Paper Coating, (New York: Andrews Paper &
Chemical Co. Inc., 1964), pp49, 53
398 Reproduction Bulletin, (Andrews Paper & Chemical Co. Inc., 1972).
399 By the Huey Company of Chicago, and the Universal Blueprint Paper
Company of Denver and Fort Worth.
400 The paper may be obtained from NaturePrint™, P.O.Box 314 Moraga,
California 94556, USA.
401 David Knierim, personal communication.
402 The only camera negatives known to have been made by John
Herschel are views of his father’s 40-foot telescope at Slough.
Specimens of these negatives are in the collections of the Museum of
the History of Science, in Oxford, and the National Museum of
Photography, Film & Television.
403 Talbot, W H F, ‘An account of some recent improvements in
photography’, Proceedings of the Royal Society, 4, 48 (1841),
pp312-16
404 Herschel recorded in his diary on 14 February 1839: ‘...pursued
without intermission the Photographic transfer of engravings.
Prepared and set going a lot sent me by P. Stewt ...’
405 Beck, H, Victorian Engravings, (London: Victoria & Albert Museum,
1973), pp19-23
406 Herschel, Sir J F W, MS Notebooks, vol III, Science Museum Library,
London.
407 I shall use the terms ‘correct’ and ‘reversed’ to denote the chirality
(handedness) of an image with respect to the original object. A
spatially ‘reversed’ photograph depicts the mirror image of the
original, when viewed from the recto. Herschel also used this word,
but rather than ‘correct’ he used ‘direct’ - which I shall avoid
because it might be taken today to imply a different meaning. These
adjectives will be used in conjunction with the nouns ‘positive’ or
‘negative’ to denote the tonality of the image. This should be
carefully distinguished from a third characteristic: the intrinsic
352
response of the process itself, which may be inherently either

positive-working or negative-working.
408 This he first did on 31 January 1839, and showed the result to Talbot
on 1 February 1839.
409 Thus, the result of making a negative of a negative is a positive, just
as -1 x -1 = +1; while a negative of a positive is still a negative:
-1 x +1 = -1, and a positive of a positive is a positive:
+1 x +1 = +1. The analogy could also be argued that the ‘sum’
of a negative and a positive is blank, i.e. zero: -1+1 = 0.
410 Schaaf, L, ‘Sir John Herschel’s 1839 Royal Society Paper on
Photography’, History of Photography, 3, 1 (January 1979), pp47-60
411 Herschel, Sir J F W, ‘On the Chemical Action of the Rays of the Solar
Spectrum on Preparations of Silver and other Substances, both
metallic and non-metallic, and on some Photographic Processes’,
Philosophical Transactions of the Royal Society, (1840) pp3, 7-8
412 The concepts of optical densitometry and the characteristic curve of a
photosensitive layer lay several decades into the future.
413 There is just one engraving diaphane in the NMPFT collection, and it
has not been completely cropped, so the partly excised title can be
matched to one of Herschel’s titles: Rosolia. There are frequent
references to this title in the Memoranda; one of Herschel’s positive
cyanotypes of this image has been published by Schaaf.
414 Art Libraries Association Portrait Index, p1377
415 Pringle, T, (ed.), Friendship’s Offering; and Winter’s Wreath: A
Christmas and New Year’s Present for MCCCCXXXVI, (London: Smith,
Elder & Co., 1836), Plate III, facing p36
416 Renier, A, Friendship’s Offering: An Essay on the Annuals and Gift
Books of the 19th Century, (London: Private Libraries Association,
1964)
417 British Museum Catalogue of Engraved British Portraits, p553
418 Pringle, T, (ed.), Friendship’s Offering; and Winter’s Wreath: A
Christmas and New Year’s Present for MCCCCXXXVIII, (London:
Smith, Elder & Co., 1838), Plate IV, facing p73
419 A postcard reproduction of a negative cyanotype made by Herschel
from this engraving is obtainable from the Museum of the History of
Science at Oxford.
420 Many of the library holdings of Friendship’s Offering are incomplete.
The National Arts Library, the British Library, the Bodleian Library,
Oxford, and Birmingham Public Library have almost complete runs.
421 In the archive at the Museum of the History of Science, at Oxford, on a
packet labelled ‘1 Cyanotype, 2 Amphitypes’ there is a note in
Herschel’s hand which reads: ‘Dr. Sir, when done with...sent to care
of Mr Stewart, 65 Cornhill. They will come safe to hand.’
422 Simcock, A V, Photography 150: Images from the First Generation,
(Oxford: Museum of the History of Science, 1989), p22
423 Britton, J, Modern Athens! Displayed in a Series of Views: or Edinburgh
in the Nineteenth Century, from original drawings by Thomas H.
353
Shepherd, (London: Jones & Co., 1831), reissued (New York:

Benjamin Blom, Inc., 1969)
424 Simcock, A V, ‘An ark for the history of science’, International
Association of Technological University Libraries Quarterly, 1, 3,
(1987), pp196-215
425 http://norman.hrc.utexas.edu/photoPublic/fullDisplay.cfm?CollID=73
426 Schultze, R S, ‘Re-discovery and Description of Original Material on
the Photographic Researches of Sir John F. W. Herschel, 1839-1844’,
Journal of Photographic Science, 13 (1965), pp57-68.
427 Simcock, A V, Photography 150: Images from the First Generation,
(Oxford: Museum of the History of Science, 1989)
428 For a transcription of this MS see: Helmut Gernsheim, Julia Margaret
Cameron, (London: The Fountain Press, 1948), p71, Plates 4 and 51
429 Schultze, loc. cit.
430 The validity of these arguments rests on two assumptions, which are
justified by the generally sharp quality of the resulting images:
namely, that Herschel did not make more than two successive
derivative contact prints from an engraving, and that he did not print
through the verso of an engraving to obtain an image of the same
chirality in one step.
431 Cat. no. 1943-39/72
432 Schaaf, L J, Out of the Shadows: Herschel, Talbot & the Invention of
Photography, (New Haven: Yale University Press, 1992), fig 82, p128.
433 The original engraving diaphane of Rosolia is also in the NMPFT. It
comes from Friendship’s Offering for 1837, Plate X, p289.
434 These four sheets of specimens are now in the Harry Ransom
Humanities Research Centre, at Austin, Texas.
435 http://norman.hrc.utexas.edu/photoPublic/fullDisplay.cfm?CollID=73
436 The ‘coating weight’ is defined as the weight of a specified component
per unit area of the surface.
437 Which corresponds typically to a volume of 1.4 cc applied to an A4
sheet.
438 Schaaf, L J, Sun Gardens. Victorian Photograms by Anna Atkins, (New
York: Aperture, 1985)
439 Children, J G, ‘Lamarck’s Genera of Shells’, The Quarterly Journal of
Science, Literature and the Arts, 16, 31 (1823)
440 Schaaf, op. cit., p24
441 ‘Fac-simile of a Photogenic Drawing’, The Mirror of Literature,
Amusement, and Instruction, 33, 945 (20 April 1839), p241
442 Hervey, W H, A Manual of British Algae, (London: John van Voorst,
1841)
443 A. A. [Atkins, Anna], Photographs of British Algae. Cyanotype
Impressions., (Halstead Place, Sevenoaks: privately printed, 1843-
53)
444 Schaaf, Larry J, Sun Gardens: Victorian Photograms by Anna Atkins
(New York: Aperture, 1985); ibid., exhibition organised by the
Crawford Centre for the Arts of the University of St. Andrews, 1988.
354
445 Hunt, R, ‘On the Application of Science to the Fine and Useful Arts.
Photography - second part.’ The Art-Union Monthly Journal, 10, (1
August 1848), pp237-8
446 Talbot, H F, ‘On Photography without the Use of Silver’, The British
Journal of Photography, 9, (9 December 1864), pp495-6

447 Hunt's and Lang's copy passed through Christie's sale rooms in 2004:
http://www.christies.com/LotFinder/lot_details.aspx?intObjectID=42785
18
448 Lang, W, ‘The Cyanotype Reproduction of Sea-Weeds’, The
Photographic News, 34, 1683 (5 December 1890), p948

449 Lang, W, 'The Cyanotype Reproduction of Seaweeds', Proceedings of
the Philosophical Society of Glasgow, 21, (1889-90), p155; idem.,

Photographic News, 34, (5 December 1890), p948; idem., British
Journal of Photography, (25 October 1889), p702.
450http://digitalgallery.nypl.org/nypldigital/dgtitle_tree.cfm?title_id=100
174&level=1&parent_id=288165&word=&s=&notword=&d=&c=&f=
&k=0&sScope=&sLevel=&sLabel=&lword=&lfield=&imgs=20&snum
=0
451 The Anna Atkins Volumes, Auction Catalogue on CD-ROM (London:
Bonhams of Knightsbridge, 1996)

452 The Guardian, (Wednesday 19 June 1996), p5; The Observer, (Sunday
23 June 1996), p20-23

453 As with most claims to be ‘first’, this can be disputed - depending on
the precise interpretation that is placed on the expression

‘photographically illustrated book’. The debate does not detract one
scintilla from Anna Atkins’s achievement, however.
454 Anne Dixon (1799-1864) was born Anne Austen (a second cousin to
the novelist); she married Henry Dixon, the vicar of Ferring, Sussex,
in 1837; the couples spent much time together, including botanising
on the Sussex seashores.
455 Atkins, A, Memoir of John George Children, Esq., Including Some
Unpublished Poetry by His father and Himself, (London: John Bowyer

Nichols & Sons, 1853)
456 Wilson, J L, Some aspects of the photographically illustrated book, a
lecture delivered to the Historical Group of the Royal Photographic

Society, (20 January 1988).
457 Wilson, J L, ‘The Cyanotype’, in Pritchard, M, (ed.), Technology and Art:
the Birth and Early Years of Photography, (Bath: Royal Photographic

Society Historical Group, 1990), pp19-25
458 Johnstone, W G, and Croall, A, The Nature-Printed British Seaweeds: a
History, accompanied by Figures and Dissections of the Algae of the

British Isles, Nature-Printed by Henry Bradbury. (London: Bradbury
and Evans, 1859), vol I, p77; ibid vol II, p143
459 The Athenaeum, 1657, (30 July 1859), pp149-150
460 The Athenaeum Review Database, III, 1713, (25 August 1860), 293;
ibid., IV, 1714, (1 September 1860), 262

461 This album, lent from the collection of Hans Kraus Jnr., was displayed
in the exhibition ‘Photography in Science: Beauty of Another Order’
355
at the National Gallery of Canada, Ottawa, in 1997. I am indebted to

Hans Kraus for kindly affording me the opportunity to examine it.
462 The Photographic Journal, 137, 7 (September 1997), p281. See also
ibid. 135, 10 (November 1995), pp418-420
463 Jennifer Howard, A Victorian woman’s experiments in botanical
illustration: A historical bibliographical approach to Anna Atkins’
Photographs of British Algae: Cyanotype Impressions, MA Thesis,
University College of London, 2011.
464 Schaaf, L J, Sun Gardens; an Exhibition of Victorian Photograms by
Anna Atkins, Crawford Centre for the Arts, The University of St.
Andrews, 1988
465 For the ‘Meteorological Observations’ for these months, see, for
example, The Philosophical Magazine, 22, 146 and 148, (May and
June 1843), pp415 and 503 respectively.
466 Dusan Stulik, Art Kaplan, Susie Clark, Philippa Wright and Brian Liddy,
private communication, National Museum of Photography, Film and
Television, August, 2006.
467 The handwriting is also not that of two other possible candidates for
authorship, Michael Pakenham Edgeworth and Hugh Cleghorn
468 The question is further complicated by the fact that some of the
author’s pencilled annotations refer to ‘modifications of the Herschel
formula’, labelled by the author as A, B, and C. There is no obvious
difference between the resulting prints so labelled, and despite the
coincidence, this notation cannot be identified with Herschel’s own
‘varieties (a), (b), and (c)’, because (b) and (c) were positive-working
processes, as has been deduced earlier, but all the cyanotypes in
Dried Ferns were obviously made by a negative-working process. It
seems likely that these ‘modifications’ A, B, and C, were only minor
variants in the processing, because they have not given rise to any
sensible difference in the final images.
469 The neologism ‘autobotanography’ is properly the creation of novelist
Jocelyn Brooke (1908-1966) in 1949 to describe his work:
http://trove.nla.gov.au/ndp/del/article/27577344
http://litlinks1.blogspot.co.uk/2013/08/jocelyn-brooke.html
470 Armstrong, Carol, and de Zegher, Catherine, (eds.), Ocean Flowers:
Impressions from Nature, (New York: The Drawing Center &
Princeton University Press, 2004)
471 I am indebted to John Sullivan, curator of the photographic archive of
the National Library of New Zealand, for this information. See also:
Sampson, F B, Early New Zealand Botanical Art, (Reed Methuen,
1985), pp109-110
472 DiNoto, Andrea, and Winter, David, The Pressed Plant: The Art of
Botanical Specimens, Nature Prints, and Sun Pictures, (New York:
Stewart, Tabori & Chang, 1999), pp126-7
473 Hamilton, A, Four Flowers: Cyanotypes, Catalogue for a touring
exhibition, (Edinburgh: Fotofeis Ltd., 1995)
356
474 [Julia Herschel], A handbook for Greek and Roman lace making,
London: R. Barrett & Sons, 1869. A copy in the NYPL may be viewed
online:
https://digitalcollections.nypl.org/items/7dd58e00-0898-0133-
038f-58d385a7bbd0
475 [Julia Herschel], A handbook of Greek lace making, London: R. Barrett
& Sons, 1870. A copy in the JPGM may be downloaded from:
https://archive.org/details/handbookofgreekl00hers
476 Janis, E P, and Sartre, J, Henri Le Secq: Photograph de 1850 à 1860,
(Paris: Musée des Arts Décoratifs/Flammarion, 1986)
477 Jammes, A, and Janis, E P, The Art of the French Calotype, (Princeton,
NJ: Princeton University Press), pp206-210
478 Le Secq’s negatives were made exclusively by the waxed paper
process.
479 Janis, E P and Sartre, J, op. cit., p39
480 Musée d'Orsay, exhibition 1996-7: Paul B. Haviland (1880-1950): a
photographer:
http://www.musee-orsay.fr/en/events/exhibitions/in-the-musee-
dorsay/exhibitions-in-the-musee-dorsay/article/paul-b-haviland-
1880-1950-a-photographer-4449.html?cHash=4edb944c85
481 Heilbrun, F, and Bajac Q, Paul Burty Haviland (1880-1950),
photographe, (Paris: Editions de la Réunion des musées nationaux,
1996), p84. I am indebted to Roger Wilson and Pamela Roberts for
first bringing this work to my attention.
482 Heilbrun, F, and Bajac Q, op cit., p66
483 I thank the owner of the album, John Moran and his family, for the
kind provision of information on this work.
484 Fairbrother, Trevor, Ipswich Days: Arthur Wesley Dow and his
hometown, (Andover: Addison Gallery of American Art, 2007).
485 http://www.mhs.ox.ac.uk/collections/imu-search-
page/results/?query=Teasdale&thumbnails=on&querytype=basic
http://www.mhs.ox.ac.uk/cameras/index.htm?item75
486 http://www.mhs.ox.ac.uk/collections/search/display-
narrative/?irn=14216&index=1
487 http://www.luminous-
lint.com/app/photographer/Washington__Teasdale/A/
488 Jacobson, K, Études d’après Nature, (Petches Bridge: Ken & Jenny
Jacobson, 1996), pp135, 162
489 Many of the Rijksmuseum’s 611 cyanotypes may be viewed online:
https://www.rijksmuseum.nl/en/search?p=1&ps=12&technique=cyanot
ypie&ii=3
I thank Photograph Conservator, Martin Jürgens for this information.
490https://library.usu.edu/Specol/photoarchive/cyanotypes/LARGE/P001
4.html
491 https://library.usu.edu/Specol/photoarchive/cyanotypes/about.html
492 Braun, Marta, 'Muybridge's Animal Locomotion: The Director's Cut',
History of Photography, 24, 1, (Spring 2000), pp52-54
357
493http://www.flickr.com/photos/publicresourceorg/sets/721576002155
29969/with/493890432/
494 Information posted to the PhotoHistory List of the Internet by Debbie
Griggs Carter. One such strip was exhibited at the Cinema’s
Centennial Exhibition in 1995.
495 Bartholomew, E, Curator, Photographic & Film Collection, National
Railway Museum, York, a private communication.

496 Ware, Mike, Annual Annan Lecture, 2013, of The Scottish Society for
the History of Photography, “Cyanotype: a Blueprint for Visual

Vandalism?”, Studies in Photography, (2013), pp24-44
497 Stevenson, S, Light from the Dark Room, (Edinburgh: National
Galleries of Scotland, 1995), p113

498 Album, Forth Bridge, Canadian Centre for Architecture, Montreal:
http://www.cca.qc.ca/en/collection/504-forth-bridge-scotland
see also: Phyllis Lambert, Casabella, October 2000:
http://www.cca.qc.ca/en/collection/660-queensferry-cantilever-as-
seen-from-the-south-approach
499 Gray, Michael, and Maggi, Angelo, Evelyn George Carey: Forth Bridge
(Milan: Federico Motta Editore, 2009).

500 Album of cyanotypes, Percement du Canal de Panama, (1888-
ca.1903), (Canberra:National Library of Australia,)

http://catalogue.nla.gov.au/Record/3044425
501 I am indebted to Philip Jackson of the National Library of Australia for
this information.
502 Michael Conner, ‘Henry Bosse and Samuel Clemens as Mark Twain –
Parallel Lives on the Mighty Mississippi’, Ramsey County History, 33,

3, (Fall 1998), pp18-25
503 Wehrenberg, Charles, Mississippi Blue - Henry P. Bosse and his Views
on the Mississippi River Between Minneapolis and St. Louis, 1883-

1891, (Twin Palms Publishers 2002)
504 http://www.mvp.usace.army.mil/About/HenryPeterBossePhotos.aspx
505 http://reflections.mndigital.org/cdm/landingpage/collection/army/
506 Neuzil, Mark, Views on the Mississippi: The Photographs of Henry
Peter Bosse (Minneapolis: University of Minnesota press, 2001).

507 Earl Johnson, e-mail communication to the author, February 2014. I
thank Earl Johnson for information regarding Bosse.

508 Neuzil, 2001, p.4. Ref 496.
509 https://www.google.com/culturalinstitute/asset-viewer/u-s-dredge-
phoenix/AAHc4n1oP5z0-A?hl=en
510 How to Paint: an Instruction Book, (St. Louis: A.S. Aloe Company,
1894), pp249-251, 352-3

https://archive.org/details/howtopaintinstru00asal
511 Pullis, Pierre, cyanotypes of New York City Subway construction 1901-
09:
http://catalogue.swanngalleries.com/asp/fullCatalogue.asp?salelot=229
8++++++69+&refno=++659886&saletype
http://tinyurl.com/q6ps8md
358
512 http://www.luminous-
lint.com/app/home/?action=ACT_VEX&p1=_PROCESS_Cyanotype_01
&p2=1&p3=0&p4=0
513 Goldblatt, R, ‘The Mafeking Blues’, The South African Philatelist,
(March 1978), pp62-79
514 The account is slightly obscured by the common error of transposing
ferrocyanide for ferricyanide.
515 Bensusan, A D, Silver Images: History of Photography in Africa, (Cape
Town: H Timmons, 1966)
516 The Dictionary of National Biography has his forename as “Edwin”.
517 I thank Julia Findlater, Curator of Leighton House, for showing me this
material.
518 Roberts, M A, ‘Edward Linley Sambourne (1844-1910)’, History of
Photography, 17, 2 (1993), pp207-213; Rimmer, R W, ‘A Note on
Linley Sambourne (1844-1910)’, The Photohistorian, 110 (January
1996), pp13-15; Gent, M, The Cyanotype, Materials and Techniques
Essay, Royal College of Art MA course in conservation, 1992
519 Simon, Robin, (ed.), Public Artist, Private Passions, Exhibition
catalogue, (London: Kensington and Chelsea Libraries and Arts
Service, 2001); Grant, Lucille, “Salacious and Satirical”, Journal of the
Royal Photographic Society, (October 2001), pp376-7
520 https://www.facebook.com/groups/13488949283/
521 http://www.flickr.com/photos/tags/cyanotype/
http://www.flickr.com/groups/61907246@N00/
522 http://www.alternativephotography.com/category/gallery-by-
process/cyanotype-photographers/
523 http://www.alternativephotography.com/cyanotypes-a-new-look-at-
an-old-technique-thesis/
524 http://www.angela-easterling.co.uk/angela-
easterling.co.uk/blueprint_for_living.html
http://www.angela-easterling.co.uk/angela-easterling.co.uk/Workshops.html
525 http://www.nickveasey.blogspot.co.uk/2012/05/from-rocks-and-reefs.html
526 http://www.dcfineart.net/
527 Farnham, Roger, and Magee, Harry, (eds.), Blueprint : photography
and engineering, (Glasgow Print Studio, 2013)
http://www.blueprint2013.com
528 Fabbri, Malin, Alternative Photography: Art and Artists, Edition I,
(Stockholm: AlternativePhotography.com, 2006); Fabbri, Malin, and
Fabbri, Gary, Blueprint to Cyanotypes: Exploring a historical
alternative photographic process, (Stockholm:
AlternativePhotography.com, 2004)
529 http://www.photographydealers.com/artists/john-dugdale/
http://www.holdenluntz.com/artists/john-dugdale
530 http://www.schaeferphoto.com/
531 http://www.ciurejlochmanphoto.com/naturalhistory/nathist01.html
532 http://alternativeimpressions.wordpress.com/category/cyanotypes/
http://lightimpressions.blogspot.co.uk/search/label/Cyanotypes?zx=521ae4
7e508555a1
359
533 Wood, John, “The Art of the Cyanotype and the Vandalous Dreams of
John Metoyer”, in The Photographic Arts, (Iowa: University of Iowa
Press, 1997)
534 http://www.brentonhamiltonstudio.net/main.html
535 http://peternappi.squarespace.com/cyanotypes/
536 http://www.museumofmemory.com/
537 Burns, Nancy Kathryn, and Wilson, Kristina (Eds.), Cyanotypes:
Photography's Blue Period, Worcester MA: Worcester Art Museum,

2016.
538 http://www.worcesterart.org/exhibitions/cyanotypes-photographys-
blue-period/
539 Crawford, W, The Keepers of Light, (New York: Morgan & Morgan,
1979); Farber, R, Historic Photographic Processes: A Guide to
Creating Handmade Photographic Images, (Allworth Press, 1998);
James, Christopher, Alternative Photographic Processes, (New York:
Delmar Cengage Learning, 2009); Enfield, Jill, Jill Enfield’s Guide to
Photographic Alternative Processes, (New York: Focal Press, 2013;
Anderson, Christina Z., Gum printing and other amazing Contact
Printing Processes, (2013).
540 An exception to this was the solar microscope, which harnessed the
sun’s rays to project an enlarged image with sufficient intensity to
be printable on the very insensitive papers of the time. A primitive
enlarger employing the sun’s light was also described by John
Stewart, the brother-in-law of Sir John Herschel, in the Journal of the
Photographic Society, 1, 8 (22 August 1853), pp101-3.
541 Herschel is generally acknowledged as the author of the word
photograph: he first used it in his diary on 13 February 1839; on 28

February 1839 he commended it to Talbot, and on 14 March he
made the first use of it publicly. Many of his initial specimens are
annotated ‘J.F.W.H. Photogr.’, the earliest of these being dated 10
February, 1839. Simcock has pointed out that it is not clear exactly
what Herschel intended by this abbreviation, i.e., ‘Photograph’,
‘Photographer’ or ‘Photographsit’, analogous to the ‘pinxit’ of
painters. The word ‘photograph’ first appeared in print on 25
February, 1839.
542 Batchen, G, ‘Photogenics’, History of Photography, 22, 1 (1998),
pp18-26
543 Talbot originally named such images skiagraphs, which translates as
‘shadow-writings’, but he later changed the name to photogenic

drawings.
544 A cyanotype made from this anatomical specimen in the 1960s was
exhibited at the National Gallery of Canada, Ottawa, in 1997 in

Photography in Science: Beauty of Another Order .
545 The paucity of our language in this respect is highlighted by the fact
that, just to describe the simple concept of a shadow-cast image, we

have had to enlist the surname of a Frenchman, M. Étienne de
Silhouette (1709-1767). Duchochois (op cit, p22) offers us another
French word, ‘Cliché’, explaining in a footnote, ‘Cliché is the term
360
applied by photographers to the negatives and diapositives, and by

extension to any drawing or print on transparent materials,
employed in the photo-mechanical and other photographic printing
methods.’ The OED definition of cliché as ‘a metal stereotype of a
wood engraving used to print from’ implies that this is a rather
inappropriate usage. However the word was also used by
Waterhouse to describe Schulze’s cut-out letters, as if it were then
current parlance. See The Photographic Journal, (June 30 1903),
p166.
546 The Shorter Oxford English Dictionary: ‘Diaphane. 1. A transparent
body or substance; a transparency 1840. 2. A silk stuff having
transparent coloured figures 1824.’
The well-known adjective ‘diaphanous’ derives from it and was used
by Herschel in his 1839 paper, but it has now degenerated to a
rather hackneyed stock-in-trade of the romantic novelist!
547 It may be objected that diaphane is inappropriate for totally opaque
objects - in which case skiaphane or scotophane might be
considered.
548 Jones, B E, (ed.) Cassell’s Cyclopaedia of Photography, (London:
Cassell and Company, 1911), p181
549 The British Journal of Photography, (15 September 1863), pp371-2.
550 Elegant Arts for Ladies, (London: Ward and Lock, nd, 1861?), pp154-5.
Among the several techniques of ‘diaphanie’, one describes the
application of engravings to glass with ‘washable varnish’ followed
by treatment with a ‘clearing liquid’. The special solutions for this
purpose were available commercially from Jabez Barnard, an artists’
colourman of Oxford Street. I am grateful to Roger Taylor for
discovering this reference.
551 The word ‘skiagraphy’ was briefly used by W.H.F. Talbot in 1835 for
what he soon preferred to call his ‘photogenic drawings’; Greek:
skia-grafein = ‘shadow-drawing’. See: Talbot, W.H.F., Notebook M,
entry for 28th February, 1835. Lacock Abbey Collection, Fox Talbot
Museum, National Trust.
552 Kemp, Martin, The Science of Art, New Haven and London: Yale
University Press, 1990.
553 Puckett, R. Campbell, Sciography or Radial Projection of Shadows,
London: Chapman & Hall, 1871; Fred Dubery and John Willats,
Perspective and other drawing systems, London: The Herbert Press,
1983, see chapter 9 on ‘Sciography’.
554 Ware, Mike, 'Prolegomena to a Skiology of Photogenic Drawing',
History of Photography, 28, 3, (Autumn 2004), pp237-246; idem.,
‘On Proto-photography and the Shroud of Turin’, History of
Photography, 21, 4, (Winter 1997), pp261-9.
Unfortunately the word ‘skiology’ has now been widely
reappropriated by the devotees of the piste, so I have had to replace
it with ‘skiagraphy’ in the present text. However, even this usage is
somewhat ambiguous, since this word was also used to refer to early
images obtained by X-rays.
361
555 Vision under high illuminance, which is photopic rather than scotopic.
556 An angle is defined in radians as the ratio of the length of the
subtended arc of a circle to its radius. Since the circumference of a
circle is, by definition, 2p times its radius, it follows that the
conversion between the two systems of angular measurement is
given by: 360 degrees º 2p radians, whence 1 radian º 180/p =
57.296° (degrees of arc). When angles are small (in the order of 25°
or less), the length of the short arc of the circle’s circumference is
closely approximated (with a precision better than 1%) by the
(slightly shorter) straight line spanning the same arc (called in
geometry, the ‘chord’), and we shall make frequent use of this
approximation to determine small angular measures easily.
557 Pirenne, M H, Optics, Painting & Photography, (Cambridge: Cambridge
University Press, 1970). The light sensitive receptors in the most
sensitive region of the retina, the fovea, are the cones, packed
together and each subtending an angle of 0.5 minutes of arc at the
pupil – i.e. the critical limit of angular resolution by the eye is twice
this: about 1 minute of arc or ca. 1/3420 radian.
558 The reader can easily test these figures personally, by viewing the
millimeter scale of a ruler (preferably one of white plastic with black
calibrations) in a good light, at distances measured with a tape, to
determine the distance (in mm) at which the mm calibrations just
become discernable; the reciprocal of this distance gives the angular
resolution of vision in radians.
559 Jacobson, Ralph E, (ed.), The Manual of Photography, (London: Focal
Press, 7th edition, 1978).
560 Or 1/100th of an inch at a distance of 10 inches.
561 The sun's diameter is about 864,000 miles (1,391,000 kilometers).
The Earth orbits the sun at an average distance of about 92,960,000
miles (149,600,000 kilometers).
The angular diameter is therefore 864,000/92,960,000 = 1/108 rad
or 0.533 degrees. Data from: Weast, R C, (ed.), CRC Handbook of
Chemistry and Physics, (Boca Raton: CRC Press, 1980).
562 Lynch, David K, and Livingston, William, Color and Light in Nature,
(Cambridge: Cambridge University Press, 1995).
563 The umbra and penumbra are most conspicuous at the time of a solar
eclipse.
564 For details of this calculation see Ware, Mike, ‘Prolegomena…’, loc.cit.
565 Likewise, any diffusion of the sunlight by clouds would also have
blurred a projected image by increasing the effective size of the light
source.
566 The new cyanotype process used for these photograms requires
shorter exposures: one minute suffices in the Aegean sun.
567 Is there a case for coining a neologism here – a tenebrogram?
568 See also: DiNoto, Andrea, and Winter, David, The Pressed Plant, (New
York: Stewart, Tabori & Chang, 1999).
569 These conditions are often agreeably met with in a Mediterranean
climate!
362
570
http://www.mikeware.co.uk/mikeware/HellenImp.html
571 Roberts, J C, The Chemistry of Paper, (Cambridge: The Royal Society of
Chemistry, 1996), pp109-110.
572 Sirdar Bilici, "Sulfamic acid for paper acidification": 2013
http://www.apug.org/forums/forum42/118153-sulfamic-acid-
paper-acidification.html
573 Ware, M, ‘A Handmade Paper for Alternative Processes’, Ag+
Photographic, 7, (1995), pp34-9

574 http://www.ruscombepaper.com/index1.html
575 UVB lies below 320 nm, at which point ordinary glass begins to absorb
strongly, and UVC lies below 280 nm.

576 http://www.pictorico.eu/productsshop/?no_cache=1&L=2
http://www.permajet.com/ProductGrp/PermaJet-Digital-Transfer-
Negative-Film-165u
577 Fabbri, Malin, and Fabbri, Gary, Blueprint to cyanotypes: exploring a
historical alternative photographic process,

578 Ware, M, ‘A New Blueprint for Cyanotypes’, Ag+ Photographic, 7
(1995), pp74-80; Barnier, J, ‘The New Cyanotype’, Photo Techniques,

18, 1 (Jan/Feb 1997), pp12-15
579 http://www.bostick-sullivan.com/techart.php
580 Tennant, J A, ‘The Blue Print Process’, The Photo-miniature, 8, 81
(1906), pp432-7
581 One well-respected practitioner complained of having to reject three
bottles of the chemical before finding a satisfactory one: Judy Seigel,

World Journal of Post-Factory Photography.
582 i.e. ca. 10% w/v potassiumn ferricyanide, corresponding to 0.3 molar
583 The solubility of potassium ferric oxalate is 6.6 g per 100 cc of water
at 15 °C.
584 If ammonium ferricyanide were available, making up the sensitizer
would be very easy: dissolve 8 g of it in ca. 70 cc of water; add 17 g

of ammonium ferric oxalate, dissolve it, and make up to 100 cc.
585 Ware, M J, ‘A New Blueprint for Cyanotypes’, Ag+ Photographic, 7
(1995), pp74-80; Barnier, J, ‘The New Cyanotype’, Photo Techniques,

18, 1 (1997), pp12-15
586 called a “graduate” in the USA
587 The chemical explanation is given in Appendix III, where it is shown
why, for ferrioxalate sensitizers, acid conditions greatly promote the

formation of Prussian blue after exposure.
588 Tim Summa, Alternativephotography.com forum, 30 July 2009
589 Mike Ware, Cyanotype: the history, science and art of photographic
printing in Prussian blue, London: Science Museum and National

Museum of Photography, Film and Television, 1999.
http://www.mikeware.co.uk/downloads/Cyanomicon.pdf
590 Nancy Kathryn Burns and Kristina Wilson, Cyanotypes: Photography's
Blue Period, Worcester MA: Worcester Art Museum, 2016, 64-7.

591 John A. Tennant, "Platinotype Processes", The Photo-Miniature, 1:7,
October 1899, 344, 349.
363
592 The Materials Safety Data Sheets can be found on the World Wide Web
at http://www.enviro-net.com/technical/msds/
593 Lucas, A D, ‘Health Hazards Associated with the Cyanotype Printing
Process’, Journal of Environmental Pathology, Toxicology and
Oncology, 11, 1 (1992), pp18-20
594 Dating from 1901, when Andrew Greig Barr of Glasgow adopted
ammonium ferric citrate as the essential active ingredient for
Scotland's other national drink: Irn-Bru. This beverage contains
0.002% ammonium ferric citrate - regrettably too dilute for making
cyanotypes!
595 http://www.nontoxicprint.com/safetyinphotography.htm
596 http://www.wrc.org.za/Knowledge%20Hub%20Documents/Water%20S
A%20Journals/Manuscripts/1999/03/WaterSA_1999_03_jul99_p363.
pdf
597 http://infohouse.p2ric.org/ref/19/18012.pdf
598 http://www.must.edu.ph/mjst/wp-content/uploads/2014/02/1.-
Hazardous-Waste-Chemicals-from-Dichromate-Chemical.pdf
599 http://inpressco.com/wp-content/uploads/2013/03/Paper524-
27.pdf
600 https://www.researchgate.net/profile/Ji-
Dong_Gu/publication/8227753_Reduction_of_hexavalent_chromium
_by_ascorbic_acid_in_aqueous_solution_Chemosphere_57609/links/
548da68b0cf225bf66a5f470.pdf?inViewer=0&pdfJsDownload=0&ori
gin=publication_detail
601 http://illumina-chemie.de/dichromatvernichtung-t4201.html
602 See the texts cited in ref 15
603 Young, W Russell III, “Traditional Cyanotype”, in Barnier, John, (ed.),
Coming into Focus: A Step-by-Step Guide to Alternative
Photographic Printing Processes, (San Francisco: Chronicle Books,
2000)
604 Mrhar, Peter, Cyanotype: Historical and Alternative Photography,
2013.
Inspection of the works of the Cyanotype Group on Facebook:
https://www.facebook.com/groups/13488949283/
suggests that about 50% of them are toned.
605 Gent, M, The Cyanotype, (Materials and Techniques Essay, MA Joint
Course in Conservation, RCA and V&A, November, 1992)
606 Hewitt, Barbara, Blueprints on Fabric: innovative uses for cyanotype,
(Loveland CO: Interweave Press, 1995), p64.
607 It is conjectured that this effect may be due to ammonia causing a
reversible deprotonation of water coordinated to Fe(III);
transformation of the ligands from OH2 to OH- would affect the
energy of the electron charge-transfer transition between Fe(III) and
Fe(II), and change the colour.
608 These deeply coloured iron complexes, when prepared directly from
iron salts and an infusion of oak-galls, have provided the basis for
iron-gall writing inks since medieval times.
364
609 Seigel, Judy, e-mail communication to the Alternative Photographic

Process Internet List, 1996.
610 Many examples may be viewed on the Facebook Cyanotype group, loc.
cit., and similar forums.

611 Mrhar, Peter, Cyanotype: Historical and Alternative Photography, 2013
612https://www.facebook.com/photo.php?fbid=983932958300164&set=
gm.10152672650639284&type=1&theater
613 I acknowledge the late Harry Milligan FRPS, former Curator of
Photographs at the Greater Manchester Museum of Science and

Industry, for first drawing my attention in 1984 to the unpublished
endeavours of John Mercer.
614 One of Mercer’s ferric processes was based on the formation of
insoluble cuprous thiocyanate, which could then be toned in various

ways, such as conversion to the brown cuprous ferricyanide.
Mercer’s work therefore antedates the ‘cuprotype’ process attributed
to J B Obernetter, which was published in 1864.
615 ‘Box of Experiments in Chromatic Photography by John Mercer’,
UDCL-8/19, Lancashire County Record Office, Bow Lane, Preston.

616 I thank Jenny Wetton, curator at the Museum of Science and Industry
in Manchester, for access to this item.

617 ‘Photographic Images in Dyes’, The British Journal of Photography, (8
July 1898), p445

618 Eder, J M, Rezepte Tabellen und Arbeitsvorschriften für Photographie
und Reproduktionstechnik, (Halle: Wilhelm Knapp, 1933), p220

619 Anthony’s Photographic Bulletin, (1901), 444. I thank Christopher
James for this reference.

620 Planchon, V, French Patent no 639,189 (13 January 1927)
621 Heron, R, ‘Blueprint into Blackprint’, Afterimage, 1, 9 (December
1973), pp4-7, 12
622 Heron, Reginald, private communication to the author, 1998
623 Brown, George E., Ferric & Heliographic Processes, (London: Dawbarn
and Ward Ltd., n.d. ?1902), p16

624 Ware, Mike, unpublished observations, 1999
625 Dr. David Saunders, National Gallery, London, private communication.
626 Ware, Mike, unpublished observations, 1998
627 http://en.wikipedia.org/wiki/Lead_poisoning#History
Engineering Chemistry, 37, 9 (1945), pp855-861

629 Whiting, Harold, ‘Process of toning blue prints’, U.S. patent no.
397,480, (5 February 1889).

630 Murray, H D, Ferro-prussiate printing, British patent no. 713,944, (10
March 1933); United States patent no. 2,014,692, (17 September

1935)
631 Hammond, A, How to Tone Prints, (Boston: American Photographic
Publishing Co., 1940), pp60-63

632 Somerville, C W, Toning Bromides and Lantern Slides, (London: Hazell,
Watson & Viney, Ltd., 1907), pp53-58; Ward, H S, The Figures, Facts
365
and Formulae of Photography, (London: Dawbarn and Ward Ltd.,

1903), p1
633 Planchon, V, ‘Le gélatino-prussiate’, Bulletin de la Société Française de
Photographie, 2 (February 1927), pp50-3

634 Kirby, J, ‘Fading and Colour Change of Prussian Blue: Occurrences and
Early Reports’, National Gallery Technical Bulletin, 14 (1993), pp63-
70
635 ‘Lead white’ is a basic lead carbonate, PbCO .Pb(OH) . It is possible
3 2
that the basic nature of this pigment may contribute to the
deterioration of Prussian blue by causing alkaline hydrolysis.
636 Kirby, J, and Saunders, D, to be published
637 http://artdaily.com/news/75394/Scientists-and-conservators-make-
startling-discovery-in-painting-by-Christen-K-oslash-
bke#.VKJVh8AwA
638 See Eder, J M, History of Photography, Epstean, E, (trans.), (New York:
Dover Publications Inc., 1978), p130, for an account of Desmortiers’

investigations and references to his original publications.
639 Ibid, p190-192
640 Brande, W T, A Manual of Chemistry, (London: John W Parker, 1841),
p755. Eder’s date of 1849, given in reference 4, for Chevreul’s

finding cannot be correct.
641 Turner, E, Elements of Chemistry, (London: Taylor and Walton, 1847),
p751
642 Reilly, J M, Care and Identification of 19th Century Photographic
Prints, (Rochester: Eastman Kodak Company, 1986), p43

643 Lathrop, A K, ‘The Provenance and Preservation of Architectural
Records’, The Americal Archivist, 43,3 (1980), pp325-338

644 Moor I L, and Moor A H, ‘The Conservation of Anna Atkins’ British
Algae’, Symposium on the 150th Anniversary of Photography, Vevey,

Switzerland, 29 June-2 July, 1989, (Vevey: European Society for the
645 McElhone, J P, ‘Determining Responsible Display Conditions for
Photographs’, Topics in Photographic Preservation, 5 (1993), pp60-

72; American Institute for Conservation, Meeting of the Photographic
Materials Group, Austin, Texas 1993.
646 I am indebted to John McElhone and the National Gallery of Canada for
permission to quote these findings in advance of publication.

647 Moor, I and Moor, A, ‘Atlantis Silversafe Photostore - a suitable paper
for photographic conservation’ Library Conservation News, 30

(1990), pp4-5
648 I am grateful to Peter Bower, paper historian, for advice and for
supplying the samples of these historic papers.

649 I am indebted to Judy Seigel for supplying me with these useful step
tablets.
650 Since exposure is defined as the product of illumination and time,
these two parameters must be combined in reciprocal proportion to

achieve the same result, i.e. if the illumination is halved, the time
must be doubled to give the same exposure. However, there comes
366
a point with all real photosensitive materials when this inverse

proportionality ceases to apply - this is the point of ‘reciprocity
failure’ - and very weak illumination must be compensated by
disproportionately lengthy exposure times.
651 Koerner, Jennifer, and Potje, Karen, “Testing and Decision-Making
Regarding the Exhibition of Blueprints and Diazotypes at the
Canadian Centre for Architecture”, The Book and Paper Group
Annual, 21, (2002), p15-23
652 Except in electrochemical cells where the two processes may be
separated by an electric current.
653 ‘Pigment blue 27: C.I. 77510, 77520’, The Colour Index, (Society of
Dyers and Colorists,) 2nd edn., pp2777, 3621
654 Kirby, loc cit
655 Christensen, P A, Harriman, A, Neta, P, and Richoux, M-C, ‘Photo-
oxidation of Water using Prussian Blue as Catalyst’, Journal of the
Chemical Society, Faraday Transactions I, 81 (1985), pp2461-66
656 The molecule contains both primary and secondary alcohol and ether
functions, which can be oxidized to give a range of substances,
loosely referred to as ‘oxycellulose’.
657 Kosar, J, Light Sensitive Systems, (New York: John Wiley and Sons,
1965), p36
658 Claire Gervais, Marie-Angélique Languille, Solen Reguer, Martine
Gillet, Sébastien Pelletier, Chantal Garnier, Edward P.Vicenzi, and
Loïc Bertrand, ‘Why Does Prussian Blue Fade? Understanding the
Role(s) of the Substrate’, Journal of Analytical Atomic Spectrometry,
28 (2013), pp1600-1611; Claire Gervais, Marie-Angélique Languille,
Solen Reguer, Chantal Garnier, and Martine Gillet, ‘Light and anoxia
fading of Prussian blue dyed textiles’, Heritage Science, 2, 26 (2014)
pp ; Claire Gervais, Marie-Angelique Languille, Giulia Moretti, and
Solen Reguer, ‘X-Ray photochemistry of Prussian blue heritage
materials: Evidence for a substrate-mediated rdox process’,
659 Kirby, loc cit
661 Holtzman’s buffer of pH 9.4 consists of a mixture which is 0.05 molar
in sodium hydrogen carbonate and 0.0165 molar in sodium
carbonate.
662 The buffer was the same as that employed by Holtzman: the solution
was 0.05 molar in sodium hydrogen carbonate and 0.0165 molar in
sodium carbonate.
663 Crawford, W, The Keepers of Light, (New York: Morgan and Morgan,
1979)
664 Moor I L, and Moor A H, ‘The Conservation of Anna Atkins’ British
Algae’, Symposium on the 150th Anniversary of Photography, Vevey,
Switzerland, 29 June-2 July, 1989, (Vevey: European Society for the
665 Wagner, S S, ‘Some Recent Photographic Preservation Activities at the
Library of Congress’, Topics in Photographic Preservation, 4 (1991),
367
pp136-149; American Institute for Conservation, Photographic

Materials Group Meeting, 1991
666 This was ‘purified’ water purchased from a retail pharmacist, and
fulfilling the standards of purity required by the British
Pharmacopoeia. It was thought most realistic to use this, rather than
de-ionised or distilled water which varies in its origins, availability
and quality.
667 Reilly, loc cit
668 Ware, M, Mechanisms of Image Deterioration in Early Photographs,
(London: Science Museum and National Museum of Photography,
Film & Television, 1994)
669 Henry, R J, Controls in Black and White Photography, 2nd edn. Boston &
London: Focal Press, 1986.
670 Blackwell, B, ‘Light Exposure to Sensitive Artworks during Digital
Photography’, Spectra (November 2000). Museum Computer
Network. www.mcn.edu/
[Note: This paper quotes values for UV contents of illumination that
are in error by a factor of 1000, owing to the mistaken use of units
of mW/lm (milliwatts per lumen) instead of m W/lm (microwatts per
lumen).]
671 Vitale, T, ‘Light Levels Use in Modern Flatbed Scanners’, RLG Diginews,
2(5) (October 1998)
http://worldcat.org:80/arcviewer/1/OCC/2007/08/08/0000070511
/viewer/file3642.html#technical
672 Neevel, J. G, ‘Exposure of objects of art and science to light from
electronic flash-guns and photocopiers’, in Contributions of the
Central Research Laboratory to the field of conservation and
restoration. Amsterdam: Central Research Laboratory for Objects of
Art and Science, 1994. 77-87.
673 Ware, M, ‘On the Illumination of Light-Sensitive Photographs’, Topics
in Photographic Preservation, 13 (2009), pp221-233; American
Institute for Conservation of Historic & Artistic Works, Photographic
Materials Group.
674 The presence of the hydronium cation, H3O+, in the zeolytic voids will
defend the Prussian blue against alkaline hydrolysis, while not
presenting a direct threat to the cellulose of the paper.
675 Keefe, L E, and Inch, D, The Life of a Photograph, (London and Boston:
Focal Press, 1984); Reilly, J M, Care and Identification of 19th-
Century Photographic Prints, (New York: Eastman Kodak Co., 1986);
Rempel, S The Care of Photographs, (New York: Nick Lyons Books,
1987); Martin, E, Collecting and Preserving Old Photographs,
(London: William Collins Sons & Co. Ltd., 1988); Hendriks, K B,
Fundamentals of Photograph Conservation: A Study Guide, (National
Archives of Canada, 1993).
676 http://www.sp.se/en/index/services/paper/sidor/default.aspx
677 Bilici, Sirdar, "Sulfamic acid for paper acidification": 2013
http://www.apug.org/forums/forum42/118153-sulfamic-acid-
paper-acidification.html
368
678 Calcium carbonate has a solubility of only ca 0.001 g per 100 cc

water.
679 ISO 2007 Imaging materials - Processed imaging materials - Albums,
framing and storage materials, 18902:2007. Geneva: International

Organization for Standardization.
680 Held, Lauren, Burge, Daniel, and Nishimura, Douglas, 'The Effects of
Acidic, Alkaline, Buffered, and Non-Buffered Enclosures on

Cyanotypes', Topics in Photographic Preservation, 14 (2011), pp237
- 249.
681 See, for example: Brinckman, E, Delzenne, G, Poot, A, and Willems, J,
Unconventional Imaging Processes, (London: The Focal Press, 1978),

p82; Jacobson, K, and Jacobson, R E, Imaging Systems, (London:
Focal Press, 1976) pp125-6; Kosar, J, Light Sensitive Systems, (New
York: John Wiley and Sons, 1965), pp28-37
682 The most recent review article on this topic at the time of writing is:
Dunbar, K R, and Heintz, R A, ‘Chemistry of Transition-Metal

Cyanide Compounds - Modern Perspectives’, Progress in Inorganic
Chemistry, 45 (1997), pp283-391
683 Norma R. de Tacconi, Krishnan Rajeshwar, and Reynaldo O. Lezna,
‘Metal Hexacyanoferrates: Electrosynthesis, in Situ Characterization,

and Applications’, Chemistry of Materials, 15 (2003), pp3046-3062
684 Sharpe, A G, The Chemistry of Cyano Complexes of the Transition
Metals, (London: Academic Press, 1976), pp121-6

685 Itaya, K, Uchida, I, and Neff, V, ‘Electrochemistry of Polynuclear
Transition Metal Cyanides: Prussian Blue and Its Analogues’,

Accounts of Chemical Research, 19 (1986), pp162-8

687 Williams, H E, Cyanogen Compounds, (London: Edward Arnold, 1948),
p191
688 Duncan, J F, and Wigley, P W R, ‘The Electronic Structure of the Iron
Atoms in Complex Iron Cyanides’, Journal of the Chemical Society,

(1963), pp1120-5
689 Bonnette, A K, and Allen, J F, ‘Isotopic Labeling for Mössbauer Studies.
An Application to the Iron Cyanides’, Inorganic Chemistry, 10, 8

(1971), pp1613-6
690 Wertheim, G K, and Rosencwaig, A, ‘Characterization of Inequivalent
Iron Sites in Prussian Blue by Photoelectron Spectroscopy’, Journal of

Chemical Physics, 54 (1971), pp3235-7
691 Davidson and Welo, Journal of Physical Chemistry, 32 (1928), pp1191
692 Louise Samain, Fernande Grandjean, Gary J. Long, Pauline Martinetto,
Pierre Bordet, and David Strivay, 'Relationship between the Synthesis

of Prussian Blue Pigments, Their Color, Physical Properties, and Their
Behavior in Paint Layers', Journal of Physical Chemistry C, 117, 19
(2013), pp9693-9712.
693 Rosseinsky, D R, Tonge, J S, Berthelot, J, and Cassidy, J F, ‘Site-
transfer Conductivity in Solid Iron Hexacyanoferrates by Dielectric
369
Relaxometry, Voltammetry and Spectroscopy’, Journal of the

Chemical Society, Faraday Transactions I, 83 (1987), pp231-243
694 Keggin, J F, and Miles F D, ‘Structures and Formulae of the Prussian
Blues & Related Compounds’, Nature, 137 (4 April 1936), pp577-8
695 See ref 667
696 Xiao-Wei Cao, Rong Wang, Gui-Hua Wang, Zong-Rang Zhang, ‘In situ
Raman Spectroelectrochemical Characterization of th Prussian blyue
modified Platinum Electrode’,
697 Dostal A, Kauschka, G, Reddy, S, and Scholtz, F, ‘Lattice contractions
and expansions accompanying the electrochemical conversions of
Prussian blue and the reversible and irreversible insertion of
rubidium and thallium ions’, Journal of Electroanalytical Chemistry,
406, (1996), pp155-163
698 Huheey, J E, Inorganic Chemistry: Principles of Structure and
Reactivity, 3rd edn. (New York: Harper International, 1983), pp72-78
699 Plichon, V, and Besbes, S, ‘Electrochemical Investigations of K+-free
Prussian blue Films’, Electrochimica Acta, 37, 3 (1992) pp501-6
700 Kulesza, P J, Zamponi, S, Berrettoni, M, Marassi, R, and Malik, M A,
‘Preparation, Spectroscopic Characterization and Electrochemical
Charging of the Sodium-containing Analog of Prussian Blue’,
Electrochimica Acta, 40, 6 (1995), p681
701 Lundgren, C A, and Murray, R W, ‘Observations on the Composition of
Prussian Blue Films and Their Electrochemistry’, Inorganic Chemistry,
27 (1988), pp933-9
702 Holtzmann, loc cit
703 Kahler, H, Komorskylovric, S, Hermes, M and Scholz, F, ‘A Prussian
Blue-based Reactive Electrode (Reactrode) for the Determination of
Thallium Ions’, Fresenius Journal of Analytical Chemistry, 356
(1996), pp204-8
704 Dresow, B, Nielsen, P, Fischer, R, Pfau, A A, and Heinrich, H H, ‘In-vivo
Binding of Radiocesium by Two Forms of Prussian Blue and by
Ammonium Iron Hexacyanoferrate(II)’, Journal of Toxicology-Clinical
Toxicology, 31, 4 (1993), pp563-9; Verzijl, J M, Joore, H C, Vandijk,
A, Wierckx, F C, Savelkoul, T J, and Glerum, J H, ‘In-vitro Cyanide
Release of Four Prussian blue Salts used for the Treatment of Cesium
Contaminated Persons’, Journal of Toxicology-Clinical Toxicology,
31, 4 (1993), pp553-562
705 Pearce, J, ‘Studies of any Toxicological Effects of Prussian Blue
Compounds in Mammals - A Review’, Food and Chemical
Toxicology, 32, 6 (1994), pp577-582
706 Buser, H J, Schwarzenbach, D, Petter, W, and Ludi, A, ‘The Crystal
Structure of Prussian Blue: Fe4[Fe(CN)6]3.xH2O’, Inorganic Chemistry,
16, 11 (1977), pp2704-9
707 Herren, F, Fischer, P, Ludi, A, and Hälg, W, ‘Neutron Diffraction Study
of Prussian Blue Fe4[Fe(CN)6]3.xH2O. Location of Water Molecules and
Long-Range Magnetic Order’, Inorganic Chemistry, 19 (1980),
pp956-959
370
708 The reader may find it helpful to be reminded of the formal oxidation
states of the iron atoms in the formulations of compounds of this
family.
709 Wilde, R E, Ghosh, S N, and Marshall B J, ‘The Prussian Blues’,
Inorganic Chemistry, 9, 11 (1970), pp2512-6
710 Christensen, P A, Hamnett, A, and Higgins, S J, ‘A Study of
Electrochemically grown Prussian Blue Films using Fourier-transform
Infra-red Spectroscopy’, Journal of the Chemical Society, Dalton
Transactions, (1990), pp2233-8
711 Ganguli, S, and Battacharya, M, ‘Studies of Different Hydrated Forms
of Prussian Blue’, Journal of the Chemical Society, Faraday
Transactions 1, 79 (1983), pp1513-1522; Bal, B, Ganguli, S, and
Battacharya, M, ‘Bonding of Water Molecules in Prussian Blue. A
Differential Thermal Analysis and Nuclear Magnetic Resonance
Study’, Journal of Physical Chemistry, 88 (1984), pp4575-7
712 Inoue, H, Nakazawa, T, Mitsuhashi, T, Shirai, T, Fluck, E,
‘Characterization of Prussian Blue and Its Thermal Decomposition
Products’, Hyperfine Interactions, 46 (1989), pp725-731
713 Kravzov, J, Rios, C, Altagracia, M, Monroynoyola, A, and Lopez, F,
‘Relationship between Physicochemical Properties of Prussian Blue
and its Efficacy as Antidote against Thallium Poisoning’, Journal of
Applied Toxicology, 13, 3 (1993), pp213-216
714 It has been determined by radio-isotope studies that this is the
mechanism, and not the cyanide ligands ‘switching’ their
coordination ends from one iron centre to the other.
715 Wilde et al, loc cit
716 Reguera, E, Fernández-Bertrán, J, Dago, A, and Díaz, C, ‘Mössbauer
Spectroscopic Study of Prussian Blue from Different Provenances’,
Hyperfine Interactions, 73 (1992), pp295-308
717 Dostal, A, Meyer, B, Scholtz, F, Schröder, U, Bond, A M, Marken, F, and
Shaw, S J, ‘Electrochemical Study of Microcrystalline Solid Prussian
Blue Particles Mechanically Attached to Graphite and Gold Electrodes:
Electrochemically Induced Lattice Reconstruction”, Journal of
Chemical Physics, 99 (1995), pp2096-2103; see also Lundgren and
Murray, loc cit
718 Robin, M B, and Day, P, ‘Mixed Valence Chemistry’, Advances in
Inorganic Chemistry and Radiochemistry, 10 (1967), pp294-9
719 Mortimer, R J and Rosseinsky, D R, ‘Iron Hexacyanoferrate Films:
Spectroelectrochemical Distinction and Electrodeposition Sequence
of ‘Soluble’ (K+-containing) and ‘Insoluble’ (K+-free) Prussian Blue,
and Composition Changes in Polyelectrochromic Switching’, Journal
of the Chemical Society, Dalton Transactions, (1984), pp2059-2061
720 Also now called the molar absorptivity.
721 Nobrega, J.A.; Lopes, G.S., Talanta 43 (1996), p971.
722 England, S J, Kathirgamanathan, P, and Rosseinsky, D R, ‘Perturbation
Calculation from the Charge-Transfer Spectrum Data of Intervalence
Site-Transfer D.C. Conductivity in Prussian Blue’, Chemical
Communications, (1980), pp840-1
371
723 Robin, M B, ‘The Color and Electronic Configurations of Prussian Blue’,

Inorganic Chemistry, 1, 2 (1962), pp337-343. Robin calculates that
the optical electron is 99% localised on the Fe(II) centre in the ground
state.
724 Braterman, P S, ‘Charge Transfer Phenomena in Some Inorganic
Complexes. Part III. Ultraviolet Spectra of Transition-Metal
Derivatives of Iron Group Cyanide Complex Anions’, Journal of the
Chemical Society (A), (1966), pp1471-5
725 To take proper account of the symmetry, the d orbitals of all six Fe(III)
centres surrounding any given Fe(II) centre are taken as symmetry-
adapted linear combinations.
726 Williams, op cit, p219
727 Nechitayilo, V B, Styopkin, V I, Tkachenko, Z A, Goltsov, Y G,
Sherstyuk, V P, and Zhilinskaya V V, ‘Structure and Electrochromic
Properties of Ferric Aquapentacyanoferrate - A New Analogue of
Prussian Blue’, Electrochimica Acta, 40, 15 (1995), pp2501-4
728 Briggs, S H C, ‘The Nature of the b-Ferricyanides and the b-
Ferrocyanides’, Journal of the Chemical Society, 117 (1920),
pp1026-1034
729 I thank Susan Blake for instrumental assistance in obtaining these
spectra.
730 Vaucher, Sébastien, Mei Li, and Mann, Stephen, ‘Synthesis of Prussian
Blue Nanoparticles and Nanocrystal Superlattices in Reverse
Microemulsions’, Angewandte Chemie, International Edition, 39, 10,
(2000), pp1793-6
731 I thank David Collison for providing access to this instrument.
732 Flynn, C M, ‘Hydrolysis of Inorganic Iron(III) Salts’, Chemical Reviews,
84 (1984), pp31-41
733 Duncan and Wigley, loc cit, found the Prussian blue formed by air
oxidation of Prussian white to have the composition
KFe[Fe(CN)6].KOH.
734 Dostal, A, Meyer, B, Scholtz, F, Schröder, U, Bond, A M, Marken, F, and
Shaw, S J, ‘Electrochemical Study of Microcrystalline Solid Prussian
Blue Particles Mechanically Attached to Graphite and Gold Electrodes:
Electrochemically Induced Lattice Reconstruction”, Journal of
Chemical Physics, 99 (1995), pp2096-2103
735 Chadwick, B M, and Sharpe, A G, ‘Transition Metal Cyanides and their
Complexes’, Advances in Inorganic Chemistry and Radiochemistry, 8
(1966), pp119-121
736 Goncalves, R, Kellawi, H, and Rosseinsky, D R, ‘Electron-transfer
Processes and Electrodeposition involving the Iron
Hexacyanoferrates studied Voltammetrically’, Journal of the
Chemical Society, Dalton Transactions, (1983), pp991-4
737 Crumbliss, A L, Lugg, P S, Childers, J W, Palmer, R A, ‘Fourier
Transform Infrared Photothermal Spectroscopic Characterization of
Prussian Blue Surface Modified Electrodes’, Journal of Physical
Chemistry, 89 (1985), 482-8
738 Williams op cit p227
372
739 Ozeki, T, Matsumoto, K, and Hikime, S, ‘Photoacoustic Spectra of

Prussian Blue and Photochemical Reaction of Ferric Ferricyanide’,
Analytical Chemistry, 56 (1984), pp2819-2822
740 Feigl, Fritz, Spot Tests in Inorganic Analysis, 6th Edn. p373
741 Mitsumasa Saito, Masanori Seki, Ken-ichiro Iida, Hiroaki Nakayama,
and Shin-ichi Yoshida, “A Novel Agar Medium to Detect Hydrogen

Peroxide-Producing Bacteria Based on the Prussian Blue-Forming
Reaction”, Microbiological Immunology, 51:9, (2007), pp889-892.
742 Mortimer, R J and Rosseinsky, D R, ‘Electrochemical Polychromicity in
Iron Hexacyanoferrate Films, and a New Film Form of Ferric

Ferricyanide’, Journal of Electroanalytical Chemistry, 151 (1983),
pp133-147
743 The Nernst Equation for the potential is given by E = E0 + (RT/nF)ln K
s
744 Neff, V D, Journal of the Electrochemical Society, 125 (1978), p884
745 Mortimer, R J, ‘Electrochromic materials’Chemical Society Reviews, 26,
3, (1997), pp147-156
746 Mortimer and Rosseinsky, loc cit
747 Christensen, P A, Hamnett, A, and Higgins, S J, ‘A Study of
Electrochemically grown Prussian Blue Films using Fourier-transform

Infra-red Spectroscopy’, Journal of the Chemical Society, Dalton
Transactions, (1990), pp2233-8
748 Itaya et al, loc cit
749 Dostal et al, loc cit
750 Murray, H D, ‘The Chemistry of Photocopying Processes’, Chemistry
and Industry, (14 September 1940), pp645-6
751 Suzuki, S, Matsumoto, K, Harada, K, and Tsubura, E, ‘Photo-
Physicochemical Behaviour of Iron and Chromium Compounds’,

Photographic Science and Engineering, 12, 1 (1968), pp2-16
752 Suzuki, S, ‘The Mechanism of the Photochemical Reaction of the Blue
Print Process, and the Photoconductivity Phenomenon of some Iron

Complexes’, Proceedings of the Royal Photographic Society’s
Centenary Conference, London, 1953, Schultze, R S, (ed.) (London:
Royal Photographic Society, 1955), pp253-6
753 Kosar, J, Light Sensitive Systems, (New York: John Wiley and Sons,
1965), pp28-37
754 Jacobson, K, and Jacobson, R E, Imaging Systems, (London: Focal
Press, 1976), pp125-6

755 Balzani, V and Carassiti, V, Photochemistry of Coordination
Compounds, (London: Academic Press, 1970), p173

756 Ware, M J, ‘An Investigation of Platinum and Palladium Printing’,
Journal of Photographic Science, 34 (1986), pp13-25

757 as the photoproduct, Prussian blue, builds up it will increasingly
absorb the incident UV radiation - the phenomenon of "self-

masking".
758 Properly, in a heterogeneous system, the optical density should be
calculated using the Koubelka-Munk theory, which takes account of

significant scattering. However the present approximation will serve
for an ‘order-of-magnitude’ calculation.
373
759 Rigorously, the yield should be calculated by integrating over the

whole waveband of actinic light, to take account of the variations of
I, f, and f with wavelength l. The present approximation will serve
for most purposes, however.
760 This corresponds to a pigment having an extinction coefficient in the
order of e = 104 dm3/mol/cm
761 The intrinsic printing exposure range of the iron-based sensitizers
generally appears to be 7 or 8 stops, corresponding to arithmetic
factors of 27 = 128, to 28 = 256.
762 According to Kosar, op cit p32, commercial blueprint paper has a
coating weight of 0.6 to 0.75 g/m2 of iron salts.
763 Brinckman, E, Delzenne, G, Poot, A, and Willems, J, Unconventional
Imaging Processes, (London: The Focal Press, 1978), pp103, 135.
There are some numerical errors in the concentration values quoted
by these authors for their calculation, but the final result is
consistent.
764 Assuming that the Beer-Lambert Law is valid for suspensions or layers
of Prussian blue: Dref = 2eCd because the light traverses the
absorbent layer a second time after being reflected by the white
paper base; where C is the concentration in mol/dm3 and d is the
path length in cm. Conversion of units to SI leads to the given
equation.
765 Magini, M, ‘Evidence of Iron(III)-Oxalato Complex Formation in
Aqueous Solution from X-Ray Diffraction’, Chemical Physics Letters,
78, 1 (1981), pp106-111
766 Cooper, G D, and DeGraff, B A, ‘On the Photochemistry of the
Ferrioxalate System’, Journal of Physical Chemistry, 75 (1971),
pp2897-2902, and references cited therein; Sato, H, and Tominaga,
T, Bulletin of the Chemical Society of Japan, 52, 5 (1979), pp1402-7;
Ionescu, S G, and Oncescu, T, ‘Photolysis of K3[Fe(C2O4)3] in
Dimethylformamide’, Journal of Photochemistry, 23 (1983), pp45-50
767 Parker, C A, ‘A New Sensitive Chemical Actinometer’, Proceedings of
the Royal Society (London), A220 (1953), pp104-115; Hatchard C G,
and Parker, C A, ‘ Proceedings of the Royal Society (London), A235
(1956), p518
768 Baxendale, J H, and Bridge, N K, ‘The Photoreduction of Some Ferric
Compounds in Aqueous Solution’, Journal of Physical Chemistry, 59
(1955), pp783-8
769 Spencer, H E, ‘Quantum Yields of Production of Iron(II) and of
Cobalt(II) in the Photolysis of Solid Potassium Trisoxalatoferrate(III)
Trihydrate and Potassium Trisoxalatocobaltate(III) Trihydrate’,
Journal of Physical Chemistry, 73, 7 (1969), pp2316-2320
770 Spencer, H E, and Schmidt, M W, ‘Photochemical Studies of Solid
Potassium Trisoxalatoferrate(III) Trihydrate’, Journal of Physical
Chemistry, 75, 19 (1971), pp2986-2990;
Brar, A S, and Randhawa, B S, ‘Mössbauer Study of Solid State
Photolysis of Alkali Tris(oxalato)ferrates(III)’, Bulletin of the Chemical
Society of Japan, 54 (1981), pp3166-9
374
771 Hatchard C G, and Parker, C A, ‘Photodecomposition of Complex

Oxalates. Some Preliminary Experiments by Flash Photolysis’, Journal
of Physical Chemistry, 63 (1959), pp22-6
772 Decurtins, S, Schmalle, H W, Schneuwly, P, and Oswald, H R,
‘Photochemical Synthesis and Structure of a 3-Dimensional Anionic
Polymeric Network of an Iron(II) Oxalato Complex with Tris(2,2’-
bipyridine)iron(II) Cations, Inorganic Chemistry, 32 (1993), pp1888-
92
773 Attfield, Chemistry: General, Medical, and Pharmaceutical, (London:
John van Voorst, 1880, 9th edn.) p172
774 Taniguchi, M, Imamura, H, Shirota, T,, Okamatsu, H, Fujii, Y, Toba, M,
and Hashimoto, F, ‘Improvement in Iron-Deficiency Anemia through
Therapy with Ferric Ammonium Citrate and Vitamin-C and the
Effects of Aerobic Exercise’, Journal of Nutritional Science and
Vitaminology, 37, 2 (1991), pp161-171
775 Valenta, E, Photographisches Correspondenz (1897), p74
776 Clerc, L P, Photography: Theory and Practice, (2nd English Edn.,
London: Sir Isaac Pitman & Sons, 1937)
777 Uphaus, R A and Blake, M I, ‘Electron Spin Resonance Characterization
of the Ferric Ammonium Citrates (Scale Salts)’, Naturwissenschaften,
68 (1981), pp270-1
778 Spiro, T G, Pape, L, and Saltman, P, ‘The Hydrolytic Polymerization of
Ferric Citrate. I. The Chemistry of the Polymer’, Journal of the
American Chemical Society, 89, 22 (1967), pp5555-9
779 Abrahamson, H B, Rezvani, A B, and Brushmiller, J G, ‘Photochemical
and spectroscopic studies of complexes of iron(III) with citric acid
and other carboxylic acids’, Inorganica Chimica Acta, 226 (1994),
pp117-127 and references cited therein.
780 Timberlake, C F, ‘Iron-Malate and Iron-Citrate Complexes’, Journal of
the Chemical Society, (1964), pp5078-5085
781 Shweky, I, Bino, A, Goldberg, D P, and Lippard, S J, ‘Synthesis ,
Structures, and Magnetic Properties of Two Dinuclear Iron(III) Citrate
Complexes’, Inorganic Chemistry, 33, 23 (November 9 1994),
pp5161-2
782 Abrahamson, Harmon B, Journal of Chemical Education, 78, (2001),
p311
783 Faust, B C, and Zepp, R G, ‘Photochemistry of Aqueous Iron(III)-
Polycarboxylate Complexes: Roles in the Chemistry of Atmospheric
and Surface Waters’, Environmental Science and Technology, 27
(1993), pp2517-2522
784 Balzani, V, and Carassiti, V, Photochemistry of Coordination
Compounds, (London: Academic Press, 1970), pp162-4
785 Rader, W S, Solujic, L, Milosavljevic, E B, and Hendrix, J L, ‘Sunlight-
Induced Photochemistry of Aqueous Solutions of
Hexacyanoferrate(II) and -(III) Ions’, Environmental Science and
Technology, 27 (1993), pp1875-1879
786 The anionic formula, FeIII[FeII(CN)6]–, used here to represent the
Prussian blue formed in the cyanotype process, is only intended as a
375
formalism; various cations may be present, e.g. K+, NH4+, H3O+, or

none, or even anions, with the consequence that the lattice may have
defects, as explained in the previous Appendix.
787 In this respect ‘cyanotype Prussian blue’ differs in its method of
preparation from the ‘normal’ form of Turnbull’s blue, which is

always obtained from uncomplexed, aquated iron(II).
788 Kosar, op cit, pp32-3
789 Suzuki et al, loc cit
790 Murray, loc cit
791 Kulesza P J, Malik, M A, Denca, A, and Strojek, J, ‘In Situ FT-IR/ATR
Spectroelectrochemistry of Prussian Blue in the Solid State’,

Analytical Chemistry, 68, 14 (1996), 2442-6
792 Itaya, K, Shoji, N, Uchida, I, ‘Catalysis of the Reduction of Molecular
Oxygen to Water at Prussian Blue Modified Electrodes’, Journal of the

American Chemical Society, 106 (1984), pp3423-3429
793 Holden, J, British Patent No.418,369 (24 October, 1934)
794 Faust and Zepp loc cit
795 Jahoda, E, US Patent No.2,517,111 (1 August 1950)
796 Williamson, H E, Cyanogen Compounds, (London: Edward Arnold &
Co., 1948), p200

797 Reynolds, A P, US Patent No.2,093,738 (21 September 1937)
798 Crowley, C A, and Goodyear, G H, US Patent No. 2,350,991 (13 June
1944)
799 Jahoda loc cit
800 ‘Pigment blue 27: C.I. 77510, 77520’ The Colour Index, (Society of
Dyers and Colorists, 19 ) 2nd edn, pp2777, 3621

801 Christensen, P A, Harriman, A, Neta, P, and Richoux, M-C, ‘Photo-
oxidation of Water using Prussian Blue as Catalyst’, Journal of the

Chemical Society, Faraday Transactions I, 81 (1985), 2461-66
802 Nishizawa, M, Kuwabata, S, and Yoneyama, H, ‘Photoimage Formation
in a TiO2 particle-incorporated Prussian Blue Film’, Journal of the

Electrochemical Society, 143, 11 (1996), pp3462-65
803 Bock, C R, Connor, J A, Gutierrez, A R, Meyer, T J, Whitten, D G,
Sullivan, B P, and Nagle, J K, ‘Estimation of Excited-State Redox

Potentials by Electron-Transfer Quenching. Application of Electron-
Transfer Theory to Excited-State Redox Processes’, Journal of the
American Chemical Society, 101, 17 (1979), pp4815-24; Julliard, M,
and Chanon, M, ‘Redox Properties of Photoexcited States’, Chemistry
in Britain, 18, 8 (1982), pp558-562; Balzani,V, and Boletta, F,
‘Transition Metal Complexes as Mediators in Photochemical and
Chemiluminescence Reactions’, Comments on Inorganic Chemistry,
2, 5 (1983), pp211-226
804 The identification of millilitres (ml) with cubic centimetres (cc) is not
strictly exact, but serves for most practical purposes.
805 See, for example: Skoog, D A, and West, D M, Fundamentals of
Analytical Chemistry, (New York: Holt, Reinhart and Winston, 1976)

806 Since the litre is not an approved unit of the SI system, the units here
should strictly be written as moles/dm3.
376
807 Though with the growth of digital imaging technology, this skill may
become rare.
808 Weast, R C (ed), Handbook of Chemistry and Physics, (Florida:
Chemical Rubber Company Press Inc, 1980), Tables D-227 to D-276
809 The Merck Index
810 Weast, op. cit., Tables F-3, F-4
377

Cyanomicon II

Uploaded by

Copyright:

Available Formats

Cyanomicon II

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cyanomicon II

Uploaded by

Copyright:

Available Formats

Cyanomicon

History, Science and Art of Cyanotype:

Honorary Fellow in Chemistry, University of Manchester, UK

Revised & Illustrated 2016

4.4.2 John Mercer’s ‘chromatic photographs’ ................................. 93

7.1.2 Mixing and coating .............................................................. 202

9.3 Peptization of Prussian blue ....................................................... 265

Warning and Disclaimer

Out of the entire gamut of photographic printing techniques left to us by

1.1 The legacy of John Herschel

Herschel also clearly entertained a practical objective: the hope of

Fig. 1.1 Sir John F.W. Herschel by Julia Margaret Cameron

limitations in the chemical knowledge of Herschel’s day pre-conditioned

1.2 Equivocal blue images

Fig. 1.2 "Cyanotype by Sir John Herschel"

In his scholarly exposition of the famous ‘Brewster Album’ in the J.

Talbot himself made no use whatever of the cyanotype process; but,

“ … no one but a vandal would print a landscape in red, or in

The anonymous “amateur” in question may have been Washington

The Scottish National Portrait Gallery holds a magnificent photographic

1.3 Acceptance of blue photographs

Fig. 1.3 Paul Burty Haviland, Florence Peterson in a Kimono, ca 1910

Neither was the process despised among the amateur community in

According to Nancy Newhall,20 he deemed the process “villainous”, a

Fig. 1.4 Edward Curtis, Misham

1.4 Commercial use of blueprint

became obsolete because, as Frances Robertson explains, it has now

Fig. 1.5 Blueprint of a steam engine

The era of the blueprint as a copying process was heralded by the

1.5 Artistic renaissance of cyanotype

1.6 A rare colour in nature

landscape, viewed through a chemist’s spectroscopic eye. The plant

Fig. 1.6 Colours of landscape and sky

In 1789 the Swiss natural philosopher Horace-Bénédict de Saussure

published a memoir in 1877 describing his experiments with cultivation

Fig. 1.8 Morpho butterfly

ultraviolet; a blue flower may be even more conspicuous to them than to

1.7 Symbolic blue in art and religion

Fig. 1.9 David Vases, 1351

The first synthetic pigment, Egyptian blue was used to decorate

The association with the colour of the celestial hemisphere adds an

Jolan de Jacobi, in her study of Jungian psychology, says:

In a religious context, blue is the colour symbolising some of the

Fig. 1.10 The Wilton Diptych

Fig. 1.11 The Lord Krishna

J E Cirlot, in his work on symbols, notes that:

night sky is therefore the colour associated with quiescence and

1.8 Liberation from blue preciousness

§2.1) by an artists’ colour-maker, heralded the liberation of painting

The Expressionist school of Munich painters - the so-named Blaue

Another turning point in the artistic significance of the colour blue

Fig. 1.13 Franz Marc, The Large Blue Horses, 1911

The colour continues to exert a fascination, as may be read in the

antithetical interpretations of the colour blue, which have some affinities

1.9 Colour language and coordinates

Blue is the colour we humans perceive least efficiently, a fact that is

19C7 Royal blue

20A8 Ultramarine (artificial)

21C8 Ultramarine (genuine)

22B7 Cobalt blue