National Gallery Technical Bulletin Volume 14, 1993

Fading and Colour Change of Prussian Blue !ccurrences and

"arly #e$orts
Jo Kirby
If one considers the palette of the mid-seventeenth-century painter, it is not immediately obvious that there was any shortage of
blue pigments: the vivid, almost overpowering splendour of ultramarine and the sparkling greener blue of azurite come
immediately to mind. Smalt, at its best and when freshly painted, could have a surprising intensity of colour.
In practice, however, all these pigments had their disadvantages: there was no one perfect blue. Ultramarine was scarce,
etremely dear and, partly because of its pronounced purplish hue, did not necessarily mi well with other colours. !y the latter
part of the seventeenth century azurite had also become etremely scarce: indeed, it appears to have fallen into disuse in the
eighteenth century. Its synthetic e"uivalent, blue verditer, seems always to have found greater favour as a decorator#s colour$ a
somewhat gritty teture and greenish cast, which tended to become more pronounced with the passage of time, made it less
popular with painters in oil and watercolour. Smalt was very pale when ground at all finely and it sank in oil. It also gradually lost
its colour in an oil medium, which itself discoloured badly %note &'. (ll these blues were transparent: their intensity of colour
depended upon their being coarsely ground. Indigo, the only organic blue pigment then used in oil, had a deep, intense colour
and, mied with yellow, was useful for making greens, but it was known to fade and was not regarded highly. )hus, by the
beginning of the eighteenth century, the choice of blue pigments was in fact very restricted %note *'.
)hen, in &+&,, an announcement for a new blue pigment appeared in -Mi scel l anea Ber ol i nensi a ( not e 3) .

Invented in
!erlin several years previously, and subse"uently known as .russian blue, the pigment was described as durable in both oil
and watercolour$ resistant to acids$ unchanged by atmospheric conditions$ easily ground to the finest powder and with ecellent
covering power. It was also non-poisonous, mied well with other colours and, the notice claimed, #it survives in "uick lime.# /ot
the least of its virtues was that its price was one-tenth that of the dearest ultramarine, with which it was compared very
favourably a few years later: #/owadays a very beautiful blue pigment has been invented in !erlin, e"ual to or ecelling
Ultramarine.... # %note 0'
/ot surprisingly, the earliest reports of the use of .russian blue appear in the neighbourhood of !erlin. )races of .russian blue
and lead white, applied over gold leaf, were detected on one of the original lead appli"ues on the -Ehr engi t t er %decorated
fencing' at 1harlottenburg 1astle, !erlin %note 2'. )he lead appli"ue sampled dates from &+&*. Use of the pigment spread
remarkably "uickly, however: it appears to have been available in 3rance and in 4ngland well before the method of
manufacture was first published by 5ohn 6oodward in &+*0, in 7ondon %note 8'. Its presence has, for eample, been confirmed
in a number of paintings by 6atteau, who died in &+*&. In his haunting picture of a clown, -Gi l l es %.aris, 9us:e du 7ouvre',
painted in .aris and probably dating from around &+&;<&=, it is found mied with lead white in the blue paint of the sky %note +'.
It has also been found in -Recr eat i on i t al i enne %!erlin, 1harlottenburg 1astle', which may date from around &+&2<&8
%note ;', and in the -Comdi ens i t al i ens %6ashington, /ational >allery of (rt', which was probably painted during
6atte?u#s stay in 7ondon, &+&=<*, %note ='.
!ccurrences of faded Prussian %lue
)he modern artists# pigment is generally held to be reasonably light-fast under normal conditions$ suppliers of artists# materials
describe it as being durable or moderately durable in oil and in watercolour, unless used in an etremely thin wash or mied
with a very high proportion of white %note &,'. ( thin wash of the pigment is said to fade in strong light, but regain its colour in
the dark %note &&'. In mitures with white, the light-fastness of the pigment is described as ecellent when the proportion of blue
is high %=,@ .russian blue, &,@ titanium white', but decreases as the proportion of white pigment is increased$ a miture of
&@ .russian blue to ==@ titanium white shows poor durability %note &*'.
Auring a recent eamination of >ainsborough#s early landscape, -Gai nsbor ough' s For est ( ' Cor nar d Wood' ) %/>
=*2', of 1.&+0; %note &B', it was noticed that the palest tint of the sky had faded in comparison with the identical paint on the
turnover edge, which had been protected by the frame. Samples of paint taken from different areas of the painting revealed that
where the sky paint had been protected by another paint layer < where the foliage of a tree had been painted over the sky, for
instance < it was found to be a brighter blue than that of the eposed sky paint. (nalysis showed that the sky had been painted
using a little .russian blue mied with a great deal of lead white. In the cross-section prepared from the faded paint, this layer
appeared greyish white in colour, with very little blue visible ecept for a few specks at the bottom where it had been protected
to some etent by the paint above.
)he same phenomenon was observed in >ainsborough#s -Mr and Mr s ndr e!s %/> 8B,&' of c.&+2, and -"he
Wat er i ng #l ace %/> &,=' of c.&+++, where faded .russian blue was found in the sky paint, mied with lead white. In -Mr s
$i ddons %/> 8;B', which >ainsborough completed early in &+;2, the dark blue stripes of the sitter#s dress, which contained
a high proportion of .russian blue, appeared unchanged, but the light blue, painted in .russian blue mied with a high
proportion of lead white, had faded. )he same miture of pigments was found in the sky of two paintings by Cichard 6ilson:
-%ol t Br i dge on t he Ri &er 'ee %/> 8&=8' and its companion -"he (al l e) o* t he 'ee %/> 8&=+'$ as before, the
blue had faded %note &0'. 6hen 6atteau#s -Recr eat i on i t al i enne was eamined in &=;2 at the Institut fur )echnologie der
9alerei, Staatliche (kademie der !ildenden DEnste, Stuttgart, the .russian blue used in the paint of the sky was similarly found
to have faded, ecept in those areas covered by the frame.
)he presence of faded .russian blue has also been observed in a number of Italian School paintings in the 1ollection. In
1analetto#s (eni ce+ , Cam-o $an (i dal and $ant a Mar i a del l a Car i t . ( ' "he $t onemason' s /ar d' ) %/>
&*+', painted between &+*8 and &+*;, faded .russian blue can be found, mied with lead white, in the light blue paint of the
sky. )he same effect was observed in -(eni ce+ 0--er Reaches o* t he Gr and Canal !i t h $an $i meone
#i ccol o %/> &8B', dating from c.&+B;, in -(eni ce+ "he Basi n o* $an Mar co on scensi on 'a) %/> 002B',
probably dating from c.&+0,, and in -(eni ce+ "he Feast 'a) o* $ai nt Roch %/> =B+', of c.&+B2, where it is less
marked %note &2'.
Fn the surface of the painting the pale blue areas < typically skies, as the above eamples show < where fading has occurred
often have a patchy, greyish appearance. )he phenomenon is easier to see in cross-sections of paint and can be illustrated by
a sample taken from the lightest area of )ime#s blue drapery in )iepolo#s -n l l egor ) !i t h (enus and "i me %/> 8B;+'
%.lates & and *, p. 8*', perhaps dated c.&+20 %note &8'. )he fading is most obvious at the top of the layer, which consists %as
in all the eamples described' of lead white mied with a little .russian blue. )he colour changes gradually from a greyish white
at the top, hardly recognisable as blue apart from the larger, more intensely coloured, translucent flakes and undispersed lumps
of blue pigment, to a more intense blue at the bottom of the layer. Gery occasionally, although not in this case, it is possible to
see flakes of blue within the faded part of the layer which have become greenish at the top. In a cross-section prepared from a
paint sample taken from a brighter blue area of the same drapery, however, it can be seen that the uppermost layer of blue
paint contains a far higher proportion of .russian blue mied with lead white and appears unchanged %see .late B'.
6here the presence of .russian blue is suspected it is important that it be confirmed, to eclude the possibility of the presence
of indigo. )his may be done by microchemical tests %the pigment is decolorised by alkali with the formation of brown ferric
hydroide' or by instrumental methods, such as energy-dispersive H-ray microanalysis %4AH', to confirm the presence of iron.
(s count rates for the principal iron peaks are "uite low for .russian blue, it is important to allow a long enough counting time in
ac"uiring spectra to give a reliable result %note &+'.
"arly re$orts of the effects of light on the $igment
)he reputation of .russian blue as a durable pigment, rivalling ultramarine, was not maintained throughout the century or so
that elapsed following the dramatic announcement of the pigment in &+&,. (t that time, because its ingredients included dried
blood or other animal matter, .russian blue was considered to be what would nowadays be called an organic pigment$ such
pigments, derived from plant or animal sources, were known to be fugitive. )he chemist .eter Shaw commented in his public
lectures, published in &+B0, that because of the animal and vegetable matters used in its preparation the durability of .russian
blue might be suspected, were it not for the fact that the colour seemed to have been rendered stable, or #fied#, by the method
of preparation %note &;'.
Shaw#s suspicions were Iustified for, by the middle of the century, writers of handbooks on painting and on artists# materials
were no longer so enthusiastic in their praise for the pigment. In the first place, the claim in the &+&, advertisement that the
pigment was stable to alkali was false: the pigment is rapidly decomposed, forming a brown hydrated ferric oide. )his reaction
takes place with lime, thus the pigment cannot be used in fresco %note &='. )he effects of light on the pigment had also become
apparent. (ntoine .ernety, in the practical treatise published with his dictionary of painting in &+2+, wrote that the colour when
used in oil painting tended to become greenish in time %note *,', a change noted by later writers, including Jeinrich StKckel,
who wrote a comprehensive handbook on oil painting in &+=;, and 5ean-3ranLois 9erim:e, the author of one of the most
widely read works on oil painting %note *&'. 1onstant de 9assoul, on the other hand, wrote that the iron in the colour tended to
regenerate and become black %note **'. Fther writers merely described the colour as being impermanent %note *B'. 5ulius
1aesar Ibbetson was a little more informative, writing, #It will not stand very long in the sun.# %note *0'
It had also been noticed that the admiture of white, either as a result of the method of preparation %discussed below' or by the
deliberate addition of white pigment, had some influence on the stability of the pigment. (s the colour was so dark and intense,
it was rarely used without the addition of some white. Cobert Aossie, writing in &+2;, stated that the darker varieties of the
pigment were usually more permanent than the lighter %the variation in colour being caused by the presence or absence of an
#earth# < hydrated alumina < precipitated with the blue during the reaction', although they lacked brightness and were very
purple. Fn the other hand, the lighter, brighter varieties were #etremely subIect to fly, or to turn to a greyish green# %note *2'.
StKkel thought that the addition of flake white to the pigment for use in oil caused no problem, but that the chalk or other
additives present in ordinary lead white went yellow-green. Je also wrote that if the pigment was to be used in oil, it was much
improved by first treating it with concentrated sulphuric acid %note *8'. Fther writers thought that lead white or other white
etender pigments used in its place contributed to the change in colour of the .russian blue %note *+'.
)he curious tendency of .russian blue decolorised by the action of light to regain its colour in the dark is vividly described by
3ranz 3ernbach, the author of a book on oil painting, in &;B0. Javing decorated a rain shelter with pigments mied with copal
varnish and thinned with turpentine, he left it to dry in the sun. Fn his return some time later, he found to his horror that those
areas painted with .russian blue had faded almost completely. Cather than repaint them straightaway, he decided to wait until
the net morning, when he was even more astonished to find the colour restored to its full strength. (fter this eperience, he
covered areas painted with .russian blue with thick paper before leaving them in the sun and had no further trouble. Je does
not report, however, how well the dried colour survived as time went by %note *;'.
)he vacillating behaviour of .russian blue under the influence of light was most systematically demonstrated by >eorge 3ield.
Auring the first "uarter of the nineteenth century, he carried out eperiments on the permanence of pigments, in both
watercolour and oil medium, eposing identical samples to sunlight and #to the foul air... of a privy# for up to five weeks, until a
wash of brasilwood or cochineal lake %both notoriously fugitive colours' had lost its colour$ further identical samples were kept
carefully folded in paper for comparison. .igment samples, many obtained from 7ondon colourmen, were painted out at three
levels of intensity: a thin glaze over white$ a shade of intermediate strength$ and as a thick layer. In oil, samples mied with lead
white were also included. .igments were considered durable in proportion to the loss of colour they displayed in comparison
with the lake pigment standard. Jis samples of .russian blue were turned purple by the damp noious air, and faded in the sun
to varying degrees, regaining their colour when removed from the testing conditions %note *='. In these particular tests 3ield did
not comment on the tendency of the colour to go green, but in -Chr omat ogr a-h)1 or2 "reat i se on Col our s and
#i gment s and o* t hei r #o!er s i n #ai nt i ng 2 published in &;B2, he stated that it sometimes became greenish, a
change he attributed to the development of a yellow iron oide, and that its tints faded under the action of strong light. It was
also, he wrote, a fluctuating pigment %note B,'.
"arly methods of $re$aration and the nature of the $igment
3ield#s eperimental conditions appear rather less severe than modern testing conditions, yet the modern pigment seems to
withstand these latter reasonably well %note B&'. It is necessary to ask, therefore, to what etent the modern pigment differs
from the eighteenth- or early nineteenth-century version.
9odern .russian blue eists in several different varieties, of which the best and most permanent is ferric ferrocyanide %iron%III'
heacyanoferrate%II'', 3e0M3e%1/'8NB.J*F %where is &0<&8', so-called insoluble .russian blue. It is difficult to obtain pure,
although it can be obtained %even in crystalline form' by the reaction between ferrous hydrogen cyanide, J0M3e%1/'8N and ferric
chloride in hydrochloric acid. )he common method of producing the pigment by miing ecess ferric chloride solution, or
another iron%III' salt, with potassium ferrocyanide %or, strictly speaking, potassium heacyanoferrate%II', D03e%1/8'.BJ*F' gives
a voluminous, deep blue precipitate or colloidal suspension. )he particle size of the product is etremely fine$ conse"uently, if
precipitation is rapid, as is usually the case, variable amounts of potassium are incorporated in the comple and the product is
a potassium iron%III' heacyanoferrate%II'. So-called soluble .russian blue %supposedly the least permanent variety', a
simplified formula for which is D3eM3e%1/'8N.J*F, is produced by adding ferric chloride solution to ecess potassium
ferrocyanide, or by using &:& molar proportions of the two compounds. If this pigment is boiled with ferric chloride solution, the
insoluble form is obtained. )urnbull#s blue, which may also be obtained in soluble and insoluble forms and is reputed to have a
permanence between that of the other two varieties, is obtained by the reaction of an iron%II' salt, such as ferrous sulphate,
3eSF0.+J*F, with potassium ferricyanide %potassium heacyanofer-rate%III', DB3e%1/8''. )he product was thought to be ferrous
ferricyanide, but 9ssbauer spectroscopy has shown that it is identical in structure to the normal ferric ferrocyanide pigment
%note B*'.
/owadays the pigment is manufactured on an industrial scale by an indirect method, based on the reaction between ferrous
sulphate and sodium ferrocyanide in the presence of ammonium sulphate to give a white insoluble product, !erlin white
%ammonium iron%II' heacyanoferrate%II', %/J0'*3eM3e%1/'8'. )his is converted to the ammonium-containing .russian blue by
oidation with sodium bichromate or chlorate. )he potassium-containing blue, although highly valued, is epensive to produce
in bulk %note BB'.
)he typical eighteenth-century method of preparing the pigment was far more curious. #/othing is perhaps more peculiar#, wrote
the 3rench chemist 5ean Jellot in &+8*, #than the process by which one obtains .russian blue, and it must be owned that, if
chance had not taken a hand, a profound theory would be necessary to invent it.# %note B0' )he account of the method sent to
5ohn 6oodward in &+*0 was as follows. (n alkaline salt %essentially potassium carbonate' was prepared by the detonation of
crude potassium tartrate and potassium nitrate. )his was mied with dried blood and heated until the miture ignited. 6hen it
ceased to burn, the black product was etracted with water and filtered off. )he heated filtrate was mied rapidly with hot
solutions of ferrous sulphate and alum %aluminium sulphate, or potassium aluminium sulphate', producing much effervescence
and a greenish precipitate. (fter filtering and washing, the precipitate was treated with hydrochloric acid to give the blue colour
%note B2'.
)he process was modified in various ways in the course of the century. 5ohn !rown demonstrated in &+*0 that, while the iron in
the vitriol was an essential ingredient, the alum %which was later recognised as also serving to neutralise ecess alkali'
appeared to serve principally as an etender: as Jellot said, another white, if desired, would do as well %note B8'. Jellot also
pointed out that complete calcination of the blood and rapid miing were e"ually unnecessary %note B+'.

!urnt hoofs or horns
were found to serve almost as well as dried blood$ different alkaline materials were used, including "uick lime$ conditions of
calcination were altered$ the proportions of the various ingredients were varied. (dditional ingredients, such as red cochineal
dyestuff, might be included: the reason for the inclusion of this particular additive was usually to give the pigment a more purple
hue, but it has also some historical interest %note B;'. )he acid, which dissolves out the alumina precipitated during the
reaction, was sometimes omitted %note B='.
!rown commented that it would be interesting to know how so singular a procedure came into being. It seems that the pigment
was first obtained entirely by accident, according to the chemist >eorg 4rnst Stahl, who gave the earliest known account of its
discovery in &+B&. Aiesbach, a !erlin colour maker, was preparing 3lorentine lake from cochineal dyestuff etracted with alum
solution, to which a little ferrous sulphate had been added. Javing run out of alkali with which to precipitate the red lake
pigment, he ac"uired some from the alchemist 5ohann Aippel$ on attempting to concentrate his colour, which appeared rather
pale, he obtained a dark blue. Aippel realised that this unepected result must have been caused by contamination of the alkali,
which had previously been used in the repeated distillation of the #animal oil# for which he was celebrated %note 0,'.
6hile it was recognised that the animal matter < be it Aippel#s #animal oil#, burnt hoofs or dried blood < played an essential part
in the reaction, the nature of its contribution was not understood: in the &+*,s chemistry as a theoretical science was
insufficiently advanced. )he animal matter was the source of the cyano- or cyanoferrate group, the -1/- or -M3e%1/'8N-part of
the pigment molecule, but at that time the understanding of compound formation was in its infancy and nitrogen had yet to be
discovered. .otassium ferrocyanide itself was obtained from .russian blue by the 3rench chemist .ierre-5oseph 9ac"uer in
about &+2*. Je recognised it as containing the #colouring principle#, which he attributed to the presence of phlogiston, that
curious hypothetical substance supposed to be the matter of fire and, by etension, of light, released during combustion %note
0&'. It was not until the &+;,s, however, with the publication of the work of the Swedish chemist 1arl 6ilhelm Scheele %better
known perhaps for his discovery of oygen and chlorine, and for the green copper arsenite pigment named after him' and that
of 1laude-7ouis !erthollet, that some understanding of the structure of potassium ferrocyanide began to be obtained. !oth
recognised that it contained alkali and iron in combination with what came to be called prussic acid and this breakthrough
provided a basis for the research of the nineteenth century %note 0*'. 4ven by &+=2, the author of - #r act i cal "reat i se on
#ai nt i ng i n 3i l Col our s was able to write that #Scheele and !erthollet have thrown much light on this subIect, and have
converted the old term of --hl ogi st i cat ed al 4al i into -#r ussi an aci d' ( not e 53) .
The effects of the method of $re$aration on $igment dura%ility
9anipulation of preparative conditions to obtain a pigment of the desired "uality was thus based on trial and error for many
years. (s the process was relatively comple there were many possible variations that could be made and, as the discussion of
the chemistry of the modern pigment has shown, different products could be obtained in conse"uence. It is very unlikely that
the product was as chemically pure as would be epected today. 4ven the conditions of the precipitation itself < the
temperature at which it takes place, the dilution of the reacting solutions and the speed of miing < have an effect on the nature
of the final product, particularly on physical properties such as particle size and colour. Fbtaining a good "uality .russian blue
was by no means straightforward. )he problems described by one >erman manufacturer, 5acob 6eber, at the end of the
eighteenth century included incorrect calcination and a greyish, sulphur-containing coating forming on the final pigment, due, he
thought, to impurities in the alkali %note 00'.
It is possible also that certain of these same variables in the method of preparation might have some influence on the stability
of the final pigment. )he varying potassium content is one possible factor to be considered. )he soluble variety of .russian blue
was described by 1hurch as the least permanent %note 02'.

9:rim:e, writing in &;B,, related solubility of the pigment to
potassium content %note 08', but it was not until &;0; that 6illiamson published a method of obtaining a potassium-free
pigment %note 0+'. (nother important factor could be the presence of other salts or contaminants arising as a result of the
method of preparation. )his was recognised at the time: Aossie commented, for eample, that the cheaper grades of the
pigment, used by decorators, were often contaminated by brown or yellow iron salts, also liable to be produced during the
reaction. In a short while, work painted blue with this low-grade pigment turned olive green %note 0;'. 9any recipes also
emphasised the need for thorough washing of the final pigment.
( far more important factor could be the presence of significant "uantities of alumina, which was not always removed from the
final pigment. )his would serve as an etender, in the same way as adding a "uantity of white pigment to the pure pigment. )he
original announcement of the pigment stated that it was prepared in two grades, one lighter than the other and the darker being
prepared from the lighter %see (ppendi'. Instructions were given in various sources for obtaining a darker blue by dissolving
out the alumina: StKckel thus converted a !erlin blue to .aris blue %note 0='. (lumina or other white pigments commonly used
as etenders, such as white earths and barium sulphate, could also have been added deliberately.
It is interesting that there appears to have been a difference in "uality between 4nglish .russian blues and other 4uropean
ones. (s early as &+*2, >eoffroy commented that the pigment prepared in 7ondon appeared more brilliant than that of !erlin
%note 2,'$ 6eber, after all his trials, was pleased to obtain a colour he reckoned to be as good as the 4nglish %note 2&'. Fne of
the most significant comments, however, is made by !ouvier, writing in &;*+, who described the 4nglish pigment as being of
higher "uality than those made elsewhere. Je claimed that, in watercolour, it lasted better than any other .russian blue, even
when eposed to the sun or mied with white or other colours. It went green or brownish in oil, admittedly, but less so than the
other blues. It was also darker in colour and, when burnt, made a less satisfactory brown %note 2*'. Je thought that the
difference in the pigments was due to different methods of manufacture$ 9:rim:e observed that the darker 4nglish blue
contained less alumina %note 2B'.
Conclusion
3rom the occurrence of faded .russian blue in some eighteenth-century oil paintings, it appears that the admiture of the
pigment with a large "uantity of white is a critical factor in its permanence. Aocumentary sources of the period provide evidence
in support of this, but they also suggest that the .russian blue in use then was, in any case, not a particularly durable pigment.
)he method of preparation was complicated and introduced a number of variable factors which may have had a bearing on the
permanence of the eighteenth-century pigment. )he method of manufacture in use today can be far more rigorously controlled,
and altered at will to obtain a pigment with particular characteristics: colour, particle size and solubility can be modified as
desired.
)he second part of this study %to be published in a future issue of the -"echni cal Bul l et i n ) aims to provide the
eperimental evidence, on the basis of which the durability of eighteenth-century and modern pigments can be compared.
Investigation of the constituents present in samples of .russian blue dating from the eighteenth century onwards may throw
some light on the role played by such variables as the presence of alumina or other etenders, the proportion of potassium or,
linked with this, the solubility of the pigment. )he information from this analysis, together with the results of fading tests on a
range of .russian blues of different dates will, it is hoped, go some way towards eplaining the behaviour of what 3ield so aptly
described as this #fluctuating pigment#.
&c'no(ledgements
I am most grateful to (shok Coy for the information on the occurrences of fading$ to Caymond 6hite for his assistance and
advice on the translation from 7atin$ and to .rof. Ar 4. -7. Cichter, Institut fur )echnologie der 9alerei, Staatliche (kademie der
!ildenden DEnste, Stuttgart, for his kind permission to "uote the results of his work.
Notes and references
1) 5. .lesters, #( preliminary note on the incidence of discolouration of smalt in oil media#, -$t udi es i n Conser &at i on 2 &0,
&=8=, pp. 8*<+0.
*) 3or a general account of the blue pigments available at the end of the seventeenth century see C. A. Jarley, -r t i st s'
#i gment s c6 7890:7;3<+ $t ud) i n Engl i sh 'ocument ar ) $our ces 2 *nd edn., 7ondon &=;*, pp. 0B<+2.
3) #/otitia 1oerulei !erolinensis nuper inventi#, -Mi scel l anea Ber ol i nensi a 2 I, &+&,, pp. B++<;. )he translation of the
7atin original in the (ppendi to the present paper is by 5o Dirby and Caymond 6hite. 3or a discussion of the eact date of the
discovery see Jarley, op. cit., pp. +0<&.
4) #9an hat heute zu )age eine sehr schKne blaue 3arbe in !erlin erfunden, so dem Ultramarin gleich, Ia wohl gar dieselbe
Ebertrift#: 5. 9. 1rKker, -'er !ohl an* =l br ende Mahl er 2 Brd edn., 5ena &+B8$ facsimile reprint, ed. U. Schiessl, 9aander
&=;*, p. &&+$ see also the editor#s comments, pp. GIII-IH. &st edn. published in &+&=$ *nd edn. &+*=.
+) 4. -7. Cichter, Staatliche (kademie der !ildenden DEnste, Stuttgart, private communication.
,) 5. 6oodward, #.raeparatio caerulei .russiaci e >ermania missa ad 5ohannem 6oodward#, -#hi l oso-hi cal
"ransact i ons 2 HHHIII, no. B;&, 5anuary-3ebruary &+*0, pp. &2<&+. .eter Shaw comments that the method was #purchased
as a very valuable Secret, both in 4ngland, >ermany and elsewhere$ but it is now got into several hands#. It is unclear whether
he is referring to the years preceding &+*0, but it is certainly possible: see .. Shaw, -Chemi cal >ect ur es -ubl i c4l ) r ead
at >ondon i n t he )ear s 7?37 and 7?3@ 6 6 6 * or t he i m-r o&ement o* r t s2 "rades and Aat ur al
#hi l oso-h) 2 7ondon &+B0, p. &;*.
-) 7. 3aillant-Aumas, #6atteau. 4tude au 7aboratoire de recherche des 9us:es de 3rance#, -Wat t eau+ t echni Bue
-i ct ur al e et -r obl mes de r est aur at i on 2 ehibition catalogue, Universit: 7ibre de !ruelles, !russels &=;8, pp. 2+<
=. Fn the dating of the painting see 9. 9organ >rasselli and .. Cosenburg, -Wat t eau 78;5:7?@7 2 ehibition catalogue,
6ashingtonO .arisO!erlin &=;0<2, pp. 0B0<&.
.) 4. -7. Cichter, Staatliche (kademie der !ildenden DEste, Stuttgart, private communication. 3or the dating of the painting see
>rasselli and Cosenburg, op. cit., p. B0B.
9) >rasselli and Cosenburg, op. cit., p. 00,. In the same catalogue, see also S. 7. 3isher, #)he 4amination and )reatment of
6atteau#sP -C t al i an Comedi ans' , pp. 082<+, especially p. 4,,)
1/) )he pigment is described as durable in 6insor Q /ewton#s -Aot es on t he Com-osi t i on and #er manence o*
r t i st s' Col our s 2 6ealdstone M&=;&N, pp. &0<&*$ in the -'al er D Ro!ne) Cat al ogue o* Fi ne r t and Gr a-hi cs
Mat er i al s 2 !racknell M&=;8N, pp. ;2, ;+<;, it is classified moderately durable. See also /ote B&.
11) 6insor Q /ewton, op. cit., pp. &&<&*$ C. 9ayer, -"he r t i st ' s %andboo4 o* Mat er i al s and "echni Bues 2 0th.
edn., 7ondon &=;* %&=;+ reprint', pp. ;+<=. )he discussion of the stability of the pigment in (. J. 1hurch, -"he Chemi st r )
o* #ai nt s and #ai nt i ng 2 Brd edn., 7ondon &=,&, pp. *&2<&8, although not modern, is still of interest.
1*) 5. (. Sistino, #3erriferrocyanide .igments: Iron !lue#, -#i gment %andboo4 2 ed. ). 1. .atton, /ew Rork &=+B, Gol. &, pp.
0,0<2. )he figures "uoted refer to enamel formulations using a typical present-day .russian blue %not necessarily that used for
artists# pigments', ammonium ferric ferrocyanide, approimate formula /J03eM3e%1/'8N.
13) 9. Aavies, -Aat i onal Gal l er ) Cat al ogues+ "he Br i t i sh $chool 2 *nd edn., 7ondon &=2=, pp. B8<+.
14) Aavies, op. cit., pp. B0<2, &,;<=. ( brief discussion of the dating of -Mr and Mr s ndr e!s appears in -"he Aat i onal
Gal l er ) Re-or t 2 7ondon &=8*, pp. *2<8.
1+) 9. 7evey, -Aat i onal Gal l er ) Cat al ogues+ "he $e&ent eent h and Ei ght eent h Cent ur ) C t al i an $chool s 2
7ondon &=+& %&=;8 reprint', pp. &;<*8, BB<8. See also A. !omford and (. Coy, #1analetto#s SGenice: )he 3eastday of S.
Coch#S, -Aat i onal Gal l er ) "echni cal Bul l et i n 2 8, &=;*, pp. 00<B, and their article in this -Bul l et i n 2 pp. B2<0&.
1,) 7evey, op. cit., pp. **;<B&$ -"he Aat i onal Gal l er ) Re-or t 2 7ondon &=+&, pp. 0;<=.
1-) S. 3oister, #)he .ortrait of (leander 9ornauer#, -"he Bur l i ngt on MagaEi ne 2 1HHHIII, September &==&, p. 8&0, note
8. I am grateful to (shok Coy for bringing this point to my attention.
1.) Shaw, op. cit., p. &;*. Jis comments on the pigment are reproduced, slightly edited, in -
, Com-endi um o* Col or s and ot her Mat er i al s used i n t he r t s de-endent on 'esi gn 2 7ondon &+=+, p. &;.
19) )he effect of an alkaline environment on the pigment is discussed in 4. Aayes, -"he Wor 4s o* t he l at e Ed!ar d
'a)es 2 7ondon &;,2, pp. B,0<&.
*/) (. -5. .ernety, -'i ct i onnai r e -or t at i * de -ei nt ur e2 scul -t ur e et gr a&ur e2 a&ec un t r ai t -r at i Bue des
di * * er ent es mani r es de-ei ndr e 2 .aris &+2+, p. ciii.
*1) J. 3. (. StKckel, -#r a4t i sches %andbuch * ur F=nst D l er 2 >ac4i r l i ebhaber und 3ehl * ar benD ust r ei cher 2
*nd edn., /uremburg &+== %facsimile reprint, (ichstetten &=;;', p. B&$ 5. -3. -7. 9:rim:e, -'e l a -ei nt ur e . l ' bui l e 2 .aris
&;B, %facsimile reprint, .uteau &=;&', p. &+;. See also .. C. de 1... 1, -"rai t de l a -ei nt ur e au -ast el 2 .aris &+;;,
pp. &20<&.
**) 1. de 9assoul, - "reat i se on t he r t o* #ai nt i ng and t he Com-osi t i on o* Col our s 2 7ondon &+=+, p. &;*.
*3) 0"he r t o* 'r a!i ng and #ai nt i ng i n Wat er D col our s 2 7ondon &++,, pp. +;<=$ - Com-endi um o*
Col or s 2 op. cit., p. &*2.
*4) 5. 1. Ibbetson, -n cci dence2 or Gamut o* #ai nt i ng i n 3i l and Wat er Col our s 2 7ondon &;,B, p. &;.
*+) C. Aossie, -"he %andmai d t o t he r t s 2 7ondon &+2;, Gol. I, pp. ++<;.
*,) StKckel, op. cit., p. *B.
*-) 0Aou&eauG secr et s des ar t s et mt i er s 2 ed. 9. Aesbri:res, .aris &;&=, Gol. II, pp. =B<0.
*.) 3. H. 3ernbach, -0eber Fennt ni ss und Behandl ung der 3ehl D Far ben 2 9unich &;B0, pp. 00<2.
*9) >. 3ield, - Hour nal o* #r act i cal Essa)s2 EG-er i ment s and EnBui r i es2 7;98 2 ff. B=H, B=v$ -EGam-l es
and necdot es o* #i gment s6 #r act i cal Hour nal 2 7;9I 2 ff. B,,v, B&8<&+. 3ield 9anuscripts, 3ieldO* and 3ieldO8,
photographic copies, 1ourtauld Institute 7ibrary, 7ondon.
3/) >. 3ield, -Chr omat ogr a-h)1 or 2 "reat i se on Col our s and #i gment s and o* t hei r #o!er s i n
#ai nt i ng 2 7ondon &;B2, pp. &&&<&*.
31) Aaler-Cowney state that pigment samples are eposed to sunlight for 2,, hours, between 9ay and September$ for their
testing conditions see the -'al er D Ro!ne) Cat al ogue o* Fi ne n and Gr a-hi cs Mat er i al s 2 op. cit., p. ;;. (
sample of 6insor Q /ewton powder pigment, modern, but of unknown date, painted out as a wash of moderately intense colour
in a"ueous gum arabic and eposed to a constant illumination of &,, ,,, lu, supplied by )horn artificial daylight fluorescent
tubes, for about si months showed no significant colour change. Fn this evidence, one would assume that, indeed, .russian
blue was a reasonably durable pigment under normal circumstances.
3*) (. 7udi, #.russian !lue, an Inorganic 4vergreen#, -Hour nal o* Chemi cal Educat i on 2 2;, &*, &=;&, p. &,&B$ (. >.
Sharpe, -"he Chemi st r ) o* C)ano Com-l eGes o* t he "ransi t i on Met al s 2 7ondon &=+8, pp. &*&<8$ (. 7udi,
#Aescriptive 1hemistry of 9ied-Galence 1ompounds#, -Mi GedD (al ence Com-ounds+ "heor ) and --l i cat i ons i n
Chemi st r )2 #h)si cs2 Geol og) and Bi ol og) 2 ed. A. !. !rown, Aordrecht, 7ondon etc. &=;,, pp. B=<00$ see also
1hurch, op. cit., pp. *&B<&2.
33) Sistino, op. cit., pp. 0,&<8$ /. Jeaton, -3ut l i nes o* #ai nt "echnol og) 2 Brd edn., 7ondon &=0;, pp. &20<;$ 5. S.
Cemington, and 6. 3rancis, -#i gment s+ "hei r Manu* act ur e2 #r o-er t i es and 0se 2 7ondon &=20, pp. =0<=$
-C nor gani c #i gment s+ Manu* act ur i ng #r ocesses 2 ed. 9. J. >utcho, .ark Cidge &=;,, pp. 0BB<=.
34) 5. Jellot, #Sur la preparation du !leu de .russe#, -%i st oi r e de >' cadmi e Ro)al e des $ci ences 2 (nn:e
9A117GI, &+8*, p. 2B: #Cien n#est peut-:tre plus bizarre "ue le proc:d: par le"uel on obtient le bleu de .russe, Q il faut avouer
"ue si le hasard ne s#en est pas m:l:, il a fallu une profonde th:orie pour l#imaginer.#
3+) 6oodward, op. cit., pp. &2<&+. 4nglish translations of this method appear shortly afterwards$ see, for eample, Shaw, op.
cit., pp. &;0<*$ 5. Joofnail, -"he #ai nt er ' s Com-ani on2 or 2 "reat i se on Col our s 2 7ondon &+B;, pp. 2=<8&. 3or
translations into 3rench see, for eample, 4. -3. >eoffroy, #Fbservations sur la preparation du !leu de .russe, ou de !erlin#,
-%i st oi r e de l ' cadmi e Ro)al e des $ci ences 2 (nn:e 9A11HHG, &+*+, pp. &22<+. )he original recipe and an
4nglish translation appear in J. 9. .owell, #)he !eginnings of 1o-ordination 1hemistry#, -#r oceedi ngs o* t he Chemi cal
$oci et ) 2 9arch &=2=, pp. +B<2.
3,) 5. !rown, #Fbservations and 4periments upon the 3oregoing preparation#, -#hi l oso-hi cal "ransact i ons 2 HHHIII, no.
B;&, 5anuary-3ebruary &+*0, pp. &+<*0.
3-) Jellot, op. cit., pp. 28<+.
3.) See, for eample, Joofnail, op. cit., p. 8&$ 1rKker, op. cit. pp. 2B2<8.
39) See, for eample, Aossie, op. cit., pp. +;<;,$ Jellot, op. cit., pp. 2+<=$ StKckel, op. cit., pp. **8<+, *0,$ 9. 7e .ileur
d#(pligny, -"rai t des coul eur s mat r i el l es 2 .aris &++=, pp. B8<;$ 7. 9arcucci, -$aggi o anal i t i coD chi mi co so-r a
i col or i mi ner al i 2 *nd edn., Come &;&8, pp. 20<+.
4/) >. 4. Stahl, -EG-er i ment a2 3bser &at i ones2 ni madD &er si ones2 CCC Aumer o2 ch)mi cae et -h)si cae 2
!erlin &+B&, pp. *;0<B<. )here are points in this account which are open to "uestion or speculation: if it is to be believed, one
must presume that the ferrous sulphate had been added to deepen or #sadden# the colour of the dyestuff. (s far as lake pigment
recipes are concerned, this is not common practice at this date: additives to make the colour more scarlet are far more likely. (
slightly different, but no less bizarre, account is given by the chemist .. 3. )ingry, -"rai t t hor i Bue et -r at i Bue sur "ar t
de * ai r e et d' a--l i Buer l es &er ni s 2 >eneva &;,B, Gol. II, p. 0,. In his version, Aippel, having thrown various waste
alkalis, previously used for the distillation of #animal oil#, into his courtyard, observed to his surprise that several of the paving
stones had become covered with a brilliant blue colour. Je then remembered that he had previously discarded some iron
sulphate residues in the same place and conducted further research into the preparation of the colour. Fne wonders how far
)ingry is being wise after the event in this story. Aippel#s #animal oil# was later found to contain pyridine and other nitrogen-
containing organic compounds.
41) .. -5. 9ac"uer, #4amen chymi"ue du !leu de .russe#, -Mmoi r es de mat hmat i Bue et de -h)si Bue de
l ' cadmi e Ro)al e des $ci ences 2 (nn:e 9A117II, &+28, pp. 80<++
4*) 1. 6. Scheele, -"he Col l ect ed #a-er s o* Car l Wi l hel m $cheel e 2 trs. 7. Aobbin, 7ondon &=B&, pp. *B;<22$ 1.
-7. !erthollet, #9:moire sur l#acide prussi"ue#, -Mmoi r es6 6 6 de l ' cadmi e Ro)al e des $ci ences 2 (nn:e
9A117HHHGII, &+;=, pp. &0;<8*. See also 5. C. .artington, - %i st or ) o* Chemi st r ) 2 Gol. III, 7ondon &=8* %&=+,
reprint' and the references therein. 3or a summary of the early history of the chemistry of .russian blue, see 7. 5. 9. 1oleby, #(
Jistory of .russian !lue#, -nnal s o* $ci ence 2 IG, &=B=, pp. *,8<&&.
43) 0 #r act i cal "reat i se on #ai nt i ng i n 3i l Col our s 2 7ondon &+=2, p. 2B
44) 5. (. 6eber, -Chemi sche Er * ahr ungen be) mei nem und ander en Fabr i 4en i n 'eut schl and 2 /euwied
&+=B, pp. &B0<82, &;=<*,,.
4+) 1hurch, op. cit., pp. *&B<&2.
4,) 9:rim:e, op. cit., p. &+;.
4-) (. 6. 6illiamson, #Fn the !lue 1ompounds of 1yanogen and Iron#, -Memoi r s and #r oceedi ngs o* t he Chemi cal
$oci et ) , . B, &;0;, pp. &*2<0,.
4.) Aossie, op. cit., p. ;*.
49) StKckel, op. cit., p. *0,.
+/) >eoffroy, op. cit., p. &20
+1) 6eber, op. cit., pp. &=2<8.
+*) .. -7. !ouvier, -Manuel des J eunes ar t i st es et amat eur s en -ei nt ur e 2 .aris &;*+$ *nd edn., &;B*, pp. B=<0,,
0=<2*.
+3) 9:riem:e, op. cit., pp. *,B<0.
&$$endi1
'Notitia Coerulei Berolinensis nuper inventi', Miscellanea Berolinensia, 1 (1710), pp. 3778.
/otice of .russian !lue lately invented
.ainters who mi oil with their colours have few that represent blue, and those such that, rightly, they wish for MsomeN more
satisfactory. Ff the common MbluesN, certainly, one does not reIect miture with the oil, but does not last for a long time in the
work: it becomes greenish, somewhat pale, rust-coloured or distinctly dirty. (nother is indeed constant and beautiful enough,
but sandy$ and this defect, which is inconvenient to the fine work of a skilled master, cannot be removed if the space of a year
were to be consumed in grinding it. )he best of all, which ordinarily they call Ultramarine or (zure, which is produced from lapis
lazuli, not only deters many by its price, but also does not freely permit the addition of other colours$ and this last is able to
show its beauty only in the light parts: it is useless K , i n ut i l i s L ( not e 7) in the shadows.
)his Iust desire of craftsmen, the blue colour, which was invented here in !erlin several years ago and MwhichN now, after
different, most careful considerations, appears boldly in the public eye, can certainly alleviate, if not satisfy. 3or instance, it has
none of these inconveniences. In oil it shows its chief splendour. In truth everywhere, in water, oil and other li"uids with which
one can paint, it is found to be durable in the highest degree. )hat nitric acid K , aBua * or t i s L 2 as it is called by chemists,
which erodes or dissolves everything, does not change it so much, far less destroy it, as render it brighter. (nd as certain blue
colours can be employed in encaustic painting ( , $chmel E!er c4M mai l ) ( not e @) and, so it is said, are not destroyed by
drying by fire, so this new colour strongly resists caustic li"uids K , i gni humi do L ( not e 3) %which name those acids and all
destructive agents can rightly be given'$ which difficulty it, out of all types of colours, can endure. )his McolourN fears even less
the more common investigations of painters: for eample, the Iuice s"ueezed from citrus fruits, etc. It is not changed MwhenN in
place K , i n l oci L ( not e 5) either by the air or by change in the weather. It survives in "uick lime and in that white which
distinguishes gold from a gem %note 2'. 5ust as it was formed out of the most finely divided materials, so it can be ground into
the finest powder. (nd where it is not sufficiently reduced by the first stage Min the processN, it may be completed by a second or
a third$ however, by these separate stages having thus been made dry, it can be wetted again by pure water. Indeed, for that
reason M-eoPN %note 8', by these means only, it is usual to complete this work in that art of painting in the smallest degree, which
they call miniature, the more skilfully, as they wish. Fr other painters, who mi colours on a board, can reduce it with a knife
with little trouble. !ecause of this MaforementionedN fineness, it covers wonderfully areas that are spread with a brush and it can
be etended by other colours. !esides, not only is it able to be covered over by less obscuring colours of its kind and is able to
set off things that stand out more, but also it can shade in folds, hollows and the greater depths of the picture.
)his McolourN is of two kinds: the one dark and more suited to epressing shadow$ the other light, not mied with ceruse or with
another white, but being in this state at once, at its first appearance %note +'. )he darker type of this colour is made from the
lighter and becomes as if reduced, or, as they say, becomes concentrated. .ainters of lesser kinds, who customarily measure
MoutN this colour for the purpose of common miing, rarely desire the lighter variety: MtheyN search for the darker alone and,
miing white with it, add light by degrees. )hose whose eyes are more discerning in fact immediately perceive MthatN a colour,
which rises to a lighter grade out of its darkness by the admiture of white, has not produced so much light and attractiveness
in its being as that which /ature has produced in a weaker form.
3or the rest, it is harmless: nothing here is arsenical$ nothing contrary to health, but rather a medicine. 6ithout danger, those
things which are made from sugar can be painted in this colour and eaten. !eginners of painting can safely draw through their
lips the brushes with which they spread their panels$ which they will do with other paints not without danger to life. )he price,
finally, by which it is procured, scarcely approaches a tenth part of Mthat ofN the dearest Ultramarine. (nd the supply of this,
which in !erlin can be had at the !ookseller of the Coyal Society of Sciences, is as much as the prodigal hand of craftsmen can
desire for the purposes of adorning their work.
Notes
1) 0C n ut i l i s makes very little sense. It seems probable that this is a misprint for -i nut i l i s 2 #useless#.
*) 0$chmel E!er c4 and -emai l both mean enamel. )his clause should be taken to refer to blue pigments that are not
destroyed by heating in a general sense.
3) 7iterally, #li"uid fire#. #1austic# here signifies li"uids that produce a burning, but not necessarily alkaline, effect.
4) #In position# or #in places#, i.e., when the pigment is painted out in its place on the picture. )he word -l oci may be a misprint
for -l oci s 2 or possibly -l oco .
+) It is not clear what this white may be.
,) )he word -eo 2 in the locative and ablative cases, can be used as an adverbial case, referring to a cause or reason given
previously, in this case the fact that the pigment wets easily$ this means that it can be used skilfully in miniature painting. Fne
could infer from these two sentences that, in watercolour, washes are easily made from the pigment.
-) In other words, it appears as a lighter colour as manufactured.
Series 4ditor: (shok Coy
T /ational >allery .ublications 7imited &==B
(ll rights reserved. /o part of this publication may be transmitted in any form or by any means, electronic, or mechanical, including photocopy,
recording, or any information storage and retrieval system, without the prior permission in writing from the publisher.
3irst published in >reat !ritain in &==B by /ational >allery .ublications 7imited 2O8 .all 9all 4ast, 7ondon S6&R 2!(.
!ritish 7ibrary 1ataloguing-in-.ublication Aata ( catalogue record for this Iournal is available from the !ritish 7ibrary.
IS!/ & ;2+,= ,BB ,
ISS/ ,&0, +0B,
4dited by Aiana Aavies and 4mma Shackleton
)he publication of this volume of the -Aational Galler) "echnical Bulletin has been made possible by
the generous support of 9r and 9rs 3rank Cichardson of /ew Rork