Munch S Paintings

Introduction
EdMunch-01-Plahter-v6.indd 1 09/10/2015 14:29

Munch’s paintings: scientific
research both recent and in
retrospect
Unn Plahter1 and Leif Einar Plahter2
1
Professor Emeritus, Department of Conservation, Museum of Cultural History (KHM),
University of Oslo, Norway
2
Former Head of Conservation, National Museum of Art, Architecture and Design, Oslo, Norway
ABSTRACT This paper presents a history of research into the materials and techniques of the Norwegian
painter Edvard Munch, research that began in Norway in the second half of the 20th century and which
continues to the present. The study concentrates on investigations of paintings in the National Museum of Art,
Architecture and Design, the Munch Museum and the Aula paintings in the Assembly Hall of the University
of Oslo. In the 1950s and 1960s, analytical techniques available in larger Norwegian museums were based
primarily on photographic methods such as X-radiography, infrared reflectography and ultraviolet radiation.
Surface microscopy, wet-chemical analyses and examination of cross-sections were also used, but in those early
days scientific investigations of Munch’s paintings were devoted mainly to how he actually made his paintings
rather than the materials used to make them. It is only recently in the last decade that research on Munch’s
materials has intensified.
Research initiated in the 1970s manner (Figure 2), the other three paintings first sub-
jected to analysis by paintings conservator Leif Einar
Introduction Plahter were all painted with opaque oil paints; here
X-radiographic studies played an essential part in
The paintings by Edvard Munch were frequently made their technical investigations (Figures 3–5).
in a straightforward manner: the composition was set The technical studies carried out by Leif Einar
out directly with a brush or a line drawing. The differ- Plahter at the National Museum of The Sick Child
ent stages of the production can often be studied (1886), Spring (1889) and The Portrait of the Author
successfully using the techniques of surface micro- Hans Jæger (1889) were all based predominantly on
scopy and infrared reflectography (IRR). Equally the X-radiographic imaging (Figure 1). Surface micro-
development of the composition of many of Munch’s scopy and IRR were used for the examination of
motifs can also be followed from preliminary draw- Death in the Sickroom (Figure 2b). These investiga-
ings and sketches. However, while Death in the tions, which began in the 1970s, were published in
Sickroom (1893) was painted in the straightforward 1974, 1994 and 1999.1

U N N PLAHTE R AN D LE I F E I NAR PLAHTE R
a c
Figure 1 (a) The Sick Child (1886); (b) Spring (1889); (c) The Portrait of the Author Hans Jæger
(1889). (Photos: Jacques Lathion, before 1980 © National Museum of Norway)
Early pigment analyses of Munch’s painting mate- included in this paper and is also integrated into the
rials focused on paint samples taken in 1974 from papers by conservation scientist Jennifer Mass et al.4
the unsigned and undated The Scream (1910?) at Date, provenance, Gerd Woll’s catalogue number
the Munch Museum (hereafter MM) and from The and size of Munch’s motifs dealt with in this article
Scream (1893) at the National Museum of Norway are given in Table 2.5
(hereafter NM). The aim of this investigation was to
determine whether pigment characterisation might
contribute to a better understanding of the date of
the Munch Museum version. Munch’s Set Design for Death in the Sickroom: surface
Ibsen’s Ghosts (1906),2 close to a probable date for examination and infrared
The Scream at the same museum, was also included reflectography
in this investigation.3 All these samples were re-
examined more recently, in the 1990s and 2000s, There are two painted versions of Death in the
and the results published in 2011. The latest research Sickroom, both of approximately the same size and
carried out by conservation scientist Unn Plahter is both painted in the same year, 1893: one belongs

M U N C H ’ S PA I N T I N G S : S C I E N T I F I C R E S E A R C H B O T H R E C E N T A N D I N R E T R O S P E C T
a b
Figure 2 (a) Death in the Sickroom (1893); (b) infrared reflectogram of Death in the Sickroom. (Photos: Jacques Lathion, before
1980 © National Museum of Norway)
a b
Figure 3 The Sick Child: (a) X-radiograph; (b) tracings show possible changes in the composition: the main lines in the finished
painting are indicated with thin lines; thick lines belong to previous stages. (Photo: Leif Einar Plahter, 1974 © National Museum
of Norway)
to the collection of the NM, and the other to the This painted drawing has a free, sketchy and spon-
MM. The motif is based on Munch’s memories of taneous character; numerous pentimenti are visible
his sister Sophie’s death in 1877. The seven figures to the naked eye. Imaging with IRR allowed them
in the room have all been identified as members of to be seen yet more clearly and also revealed many
Munch’s family. more hidden examples (Figure 2b). The presence of
The NM version is executed on canvas in a mixed extensive pentimenti indicates that much of the crea-
technique.6 The canvas is covered with a greyish- tive process took place directly on the greyish-white
white ground upon which Munch drew the main ground. Leif Einar Plahter’s publication of the infra-
lines of the composition with a dark, fluid paint. red images in 1994 supported the opinion expressed

a b
Figure 4 The Portrait of the Author Hans Jæger: (a) X-radiograph; (b) perspective lines illustrate changes in composition in the
area encompassing his lap and the tabletop. (Photo and illustration: Leif Einar Plahter, 1974 © National Museum of Norway)
earlier by Reinhold Heller that the painting at the preliminary drawings and sketches are known to sur-
National Gallery (since 2003 part of NM) is the ear- vive for many of Munch’s paintings, none is known
lier version of the two.7 While Heller’s observations for these works. Therefore X-radiographs have been
had been based on the appearance of the painting essential in the process of clarifying the genesis of
alone, Plahter’s interpretation of the IRR reinforced the composition of these paintings.
the argument by revealing the complexity of the pre- In The Sick Child, perhaps the most significant
viously unknown pentimenti, then newly discovered. development revealed was that the mother, now lean-
ing forward, was originally seated in a significantly
more upright posture while the position of the child,
the chair and her pillow remained the same through-
The Sick Child, Spring and The out (Figure 3). In 1886 after this major change had
Portrait of the Author Hans Jæger: been executed, The Sick Child was exhibited at The
X-radiographic examination Autumn Exhibition in Christiana (from 1925, Oslo).
Subsequently, the area in the right-hand side of the
The three paintings examined by Leif Einar painting was further reworked until its final stage
Plahter – The Sick Child, Spring and The Portrait of was achieved in 1896, the year that a ‘copy’ was
the Author Hans Jæger – are executed with opaque made for the art collector Olaf Schou. Schou’s paint-
oil paints on canvas. Analyses of X-radiographic ing matches the present state of the original version,
images have made significant contributions to the thus confirming its completion in 1896.8
understanding of how these paintings were reworked The Portrait of the Author Hans Jæger, with some
during their creation. X-ray imaging also revealed adjustments, was finished in 1889. X-radiographic
how the early stages in the development were sub- study of this work reveals that an area across Jæger’s
sequently hidden by new layers of paint. Although knees was repainted and that the angle of the tabletop

Figure 5 Spring: (a) X radiograph; (b) tracings in thick lines belong to previous stages in the composition; (c) tracings of a
possible earlier stage in the composition. (Photo and illustration: Leif Einar Plahter, 1974 © National Museum of Norway)
a b c
Figure 6 (a) Study for a Portrait (Betzy Nilsen) (1887). (Photo: Jacques Lathion, before 1980) (b) X radiograph; (c) detail of
X-radiographic image with the signature in the lower left corner showing the incised signature E M 87. (Photo: Leif Einar
Plahter, 1974 © National Museum of Norway)
was changed from a ‘correct’ geometric perspective Based on this, one might suggest that The Sick Child
to its current angle. The edge of the tabletop was may have functioned in part as a preliminary study
now given a vanishing point higher up. By changing for Spring.10
the angle, the tabletop is seen from a higher point of Study for a Portrait (Betzy Nilsen) from 1887, also
view than the rest of the painting (Figure 4). Effects oil on canvas, was formerly believed to have been a
of ‘incorrect’ perspective such as this are commonly preliminary study for The Sick Child.11 X-radiographic
employed by Munch and may be seen as a means of imaging carried out in the 1990s, however, dem-
giving the observer the feeling of being drawn into onstrated that the portrait was in fact painted on
the picture.9 a reused piece of canvas cut from a larger painting
The X-ray image of Spring, also completed in 1889, that was signed and dated to 1887 (Figure 6), hence
shows that at earlier stages the composition was full the portrait was painted at least a year after The Sick
of details such as a tabletop and flasks as well as two Child was completed. At some unknown date it was
figures to the right of the child (Figure 5). Most sur- lined and put on a stretcher. A mismatch between
prisingly, however, Munch had originally placed the the edging of the original canvas and the new lining
head of the child in the same position as seen in The can be observed. Unfortunately the study of Betzy
Sick Child but then adjusted the head and the pillow Nilsen was stolen in April 1994 and has never been
to the configuration now seen in the final stage. recovered.

a b
Figure 7 Photocopies of The Sick Child obtained from the glass negative taken in Berlin 1892/93. (a) A photocopy from Munch’s
photo album (Munch Museum). Scratches in the paper copy appear as white patches. (b) A newly made photocopy from the
same glass negative as the previous paper copy (Norwegian Museum of Science and Technology). (Courtesy of the Munch
Museum and the Norwegian Museum of Science and Technology)
The Sick Child: interpretation of wrote in fact: ‘I discovered furthermore that my own
early photographic prints eyelashes had contributed to the visual expression ...
I therefore included a hint of them as shadows on
The appearance of The Sick Child was documented the surface of the picture.’15 A free search on ‘tears
in the winter of 1892/3 when a glass negative was and trembling eyelashes’ in Edvard Munch’s Writings
obtained from a photography session undertaken at (English edition) digital archive gave no hits.16 The
The Berlin Exhibition at the Equitable Palast.12 It pro- origin of this statement may therefore seem to derive
vides an early image of the painting before Munch from Eggum’s interpretation of Munch’s own text
completed it in 1896 thus it is a rare and valuable where he mentioned eyelashes and shadows, but no
document attesting to the development of this mas- ‘tears and trembling eyelashes’.
terpiece. Subsequently prints were made from the According to Eggum, stripes and diffuse areas
negative, which itself was lost temporarily. It was are no longer visible and therefore he claimed
deposited in the archives of the Teknisk Museum that Munch himself ‘overpainted’ these areas after
(Oslo) in 1934 until its ‘rediscovery’ in 2005.13 It is the painting was exhibited and photographed in
this early print, a paper copy made from the glass 1892/93.17 This ‘overpainting’ theory was refuted by
negative (Figure 7a), presumably obtained between Leif Einar Plahter in his publication in 199218 where
1893 and 1896, which has been the source of much he explained that the diffuse areas were due to poor
speculation about the development of the painting. darkroom work and not related to properties of the
Therefore, based on diffuse areas across the paint- painting (Figure 7a).19 A newly made print taken
ing that can be observed in this historical print of from the recovered glass negative confirmed that it
1892/93, in 1978, Arne Eggum suggested that the was in fact well focused.20 Based on a study of the old
painting originally had a significantly stronger avant photographic documentation and the painting itself,
garde character than can be seen today. He also called Plahter noted that the ‘stripes’ seen in the historical
attention to vertical stripes of ‘paint’ visible in the print are indeed still visible in the painting today.
historical print, ascribing these to a sensation that Plahter’s article also included comments on
Munch had when looking out according to Eggum the public reaction to The Sick Child when it was
through tears and trembling eyelashes.14 Munch exhibited in Christiania in 1886. The painting was

Figure 8 The Sick Child: ultraviolet fluorescent image showing the fluorescent varnish
which covers the child and the forward-bending mother. (Photo: Trond E. Aslaksby, 2008
© National Museum of Norway)
regarded as unfinished, which caused general indig- streaks visible today appear where the paint remains
nation and led to articles expressing various views uncovered with varnish while areas with varnish
on the picture’s condition and the manner in which may appear to be slightly opaque and light in hue.
it was executed.21 An ultraviolet (UV) image recently The present authors now suggest that the runny var-
published by Trond Aslaksby has confirmed that a nish was perhaps originally transparent when freshly
varnish was applied after the final position of the applied and appeared as dark streaks at the exhibi-
child and mother was established, but before the tion. Possibly loss of solvent over time gave rise to
placement of curtain, table and glass was adjusted voids and light scattering and a rather pale appear-
(Figure 8).22 Analyses by Brian Singer have identified ance seen on the historical print as well as on the
the varnish as an un-pigmented compound com- painting today. Thus the fresh runny varnish might
posed of pine resin with a small addition of linseed have been more visible than it is today or even when
oil.23 The UV image of the painting indicates that the it was photographed in 1892/93.24
varnish was applied carelessly; it is seen to have been When The Sick Child was exhibited in 1886 the
runny and was allowed to flow down the surface of whole pictorial area was probably covered with the
the painting. The vertical streaks visible on the UV runny and carelessly applied varnish, thus giving rise
image are now faintly visible to the naked eye, but as to Jæger’s comments. Based on the evidence pro-
Leif Einar Plahter and Trond Aslaksby have argued, vided by the photograph of 1892/3, the reworking of
some striped effect must have been observed by the curtain, table and glass had not yet taken place
Jæger. Aslaksby has suggested that the varnish may (Figure 7b) but was undertaken between 1893 and
possibly have accounted for these observed ‘stripes’, 1896, possibly as late as spring 1896, according to
which may have been fairly prominent when it was Aslaksby. The version made for Schou provides the
freshly applied at the time of the exhibition. Dark terminus ante quem for any changes.25

The Sick Child: versions, replicas or of the prints also meticulously matches the design in
repetitions the painted versions. A copy of the draft could have
been made either from the original painting or from
From 1896 to 1927 Munch made five additional ver- Schou’s version as both were in Paris that year.34 Woll
sions of The Sick Child.26 In order to compare the extent also maintained recently that the lithograph of Jæger
to which the original painting matches his later versions, made in 1896/97 might also have been copied from
digitised images of the composition in all six variants the painted version of 1889.35 This also applies to the
of The Sick Child were rendered in the same scale and lithograph of 1943/44.36 Digitised images of the paint-
at the same resolution.27 By placing semi-transparent ing and the lithographs show that the draft of the
images of the later versions over the original one, the printed heads matches the draft of the painted head
match between the contours of the original composi- of Jæger extremely well. In this case ‘a true copy of
tion and those in his ‘copies’ proved to be astonishingly the part with the head’ in the correct orientation was
exact in all details. Thus although the painting process made from the painted version. In addition inter-
in all five repeated versions differ, they were ultimately nal measurements of the paintings match equivalent
based on an exact copy of the composition as outlined measurements of the lithographs, confirming that the
in the original painting. The match is so precise that draft of the heads in the painting is repeated exactly
one may suppose that some transfer device had been 1:1 in the lithographic renderings.37
used. Transfers made by ‘eyeballing’ or the use of grid Woll described how the knowledge of transmit-
constructions would probably have resulted in less ting a motif from one medium to another was an
exactitude as well as being rather tedious to execute. important part of the visual artist’s ‘trick’. She found,
A simple device such as a pantograph used for copying however, that Munch himself claimed that he never
and scaling drawings would have ensured the accuracy ‘copied’ his own works but rather engaged in further
observed but would have required the original paint- development of his motifs.38 Jay Clarke has noted that
ing or an exact copy of it to be physically available. Munch referred to his multiple versions as ‘experi-
Apparently, as will be shown, Munch preferred to have mental copies’.39 This holds true for the painting
the original present while he made his later versions. process, but there was nothing experimental about
In 1909 Munch wrote in a letter to Schou that ‘As the draft of the composition: all his versions of The
long as Nørregaard has it [the original], I know that I Sick Child are in fact painstakingly faithful to the orig-
can at any time take it on loan for copying.’28 As the inal composition (not at all ‘experimental’), but they
original painting remained in the possession of Harald are at the same time freely executed in the handling
and Aase Nørregaard from 1903 until 1931, the paint- of the paint. According to Munch: ‘I returned to it
ing was in fact physically available to Munch when he again in 1896 and 1906 ... there I achieved more of
made all his later versions. The original painting was the intense colour I had wanted to give it ... I painted
even brought to Paris where he painted the first ver- 3 versions ... These are all different, each making its
sion for Schou in 1896.29 Likewise, when the two later own contribution in evoking what I felt at the time of
versions were painted at Warnemünde in 1907, the that first impression.’40
original was sent to Munch there.30 The question is: if he did transfer by freehand, why
During his stay in Paris in 1896, Munch also make such an exact version so time-consuming to
made a lithograph version of The Sick Child.31 Woll achieve that was not reflected in the actual painting
has referred to the lithograph as ‘a true copy of the process? His faithful repetition of the draft demon-
part with the head on the pillow as they appear on strates his strong devotion to the composition each
the 1896 painting’.32 The lithographs are rendered in time he painted new versions.
the correct orientation which implies that the draw-
ing was copied to a secondary support before it was
transferred to the stone. Woll wrote that Munch ‘in
several cases has transferred a draft of his motif before Pigment analyses
finishing it on the stone’.33 A comparison between
the digital images of the lithograph and the original As has been discussed, the earliest studies of Munch’s
painted version again reveals that the design drawing paintings concentrated on the use of imaging
10

technologies to better understand his working pro- d-spacings of the partially crystalline hexagonal and
cesses, while analyses of painting materials were rare. cubic cadmium sulphide (CdS), while XRD patterns
The same paint fragments sampled in 1974 from The of the yellow from the water in The Scream (MM)
Scream paintings were repeatedly analysed with new and wall in Ghosts (MM) revealed cadmium car-
instruments over the years. Beginning with X-ray dif- bonate (CdCO3) as a major component along with a
fraction (XRD) in 1974,41 further work subsequently small proportion of CdS.47 The amorphous character
focused on investigation of cadmium yellow using indicated by the diffuse character of the d-spacing
energy dispersive X-ray (EDX) analysis in the scan- of CdS showed that the cadmium yellow was not
ning electron microscope (SEM) in the 1990s and sufficiently calcined. Calcination involves heating to
finally Fourier transform infrared (FTIR) spectro- 600–800°C in this case in an atmosphere of sulphur;
scopy in 2006. Thereafter non-invasive analyses with this manner of production provides a crystalline and
portable X-ray fluorescence (pXRF) analyses were more stable product, but as temperature control
carried out on the actual painting.42 These results could be challenging in those early years calcination
were published in 2011.43 was not always carried out.48 Surprisingly the results
Within the last decade, Brian Singer et al. con- thus suggested a closer resemblance between The
ducted a broad study of Munch’s pigments and Scream (MM) and the 1906 painting than between
media, including investigation of new samples taken the two painted versions of The Scream (Figure 9).
from the two versions of The Scream (MM and NM), In the 1990s SEM-EDX analyses were carried
which was published in 2010.44 Since June 2013, out on the fragments of paint un-embedded and
Jennifer Mass has conducted non-destructive analy- mounted on carbon tape.49 Analyses of yellow sam-
ses of The Scream (MM) with special attention to pled from The Scream (MM) confirmed the presence
cadmium yellow.45 Recent research on the NM and of a cadmium-based yellow pigment containing a low
MM versions carried out by Unn Plahter is pre- proportion of sulphur which agrees with the high
sented below. proportion of CdCO3 revealed by XRD (Figure 10).
Other significant analyses that were not previ- EDX analyses of yellow sampled from The Scream
ously published include those of 1999: eight paint (NM) also confirmed use of cadmium-based yellow
samples from seven paintings executed by Munch in but, in this case, equal atomic percentage amounts of
1920.46 The aim of the study was to elucidate whether Cd and S were observed (Figure 11).
Munch used titanium white at this early date, when In the backscattered image (BSI) of the cadmium-
it is reported to have been in production in Norway based yellow pigment of The Scream (MM), the
(thought to be one of the earliest known instances of presence of crystal fragments close to and less than
its manufacture). Based on study of cross-sections 10 µm long can be observed. According to XRD anal-
and SEM-EDX analyses of the samples, the research yses they have been identified as otavite (CdCO3).
concluded that titanium was not present in either the Otavite is noted for having a characteristic ada-
paint or preparation layers of any of the paintings. mantine (pearly) lustre, a phenomenon recognised
as ‘glittering’ particles in paint fragments (Figure
10a). The smooth surface of these otavite fragments
(Figure 10c) suggests that in the end product otavite
The Scream: pigment analyses functions as a substrate for the cadmium sulphide-
based pigment. It should clearly not be identified
In 1974, XRD analyses of the fragments sampled as a later result of photodegradation as degrada-
from areas of blue, green, yellow and red paint indi- tion products formed in situ would be expected to
cated that the main pigments in all three paintings manifest in large clusters of fine crystals rather than
(as described above) were much the same: ultra- relatively large, scattered, individual crystals.
marine for the blue, hydrated chrome oxide for The open structure of paint film observed (Figure
the green and vermilion for the red. However, one 10c) suggests that the yellow paint is rather lean. This
important discrepancy was found: XRD patterns may to some extent be responsible for the serious
of the cadmium yellow paint from the railing of flaking seen in the close-up of the surface (see Figure
the road in The Scream (NM) revealed only diffuse 16c) where the underbound paint has been applied
11

a b
Figure 9 (a) X-ray powder patterns of Cd yellow to the left. The upper diagram is obtained from yellow paint on the railing
of The Scream (1893) at the National Museum. The strongest d-spacings d(Å): 3.36 (100) for cubic CdS and 3.16 (100) for
hexagonal CdS, are visible and show that the pigment is composed of a mixture of the hexagonal and cubic form. The diffuse
appearance suggests that the pigment is only partially crystalline. The lower diagram is obtained from yellow paint on the water
of The Scream (1910?) at the Munch Museum. The d-spacings, d(Å): 2.95 (100), 3.78 (80), 2.46 (35), 1.825 (35), 2.066 (25) are
all assigned to otavite and only faint lines that can be ascribed to the cubic and hexagonal structures of CdS (d-spacings not
identified indicate that additional components are present). (Photo: Unn Plahter, 1974) (b) Close-ups of Cd yellow paint to the
right: the upper image is obtained from the railing of The Scream (NM), the lower image from the water in The Scream (MM).
(Photos: National Museum and Munch Museum, after 2000)
in pastose strokes. A more dense yellow paint (Figure the sample (Figure 12b) the content of both sodium
11a) containing cadmium yellow was sampled from (Na) and S indicate the presence of sodium sulphide
The Scream (NM). The BSI of this paint (Figure 11b) (Na2S). Conceivably Na2S served as a synthesis rea-
indicated that here the pigment which is well embed- gent that, due to poor washing, remained in the end
ded in the medium is less well defined in the image product (Figure 12b, Table 1).
than those observed in the sample from MM version. The more recent FTIR analyses also confirmed a
A yellow paint fragment sampled from the water high content of CdCO3 in the cadmium-based yellow
in The Scream (MM) was prepared as a cross-section of The Scream (MM) (Figure 13a). Unexpectedly,
(there was no orientation related to top and bottom however, the FTIR analyses of cadmium yellow
in the resulting cross-section). EDX mapping of of The Scream (NM) (Figure 13b) which, accord-
the section revealed a small content of sulphur (S) ing to the XRD diagram, suggested the presence of
in all areas of the sample, suggesting that CdS is pure partially amorphous CdS, indicated a content
fairly evenly distributed throughout the sample and of a sulphate with large absorptions in the region
not localised in lumps (Figure 12b,c).50 The table 1000–1200 cm–1 assigned to sulphate (SO4–2). The
of results (Table 1) shows generally a low propor- large absorptions in the region 4000–3000 cm–1
tion of S. In points 3 and 5 in the upper left part of suggest a hydrated form. As SEM-EDX analyses of
12

a b d
Figure 10 The Scream (MM): analyses of yellow paint fragments sampled from the
water. (a) Microphotograph in normal light showing particles with a pearly lustre;
(b) secondary image (SEC) (5000×); (c) backscatter image (BSI) (5000×) showing
crystal fragments (grey) of otavite (CdCO3) and particles (white) with HgS; (d) EDS
spectrum of the whole area; (e) EDS of long grey crystal showing a low proportion
of S ascribed to a small content of CdS. (Photos and spectrum: Unn Plahter, during
the 1990s)
a b d
Figure 11 The Scream (NM): analyses of yellow paint fragments sampled from the
railing. (a) Microphotograph in normal light; (b) secondary image (SEC) (5500×);
(c) backscatter image (BSI) (5500×); (d) and (e) two EDS spectra obtained from
fields marked 2 and 4 in the BSI image, both with high levels of S compared to Cd.
(Photos and spectra: Unn Plahter, during the 1990s)
13

b c
Figure 12 Yellow paint sampled from the water in The Scream (MM): (a) section in normal light; (b) EDS mapping where Cd is
blue, S is red and Na is given the colour green; (c) BSI image denoting the analyses (2 to 7) given in the table below with atomic
%. (Images: Unn Plahter, during the 1990s)
Table 1 The Scream (MM): EDS analyses marked in Figure of cadmium sulphate (CdSO4) is probably in an
12c (normalised atomic %).
amorphous state and therefore not detected in the
Spectrum
XRD diagrams. As CdSO4 is hygroscopic it can be
The Scream
expected to be present as an amorphous phase.51
(MM) O Na S Cl Zn Cd
2 57.93 8.11 8.84 1.15 23.97 Examination of both the BSI (Figure 11c) and the
3 37.34 24.52 18.39 4.99 1.41 13.35 close-up photograph (Figure 9b upper image) indi-
4 67.84 3.37 2.66 1.39 24.74 cated that the cadmium sulphate in The Scream
5 76.79 12.50 3.31 7.40 (NM) is not present as a degradation product in
6 65.12 3.23 4.17 1.33 26.16 the form of globules or as a greyish crust such as
7 68.80 0.73 4.12 0.80 25.55 has been reported by Geert Van der Snickt et al.52
However, cadmium sulphate as an amorphous form
could be present as an occluded material trapped in
the somewhat poorly crystallised CdS.53
the NM sample confirmed the presence of only Cd In 2007, The Scream (MM) was examined using
and S, sulphates of cadmium seem highly probable. pXRF, which suggested that a cadmium-based yellow
As d-spacings noted in the XRD analyses were all pigment was likewise present in the pale, faded
ascribed to CdS (hexagonal and cubic), the content yellow strokes in the sky and on the neck of the main
14

a b
Figure 13 FTIR analyses of Cd yellow: (a) spectrum from the yellow paint on the water in The Scream (MM); (b) spectrum
from the yellow paint on the railing in The Scream (NM). (Spectra: Unn Plahter, after 2005)
a b
Figure 14 Detail from the sky in The Scream (MM): (a) image in normal light. pXRF analyses marked with yellow circles
indicate where Cd was registered, circles in black indicate a lack of Cd. (Photo: Biljana Topalova-Casadiego, 2008 and analyses
by Unn Plahter, 2009) (b) UV induced infrared fluorescence revealing the distribution of CdS. (Photo: Jennifer Mass, 2013)
figure (Figures 14–16).54 Based on these findings it of cadmium-based yellow paint on the railing have
was concluded that a cadmium-based yellow used retained a bright yellow hue. In both cases, however,
in The Scream (MM) had indeed faded. This theory it seems likely that the colourless components present
was recently confirmed by work by Jennifer Mass may be identified as partially unconverted synthesis
and her team using pXRF and ultraviolet induced reagents employed in the production of cadmium
infrared fluorescence imaging.55 yellow. Two methods appear to be relevant: a direct
The cadmium-based yellow pigments in both ver- wet method and an indirect wet method.
sions of The Scream contain colourless cadmium The wet method may have been the technique used
compounds in addition to CdS; it is unclear to what to produce the cadmium yellow for the NM Scream;
extent these colourless components represent syn- it is based on the precipitation of cadmium sulphide
thesis regents or degradation products. Obviously a from a solution of a soluble cadmium component
sufficient amount of CdS is present to account for such as cadmium sulphate. Some unconverted cad-
the yellow hue of the water in The Scream (MM), mium sulphate may remain in the end product due
while in the thin off-white strokes in the sky and the to poor washing and to occluded material trapped
neck most of the CdS appears to have turned colour- in the structure of the pigment. The indirect wet
less. On The Scream (NM) the rather pastose strokes method may have been employed in the production
15

a c
Figure 15 Detail from the neck in The Scream (MM): (a) pXRF spectra analysis. (b) High levels of Cd in spots 42 and 43 are
marked with a yellow circle while low levels in spots 37 and 44 are indicated within the black circle and are comparable to the
spectrum obtained from cardboard labelled 32 in (a). (Spectra: Unn Plahter, 2009. Photo: Biljana Topalova-Casadiego 2008) (c) UV
induced infrared fluorescence with highlights revealing the distribution of CdS. (Photo: Jennifer Mass, 2013)
of cadmium-based yellow in the MM version of The In 1908 Alexander Eibner (a German chemist and
Scream of 1910(?) and is based on the precipitation a scholar of art materials) had already claimed that the
of CdS from a dispersion of a sparingly soluble cad- content of CdCO3 could not be due to adulteration
mium component such as CdCO3. With the addition simply because it was just as expensive as the pigment
of a soluble sulphide such as Na2S to a dispersion of itself. However, he maintained that such a mixed prod-
CdCO3 in water, CdS will precipitate on the surface of uct was unstable and apt to lose its intensity of colour.58
the dispersed CdCO3 particles where the Cd-ion con- As already mentioned, pXRF analyses of The Scream
centration is the highest. As paler shades of yellow (MM) revealed that pale off-white strokes in the sky
may appear to have been popular at the time it seems and on the neck contained Cd with a low proportion
possible that a precipitation on the solid surface of of S, and it could only be assumed that the sulphide
CdCO3 could have enabled a controlled precipitation originally present in the cadmium yellow had oxidised
of the CdS and a better chance of obtaining a particu- and formed colourless compounds of cadmium such
lar hue. This would have made CdCO3 an important as a sulphate. Only a vague tinge of yellow can now
ingredient used as a synthesis reagent that remained be seen in the cadmium-containing strokes of paint.
partially unconverted in the end product rather than In the absence of lead white it has been observed that
as a material added as filler. Inge Fiedler et al. noted any degradation of the CdS has not led to the form-
that in 1887 Georg Buchner referred to CdCO3 as ation of PbS causing a brownish decoloration such
an adulterant.56 Interestingly Buchner described as the colour Mass et al. found in their investigation
Cd(OH)2 as a substrate for CdS, which implies that a of Henri Matisse’s Le Bonheur de vivre (1905–06).59
sparingly soluble cadmium compound was used as a Unfortunately, sampling of the paint in the sky or on the
reagent for the indirect method and not as a filler. His neck has not been undertaken and conclusions rely on
statement that cadmium carbonate was added as an pXRF and UV induced infrared fluorescence imaging.
adulterant may simply be due to an imprecise use of It is interesting to note that SEM-EDX analyses
terminology. In 1915, A.H. Church wrote that adul- of yellow sampled from Women Turned towards the
teration of cadmium yellow may include components Sun (1914–16) indicated an atomic ratio of Cd to S
such as various chromates, but did not mention the at 1:1, suggesting that in this case cadmium carbon-
more expensive colourless sparingly soluble cadmium ate is likely to be absent, while a content of cadmium
compounds.57 sulphate cannot be ruled out.60
16

a b c d
Figure 16 The Scream (MM) with images of Cd yellow on the water: (a) detail of the water and the railing in normal light.
(Photo: Biljana Topalova Casadiego, 2008) (b) UV induced infrared fluorescence with highlights revealing the distribution of
CdS. (Photo: Jennifer Mass, 2013) (c) Close-up in visible light; (d) close-up of fluorescence in ultraviolet light, CdS with orange
fluorescence, zinc white with a fluorescence in turquoise. (Photos: Biljana Topalova-Casadiego, 2008)
The Scream: reconstruction white keeping with the contrasting colour tonality Munch
to yellow imaged for the sky.
As mentioned above, pXRF analyses of thin, off-

white strokes in the sky and on the neck of the
central figure of The Scream (MM) suggest the Research since 2005
presence of a cadmium-based yellow pigment that
has faded over time. In Figure 14 yellow circles It is only in the last decade that research on technical
marked 22, 10, 24 and 5 and in Figure 15 a yellow aspects of Munch’s work has increased significantly,
circle marked 42/43 indicates areas in which cad- with publications by both paintings conservators
mium was identified. Circles in black denote areas and conservation scientists alike. A study of Munch’s
where little or no cadmium was found. The UV materials and techniques was published by paint-
induced infrared fluorescence images indicate the ings conservator Biljana Topalova-Casadiego in
presence of cadmium sulphide bonds (Figures 14b, Gerd Woll’s Edvard Munch: Complete Paintings in
15b, 16b). Obviously these bonds still remain in the 2009. Investigations of the Aula paintings headed by
yellow strokes on the water (Figure 16a,c), in the paintings conservator Tine Frøysaker were published
pale yellow strokes in the sky where there is a tinge 2007–2013. A study of The Scream undertaken by
of yellow (Figure 14a), and on the neck and hand Topalova-Casadiego was published in 2008 and some
of the main figure (Figure 15a). The UV induced new studies on the perspective design carried out by
infrared fluorescence is also visible where the pig- Unn Plahter are published in this paper. A contin-
ment still remains in mixtures with other pigments ued study of The Sick Child by paintings conservator
such as the pinkish strokes in the sky and on the Trond Aslaksby was published in 2009. Also of note is
railing (Figures 14 and 16a,b). However, in the off- that over the last decade, investigations of the signifi-
white part of the stroke in the sky extending to the cance of the ‘kill-or-cure’ technique (see below) have
right, pXRF analyses (marked 24 and 5) revealed been undertaken by paintings conservator Mille Stein.
a content of cadmium associated with an almost Conservation scientist Brian Singer and colleagues
complete lack of UV induced infrared fluorescence conducted a broad study of Munch’s pigments and
corresponding to the absence of CdS bonds.61 media, including investigation of new samples taken
Based on these studies a reconstruction of the orig- from the two versions of The Scream (MM and
inal coloration of the painting has been attempted. In NM), which was published in 2010. Since June 2013,
Figures 17a and 17b to the left, the present state is conservation scientist Jennifer Mass and her team
shown: to the right the now off-white stroke in the have conducted non-destructive analyses of The
sky is given a false yellow colour. How yellow this Scream (MM) with special attention to cadmium
stroke in the sky was originally is hard to tell, but yellow. These latter results are presented in papers
an effort is made to visualise an approximate orig- elsewhere in this volume. Recent research on this
inal appearance. A stronger yellow would seem in topic by Unn Plahter is also included.
17

a b
c d
Figure 17 The Scream (MM): the images to the left (a and c) show the painting in its present state; in the images to the right
(b and d) the whitish bleached stroke in the sky is given a false yellow colour. (Photo: Biljana Topalova-Casadiego. False colour:
Unn Plahter)
Munch’s materials and painting Munch used various types of supports and media
technique combinations but oil painting on canvas (both fine
and coarse weaves with or without priming) was the
The most comprehensive study of Munch’s materi- most common. Apart from the period before 1900
als and painting technique, mainly involving surface where the use of substantial opaque layers of paint
examination, was carried out by Topalova-Casadiego was standard, after 1900 the artist exploited the prop-
and published in 2009.62 In addition a select group erties of dilute and fluid translucent paints. In these
was subject to investigations involving sampling paintings the texture and colour of the support played
and scientific analysis.63 Topalova-Casadiego’s essay an important role as did the laying in of the composi-
included information accumulated from published tion, which was set out with a brush or by means of a
material as well as information found in the archives linear drawing and which, like the support, remained
of a number of Norwegian institutions. partially visible in the finished painting.
18

a b c d
Figure 18 Four Aula paintings: (a) Harvesting Women (1916–16); b) Chemistry (1914–16); (c) The Source (1915–16); (d) New
Rays (1914–16). (Photos: Svein Andresen and Sissel de Jong, 2005 © Munch Museum)
Topalova-Casadiego found that Munch did not presence of metal soaps and zinc oxalate that had
create paintings according to a fixed set of rules but formed.65
constantly experimented. His works are character- Four paintings – Chemistry (1914–16), New
ised by the use of a wide variation in the properties Rays (1914–16), Harvesting Women (1915–16) and
of the paint: thickness, texture and the degree of The Source (1915–16) – were examined primarily
gloss and matt surfaces. She noted that it is impos- by means of visual observation and pXRF analy-
sible to discern a simple chronological development ses (Figure 18).66 All four have thinly applied white
in his use of materials and technique. Both variation grounds that are matt, lean and highly absorbent,
in gloss and matt surfaces as well as the use of thin and probably glue-bound. They are composed of
washes and thick pastose application of paint occur zinc white, lead white and chalk. Underdrawing of
throughout his oeuvre. the human figures is applied directly onto the ground
with fine, matt grey or black lines. Subsequently flesh
colours were worked up and thereafter the remain-
ing surrounding areas were painted, although large
Munch’s monumental Aula parts of the ground received no further applications.
paintings Visual assessment suggests that the media of the col-
ours may be based on drying oil. No overall varnish
In 2005, Conservation Studies of the University was found.
of Oslo initiated the Munch Aula Painting Project FTIR analyses were carried out on paint sam-
(MAP), at first a three-year research project headed pled from a light green tone in Men Turned towards
by Tine Frøysaker.64 The Aula (the Assembly Hall the Sun (1914–16) and on paint and ground sam-
of the University of Oslo) houses 11 large-scale oil pled from a purple beam in The Sun (1911).67 The
paintings on canvas created by Munch between resulting spectra of the two paint samples show
1909 and 1916. The project aimed to clarify the absorptions assigned to a drying-oil based medium.
conservation history of the paintings as well as to Spectra of the ground layer revealed a high propor-
analyse their materials and painting techniques and tion of calcite, probably chalk, as well as absorptions
carry out conservation of the works. A challenging assigned to oil. Conceivably the ground was not
part of the task was to develop suitable cleaning bound in oil but visual assessment suggested that oil
methods for these unvarnished paintings. Special was locally absorbed from oil paint when applied to
attention was given to the problems concerning the the surface.
19

Normally a major pigment is adjusted with smaller Topalova-Casadiego maintained that the manner
admixtures of other pigments, but bright contrast- in which The Scream (MM) is painted suggests that
ing colours rather than muted and subdued hues are it was based on a familiar composition, while The
favoured.68 Pigments such as ochre or Naples yellow Scream (NM) is characterised by a more tentative
are absent or hardly used, and brown tones created approach. This is consistent with the present dating
of mixtures of yellow and black or red and green suggested for The Scream (MM). She also noted the
are rare. Coloured areas and lines are applied with a fact that the space occupied by the two men in the
brush in single layers directly onto the bare ground. background of the MM version was left in reserve
Some paints are diluted with a solvent, while others when the railing was laid in, and could perhaps sig-
display a range of impasto from low to high. Surface nify that the composition was set in the mind of the
properties shift from matt to glossy and colours are painter. In the NM version the figures superimpose
applied in various ways: semi-transparent or opaque the painted railing (Figure 19).
to some extent modelled, in some cases quite flat. Interestingly Pål Hougen noted that the figures
Overall the same support, ground and paint- in the background in the 1893 version are wearing
ing technique was found in the four Aula paintings. diplomatic frocks and top hats, while in The Scream
However, recent examination of the seven remaining (MM) the figures are rendered as working class men,
paintings revealed that the preparation layers differ a depiction that is consistent with trends that devel-
significantly within this group, while the materials and oped in the first decades of the 1900s.73 Arne Brenna
techniques of the paintings themselves correspond wrote that indeed in 1909–10, Munch took up the
to those found earlier in the other four paintings.69 working motifs that had flourished in Europe 10–20
Referring to a quote published by Paul Erik Tøjner, years earlier.74 This supports the present dating
Frøysaker observed that Munch had stated that the which is close to 1910.75
Aula paintings ‘are executed in the same way as all
my other easel paintings that are shown in galleries’.70
The Scream: perspective studies76
The Scream: painting technique A recent study of the perspective of the two versions
of The Scream likewise contributes to the supposition
Following the theft and the subsequent recovery, that the MM version is the later of the two. The main
The Scream (MM) underwent a thorough study that focal point in the various renderings of The Scream
was published in 2008.71 This publication includes a and Anxiety lies where the vanishing lines of the
chapter on conservation, materials and technique of three railings converge to a point on the horizon line
the painting as well as a chapter that considers the close to the left edge of the painting where the main
manner in which both the MM and NM versions vertical lies (Figure 20). The horizon line in all ver-
of The Scream were made.72 As Topalova-Casadiego sions of The Scream (Figure 21b,c,d,e) is positioned
noted, there are differences in the way in which the passing near the heads of figures in the background
compositions were set out: in The Scream (MM) the as well as the boats on the water and the peninsula.77
composition was laid in with a line drawing while in Øyvind Storm Bjerke has drawn attention to the fact
The Scream (NM) a painted sketch is seen to have that various portrayals of Anxiety and The Scream all
been executed with a brush and applied in wide spon- have a common origin in a drawing from a sketch-
taneous strokes comprising thin washes of paint. book used around 1889 (Figure 22).78 Evidently the
The Scream (MM) is largely painted in opaque draft of the later images is founded on basically the
paints with pure intense colours; in The Scream same perspective lines that seem to have been estab-
(NM) intense opaque colours are applied to the sky lished already in the drawing of 1889. The position
while most of the rest of the image is rendered in of the figure in the background of the 1889 draw-
subdued mixed tones. The addition of variously col- ing coincides with the location of the background
oured drawn lines in crayon or pastel also features in figures in these later images. The major difference
this work. between the drawing of 1889 and the versions of The
20

a b
Figure 19 Details of The Scream (NM) and The Scream (MM) with figures in the background: (a) The Scream (NM) with figures
wearing diplomatic frocks and top hat overlapping the railing; (b) The Scream (MM) with figures in working-class attire painted
in reserve when the railing was laid. (Photos: Biljana Topalova-Casadiego, 2008)
a b
Figure 20 Perspective lines of the thee railings. (a) The Scream (NM): the vanishing lines (in white) for the railings in front
of the main figure meet above the vanishing lines (in red) for the railings behind the front figure. (b) The Scream (MM): the
vanishing lines of the entire railing meet in one point. (Line drawings: Unn Plahter, 2014)
21

a b c d e
Figure 21 (a) Sketch on the reverse side of The Scream (NM) from 1893; (b) The Scream (NM) from 1893; (c) The Scream
(1893) in pastel and crayon; (d) The Scream (1895) in pastel; (e) The Scream (MM) from 1910(?). In order to interrelate
internal measurements all images are rendered with the same height and resolution. White lines indicate their horizon lines.
(Illustrations: Unn Plahter, 2014)
Scream and Anxiety is found in the position of the the left corner of the pictorial area; in addition the
main vertical. In his later versions Munch obviously angle of the railing was slightly sharpened. In conse-
removed the field to the left of the vertical included quence the area occupied by the road and railing was
in the drawing. reduced while that occupied by the landscape and
Although the location of the horizon line, the ver- sky was increased. In addition the front figure was
tical line as well as the railing is repeated in all the reduced in size and brought further down in relation
portrayals of The Scream and Anxiety, the angle to the position of the new horizon line.79 Basically all
between the vertical line and the railing varies. The later versions of The Scream appear to be based on
angle in the 1889 drawing is the largest, close to 75 the NM versions. The closest resemblance is found
degrees. In the renderings of Anxiety made 1891–92 between The Scream (NM) and the two renderings in
the angle of the railing is sharp, 35–40 degrees, while pastel. In these pictures the ratio between the height
in the versions of The Scream the angle is less sharp and width of the pictorial area corresponds with the
(close to 55 degrees) – in the case of The Scream height and width of the pictorial area of The Scream
(MM) it is almost 60 degrees. (NM), the red strip not included. The pictorial design
Interestingly the angle in the sketch on the reverse in the area on the reverse corresponding to the area
of The Scream (NM) is also close to 55 degrees. One occupied by the red strip on the front was omitted
might assume that this sketch represents the first in the final painting. An attempt to remove the area
version of The Scream and the first trial with an by cutting off the cardboard was made. Subsequently,
angle larger than those in his versions of Anxiety. A however, Munch instead covered the area with red
small painted dot near the left vertical edge (Figure paint. As for The Scream (MM), the ratio between
23) may have functioned as the main focal point the height and width of the cardboard corresponds
and a starting point for the placement of the hori- to the total dimensions of the NM cardboard, i.e.
zon line and the vanishing point for the railing. including the red strip. Therefore the pictorial area in
Munch then sketched the railing possibly with the The Scream (MM) is slightly more elongated than the
same paint as used for the focal point. At this stage earlier versions, and the angle of the railing slightly
items such as background figures, boats and penin- larger, close to 60 degrees, less sharp than the earlier
sula would have been well known to him. When the versions. In addition the front figure is drawn further
sketch was rejected and the final version painted, the to the left and appears slightly more dominant than
main focal point and the horizon line were lowered in the previous versions.
(Figure 21a,b). In Figure 24a the finished painting is Obviously Munch was true to a draft based on a
superimposed by a semi-transparent version of the common perspective drawing applied in all his por-
sketch. The white line follows the upper railing in trayals of Anxiety and The Scream. The painted dot
the sketch while the red line follows the railing in the (the main focal point) on the reverse of The Scream
painting. Thus a lowering of the focal point led to (NM) suggests that he was concerned with perspec-
a parallel displacement of the railing down towards tive lines to guide his composition and did not rely
22

Figure 22 From Ljabruchausseen 1889–90, a drawing from a sketchbook used around 1889
(after Bjerke 2008:18).
Figure 23 The Scream (NM): (a) sketch on its reverse side, the vanishing lines of the three railings
converging to the main focal point. (b) Detail with the main focal point, here a painted dot near
the left vertical edge; (c) a close-up of the same photo showing the painted dot. (Courtesy of the
National Museum of Norway)
on freehand alone. This recent study exemplifies in The Scream (NM) lies slightly above the main
how systematically he based these compositions. focal point (Figure 20). This use of a raised vanishing
Unlike The Scream (MM) however, the vanishing point for the railing in front of a figure leaning over it
point of the three railings in front of the main figure appears to be a common trend in Munch’s images, as
23

Figure 24 The Scream (NM): (a) the photo of the finished painting is superimposed by a semi-transparent version of its reverse
side sketch. The white line follows the upper railing in the sketch while the red line follows the upper railing in the painting.
A parallel displacement of the railing in the painting, down towards the left, is evident. (b) Photo of the sketch on the reverse
side. (Image manipulations: Unn Plahter, 2014)
for instance in his portrayals of Anxiety, but appar- entire pictorial area is covered with paint in various
ently not when he painted The Scream (MM).80 Such colours and hues.81
an incorrect perspective leads to the impression that Like Leif Einar Plahter in 1994, Trond Aslaksby
the railing in front of the main figure is seen from a began his account by citing the indignant reaction of the
higher point of view than the rest of the painting, an public when the painting was exhibited in Christiania
aspect apparently no longer an issue when he painted in 1886.82 He included Jæger’s frequently cited news-
The Scream (MM). The same type of incorrect per- paper account related to the making of the painting
spective can be observed in The Portrait of the Author and referred to the reactions of Jæger and the public
Hans Jæger. In this case the tabletop is seen from a when the painting was shown. Aslaksby, in common
higher level than the rest of the painting (Figure 5). with Plahter, cited Munch’s own statement referring to
his perception of his work and his state of mind during
its making, which the artist recorded 40 years after the
completion of his painting.83 Aslaksby drew attention
The Sick Child to these written sources and to the technical charac-
teristics observed in the painting itself. He found that
In a letter of 1974 to Carl Adam Nordenfalk, Leif it seemed little different from Jæger’s description and
Einar Plahter wrote that The Sick Child is painted suggested that a change in taste is more likely to have
mostly with relatively thick opaque oil colours. The occurred rather than a drastic change in the physical
paint is applied in short strokes, largely modified condition of the painting.84
further on the pictorial surface. The paint was Based on observation of the X-radiograph pro-
worked wet-in-wet and lower levels of paint were duced in 1974, the UV image showing the fluorescent
uncovered where top layers have been scraped varnish and retouching, and the Berlin photograph
off. The painting is meticulously executed and the from 1892–93 as well as careful surface examination,
24

Aslaksby, much in agreement with Leif Einar Plahter’s staining, SEM-EDX, FTIR as well as gas chromato-
previous statement, divided the process of making graphy-mass spectrometry (GC-MS). Pigments
the painting into four phases, roughly as follows:85 identified corresponded to those found in other
examples in contemporary art: cobalt violet, cerulean
1. Rendering of the child, the pillow and with the blue, ultramarine, Prussian blue, viridian, cadmium
mother in an upright position; yellow, yellow chromates and vermilion were all fre-
2. Execution of the mother bent forward and the quently noted. A colouring material identified as
application of a varnish (the painting was in this brass flakes was observed, employed as gold paint in
stage when exhibited in 1886 and photographed the second version of Separation from 189689 and a
in Berlin in 1892–93); lake based on buckthorn berries was found in yellow
3. Partial retouching (by Munch?); paint in The Scream (MM).
4. Alterations of the curtain, table and glass as To some extent it appears that Munch seems
shown in the copy made for Schou in 1896. to have favoured bright contrasting colours. This
applies especially to the Aula paintings investigated
In addition Aslaksby made numerous observations by Tine Frøysaker and Mirjam Liu.90 Within the
linked to the application of various specific strokes group of paintings investigated by Singer et al. ochres
of paint. are identified on The Sick Child and Madonna but
In his earlier article of 1999, Aslaksby found are absent or rare in the remaining paintings studied
evidence for flaking paint at an early stage in the here. On the whole, pigment mixtures found in The
repainted vertical third of the painting to the right Sick Child appear to be more complex than in other
side, visible in the 1892–93 photograph. Aslaksby paintings investigated.
also claimed that a signature and date, ‘E. MUNCH Oil-based media on canvas have been domi-
85’ (in block letters) can be detected in the lower nant throughout Munch’s oeuvre. The Sick Child,
right corner of the Berlin photograph.86 He admit- Madonna (MM) and The Human Mountain are all
ted, however, that the interpretation of this signature painted in drying oils: The Sick Child with linseed oil
is tentative. Leif Einar Plahter has remained uncon- or linseed oil mixed with poppy or possibly walnut
vinced by this claim, and notes that it would have oil; Madonna with heat-bodied linseed oil and mix-
been highly unusual for Munch to have signed in tures of linseed and poppy seed oil as well as linseed
block letters, as this form has not yet been recorded.87 oil mixed with beeswax; The Human Mountain with
heat-bodied linseed oil in paint layers as well as in
the upper ground while white paint contains a mix-
ture of linseed and poppy seed oil.
Analyses of Munch’s painting Analyses of the two painted versions of The Scream
materials on cardboard and the five sketches on canvas, on the
other hand, revealed a wide range of materials and
In 2010 Brian Singer together with Trond Aslaksby, confirm the use of several types of media within a
Biljana Topalova-Casadiego and Eva Storevik painting. On the NM Scream, drying oil-based paint
Tveit published analyses of pigments and binders and media modified with egg or resin was identi-
in a number of paintings by Munch.88 Their arti- fied. In addition painted passages are accentuated
cle included analyses of paintings from the period with streaks of pastel and crayon, reported as wax
1886–1927, the earliest being The Sick Child (1886), crayon, casein pastel and oil pastels with beeswax
the latest The Human Mountain: Towards the Light and in one case Japan wax. Casein pastels were iden-
(1927–29). Special attention was paid to The Sick tified as used in the blue and green landscape. A pink
Child, and the painted versions of The Scream (NM crayon employed in the sky of the NM Scream was
and MM) and Madonna in the Munch Museum. The bound in paraffin wax, stearin wax and pine resin.
investigation also included a study of five sketches on In dark blue strokes in the landscape, a mixture of
canvas at MM that were painted as studies for the protein and gum arabic was recorded. The MM
final Aula paintings. Methods used included polar- Scream is largely painted in drying oil-based paints;
ised light microscopy (PLM), cross-sectioning and the use of both linseed and poppy seed oil is possible,
25

conceivably modified with beeswax or traces of pine the Munch Museum.94 The aim has been to elucidate
resin. Tempera paints or pastels with casein and egg whether Munch deliberately exposed the paintings to
are also found. FTIR and EDX analyses of the lower natural weathering outdoors.95 Many of the paintings
yellow streak in the sky of the MM Scream indicate and painted drafts that Munch bequeathed to the city
a high level of kaolin tinted with lead chromate and of Oslo (the Ekely Collection) were marked by bird
seemingly a very small amount of organic material. droppings and water stains; the paint surfaces were
This suggests that in this case a soft pastel with a low weathered and often showed large areas of flaking col-
content of binder was used.91 ours. Some of the canvases had holes in them and were
rotted along the edges of the stretchers. According to
Rolf Stenersen, the author of the well-known biog-
raphy on Munch, the poor state of preservation was
Puberty due to the fact that the artist deliberately exposed the
paintings outdoors in order to change their appear-
Investigations of the painting technique of the ance.96 Stenersen called this Munch’s hestekur97
MM and NM versions of Puberty were published (literally his ‘horse-cure’) now referred to as the ‘kill-
by Topalova-Casadiego in 2012.92 Munch painted or-cure’ technique. Ever since this idea was raised in
four versions of Puberty: the first version now lost, 1944, it has been an issue of some debate.98
is dated from 1886 whereas the MM and the NM A variety of photographic documentation shows
versions discussed here were both painted in 1894 paintings and painted drafts hanging in Munch’s out-
(Figure 25). While Puberty (MM) has traditionally door studios and a number of visitors have related
been considered as the earlier of the two, Topalova- how Munch allowed the artworks to be treated
Casadiego found that the evidence derived from roughly out in the open. This was apparently due
study of the materials and technique of the painting to the lack of space for indoor storage. However, it
contradicted this theory. In fact neither of the two is not documented that Munch himself ever used
can be viewed as a study or a prototype for the other. the term hestekur or that he explicitly used weather
Pigment identification of Puberty (MM) was based exposure as an artistic tool. Among the researchers
on pXRF analyses and visual assessment.93 The violet on Munch there exists, therefore, divergent views on
on the back wall is composed of a blue pigment, whether or not the weather exposure was an artistic
most probably ultramarine mixed with vermilion. instrument or had other causes. Contradictory views
Ultramarine is also the major pigment in the blue on this matter include those of Sigurd Willoch, who
on the dark blue back wall and outlines on the hand. believed that Munch was merely careless, and Dieter
Pale blue contains ultramarine lightened with zinc Buchhart, who asserted that marks from weather-
white. Dark green in the shade and light green on the ing resulted from a deliberate ‘kill-or-cure’ process
sheet appears to be based on chrome green lightened as part of Munch’s artistic expression and formal
with zinc white and perhaps some lead white. Yellow vocabulary.99
on the sheet may be created by means of a yellow Paintings conservator Jan Thurmann-Moe
toned varnish while in a beige area by the sitter’s foot (Munch Museum) has long argued that the ‘kill-or-
cadmium yellow was recorded. Brown on the front cure’ was a deliberate action from Munch’s hand, and
of the bed contains ochre and a content of manga- therefore traces of weathering left on the paintings
nese (Mn) suggests the use of umber. Flesh appears should be preserved.100 Thurman-Moe’s argumen-
to be based on zinc white and lead white, its tonality tation has been continued by, among others, the
adjusted with various pigments. Danish art historian Tøjner101 and Buchhart, who
believed the result of weather exposure accentuates
the painting’s motif.102 These scholars based their
theories on a very few rather seriously maltreated
The ‘kill-or-cure’ technique paintings from the estate.
More recent research conducted by Mille Stein
Since 2011, Mille Stein has studied the state of preser- has found that the theory on the use of weathering as
vation of the paintings of the Ekely Collection, now in an artistic means is based on the visual assessment of
26

a b
Figure 25 (a) Puberty (1894) in the Munch Museum; (b) Puberty (1894–95) in the National Museum. (Courtesy of the Munch
Museum and the National Museum of Norway)
a limited number of Munch’s paintings and painted the Beach (1921–23).107 In addition, no correlation
drafts from the Ekely Collection, as well as on a series was found between motif and weathering or between
of quotes from his friends and acquaintances and Munch’s duplicates and weathering. Weather marks,
by Munch himself. The conclusion that the ‘kill-or- however, increase with the size of the motif.108 The
cure’ technique was an intentional action is therefore results of the quantitative analysis, combined with a
based largely on circumstantial evidence and a lim- critical assessment of the written sources and exam-
ited group of artworks. ination of Munch’s exhibition policy indicated that
Stein has taken a new methodological approach the signs of weathering on works from his estate
to the ‘kill-or-cure’ suggestion. Her investigation are likely to be due to neglect and not the result of
has involved quantitative analysis of the roughly an artistic action based on a systematic method.109
1000 pictures from the Ekely estate. The analyses Notably, most of the typical ‘kill-or-cure’ pictures
included motifs stained with bird droppings,103 with were not signed or exhibited by Munch.
water stains104 and with severely weathered sur-
faces.105 She concluded that the arbitrary creations
of bird droppings and water stains can hardly have
been used as artistic tools. Moreover, analyses of Conclusions
the degree of weathering showed that only 5% of the
Ekely Collection may be described as having severely Early studies of Munch’s painting technique published
weathered paint surfaces.106 It is within this group in 1974 by Leif Einar Plahter aimed to illuminate
of maltreated works where the typical ‘kill-or-cure’ Munch’s working process: how Munch developed the
works can be found that have suffered heavy paint different stages of a painting rather than analyses of
losses, such as Death and Life (1894) and Dance on the painting materials. These investigations drew on
27

28
Table 2 Motifs by Munch included in this paper.
Dimensions
EdMunch-01-Plahter-v6.indd 28
Title Date (cm) Provenance Woll Selected literature
The Sick Child 1885–86 120 × 118.5 NM 130 L.E. Plahter 1974, 1992, 1999; Aslaksby 1999, 2009; Singer et al. 2010:
NG.M.00839 276−7
Betzy Nilsen 1887 25.5 × 29 NM 144 L.E. Plahter 1992, 1999
NG.M.03054
Previous title: Study for a Portrait
Spring 1889 169.5 × 263.5 NM 173 L.E. Plahter 1974, 1992, 1999
NG.M.00498
Hans Jæger 1889 109 × 84 NM 174 L.E. Plahter 1992, 1999
NG.M.00485
Previous title: Portrait of the

Author Hans Jæger
Ljabruchausseen 1889−1890 23.1 × 30.8 MM T 126-10 Bjerke 2008
Anxiety 1892 92.0 × 86.0 TG
Death in the Sickroom 1893 152.5 × 169.5 NM 329 L.E. Plahter 1992, 1999; U. Plahter 1994
NG.M.00940
Death in the Sickroom 1893 134.5 × 160 MM 330 L.E. Plahter 1992, 1999
M 418
The Scream 1893 91 × 73.5 NM 333 U. Plahter 1975; Singer et al. 2010: 278−82; Plahter and Topalova-
(front and reverse) NG.M.00939 Casadiego 2011: 247−52
The Scream (pastel and crayon) 1893 74 × 76 MM 122b
Puberty 1894 149 × 112 MM 346 Topalova-Casadiego, 2009b, 2012; U. Plahter 2009; Topalova-Casadiego et
M 281 al. (in this vol.)
Puberty 1894–95 151.5 × 110 NM 347 Topalova-Casadiego 2012
NG.M.00807
Anxiety 1894 94 × 74 MM 363 Topalova-Casadiego et al. (in this vol.)
M 515
Madonna 1894 90 × 68.5 MM 365 Singer et al. 2010: 277−9
M 68
The Scream (pastel) 1895 79 × 59 Private collection
Vampire 1895 91 × 109 MM 377 Topalova-Casadiego et al. (in this vol.)
M 679
The Sick Child 1896 121.5 × 118.5 Gothenburg 392 Not analysed
Museum of Art
Separation 1896 96.5 × 127 MM 393 Topalova-Casadiego 2004
M 24
Set Design for Henrik Ibsen’s Ghost 1906 60.5 × 102.2 MM 699 U. Plahter 1975
M 984
09/10/2015 14:29
The Scream 1910 (?) 83.5 × 66 MM 896 Topalova-Casadiego 2008, 2009a; Singer et al. 2010: 282−3; U. Plahter and
M 514 Topalova-Casadiego 2011: 247−52; Mass et al., parts 1 and II (in this vol.)
EdMunch-01-Plahter-v6.indd 29
History (Aula draft) 1910–11 314 × 875 MM 919 Singer et al. 2010: 283−5
M 960 .
Harvesting Women (Aula draft) 1910–11? 490 × 205.5 MM 942 Singer et al. 2010: 283−5
M 915
New Rays (Aula draft) 1910–13 260 × 146 MM 945 Singer et al. 2010: 283−5
M 918
History 1911/ 455 × 1160 UiO Aula 968 Frøysaker (ed.) unpublished report
1914–16
The Sun 1911 455 × 780 UiO Aula 970 FTIR purple beam (sample 5); Liu et al., forthcoming 2015
Alma Mater (Aula draft) 1912–13 253 × 710 MM 1021 Singer et al. 2010: 283−5
M 965
Alma Mater 1916 (1915–16) 455 × 1160 UiO Aula 1220 Frøysaker (ed.) (unpublished report)
Geniuses in Lightstream 1914–16 455 × 305 UiO Aula 1221 Frøysaker (ed.) (unpublished report)
Awakening Men in Lightstream 1914–16 455 × 305 UiO Aula 1222 Frøysaker (ed.) (unpublished report)
Women Turned towards the Sun 1914–16 454 × 165 UiO Aula 1223 Frøysaker (ed.) (unpublished report)
Men Turned towards the Sun 1914–16 455 × 165 UiO Aula 1224 FTIR light green (sample 13)
New Rays 1914–16 455 × 225 UiO Aula 1225 Frøysaker and Liu 2009
The Source 1915–16 455 × 225 UiO Aula 1226 Frøysaker and Liu 2009
Chemistry 1914–16 455 × 225 UiO Aula 1227 Frøysaker and Liu 2009
Harvesting Women 1915–16 455 × 225 UiO Aula 1228 Frøysaker and Liu 2009
The Human Mountain: Sphinx 1927–28 141×x 103 MM 1607 U. Plahter 1999
M 801
The Human Mountain: Kneeling 1927 149.5 × 102.5 MM 1608 U. Plahter 1999
Nude M 804
The Human Mountain: Towards 1927–29 300 × 420 MM 1613 Singer et al. 2010: 285−6
the Light M 978
Naked Figures 1927–29 100 × 72 MM 1616 U. Plahter 1999
M 1021
Previous title: Naked Nudes
Kneeling Female Figure 1927–29 100 × 75 MM 1617 U. Plahter 1999
M 1020
Workers on the Building Site. Draft 1931–33 340 × 570 MM 1685 U. Plahter 1999
for a Decoration of Oslo City Hall M 977A and 977B
Karl Wefring 1934–35 89.5 × 82 MM 1725 U. Plahter 1999
M 461
Alma Mater: Right Part (Aula 1914–16 470 × 396 MM 1816 Singer et al. 2010: 283−5
draft) M 961
29
09/10/2015 14:29
X-radiographic and infrared radiographic imaging. the step-by-step evolution from the drawing in 1889
Gradually, in accordance with the increase in instru- to The Scream in 1893, The Sick Child was developed
mental analyses available to the museum laboratories on the very same canvas, layer by layer.
and the first research performed by Unn Plahter, As the repetitions of the versions of Anxiety and
the scope of analyses of Munch’s painting materials The Scream could be constructed from a common
increased. perspective drawing, the size of the paintings could
As Topalova-Casadiego has summarised, various vary. Large paintings and small sketches could be
types of support were used, but oil painting on canvas made within the same frame. Munch’s repeated ver-
was the most common. Munch’s palette, thoroughly sions of The Sick Child (paintings and lithographs) on
investigated by Singer et al. and Frøysaker and Liu, the other hand, indicate that the draft of the com-
was not extensive. Singer et al. found that Munch position in all cases was based on an exact copy of
was ‘reasonably willing to introduce new materials as the original painting rendered 1:1. Likewise the litho-
they became available, as illustrated by his use of a graphs of the head of the author Hans Jæger were
petroleum-based wax crayon in 1893, soon after their based on a draft copied from the painting also 1:1. In
introduction, his use of oil pastel (oil crayon) perhaps all cases the painterly completions varies. The inten-
as early as 1893 and his use of cadmium red by 1927– sity of the striped effect observed by Jæger when The
29’.110 Non-invasive analyses of a selected number of Sick Child was exhibited in 1886 remains unclear as
paintings by Munch carried out by MOLAB revealed this is less pronounced today. In this paper it has been
no clear evidence for a chronological development hypothesised that resinous varnish upon application
in Munch’s choice of the palette.111 This is consistent was transparent and would have appeared as dark
with Topalova-Casadiego’s findings that a simple and stripes, very unlike how they now look.
linear technical development cannot be observed in The composition of the cadmium yellows identi-
Munch’s oeuvre. fied in The Scream paintings suggests differences in
Recent studies of The Scream and The Sick Child the manufacture of this pigment as well as variations
have shed light on how methodically Munch worked. in the degree of degradation. In both paintings the
Storm Bjerke has pointed out how Munch’s versions of cadmium yellows contain colourless cadmium com-
The Scream and Anxiety all have a common starting ponents: in The Scream (MM) cadmium carbonate
point in the drawing from 1889. Further investigations and in The Scream (NM) cadmium sulphate. These
of these versions, however, show how Munch, with a colourless components may appear to have func-
systematic use of the perspective drawing, created the tioned as starting reagents for pigment production,
versions of Anxiety and The Scream. All these paint- but it is still unclear to what extent they remained as
ings have the main vertical close to the left edge of unconverted constituents in the end product or were
the image and most have common elements such as formed from photodegradation. Obviously a suf-
a foreground figure, figures in the background, boats ficient amount of the cadmium sulphide present in
and the peninsula. The new challenge when making the yellow paint of The Scream (MM) is responsible
The Scream appears in fact to have been limited to for the yellow hue on the water while in the thin off-
the level of the horizon line with the main focal point white strokes in the sky and the neck, most of the
as well as the size of the angle between the upper rail- cadmium sulphide has turned to colourless com-
ing and the main vertical. pounds. Notably, however, on The Scream (NM),
The arrangement of these features defines the pro- cadmium-based yellow on the railing retains a bright
portion of the space given to the railing and the road yellow hue.
in relation to the space occupied by the landscape and These analyses have proved to be of general inter-
sky. The sketch may appear to have been the first est for the understanding of pigment manufacture in
documented attempt for creating The Scream and the late 1800s and early 1900s as well as pigment deg-
evidently Munch was unhappy with the position of radation. Photodegradation and colour changes in
the focal point in the sketch. After having rejected the cadmium-based yellows have been recorded in paint-
sketch, the second version (the original) was based ings by Vincent van Gogh and Henri Matisse. Thus
on a lowering of the focal point thereby reducing the research attached to the photodegradation and prob-
space given to the railing and the road. In contrast to lems linked to the manufacture of cadmium yellow has
30

intensified over the last decade. The topic is also dealt Notes
with in the articles by Mass et al. in this volume and
continued research on these issues will be conducted 1. L.E. Plahter 1974; 1994: 5−10; 1999: 111−14.
by Jennifer Mass and her team.112 The cadmium car- 2. Hereafter referred to as Ghost.
bonate-rich pigment has so far been identified on two 3. U. Plahter and Topalova-Casadiego 2011: 247−52.
4. Mass et al., in this vol.
paintings by Munch − Ghosts and The Scream (MM),
5. Woll 2009, 1: 1−4.
both dated to the first decade of the 20th century − 6. U. Plahter 1994: 42−6; L.E. Plahter 1999: 121−7. A
but it was not found on Women Turned towards the report on solubility tests suggested that the medium
Sun in the Aula (1914–16).113 When the carbonate- of the orange colour of the floor is based on a drying
rich cadmium yellow became available on the market oil, while the preparation layer could be classified as
and when it ceased to be sold is not known but it is tempera. The greenish flesh colour on Laura’s hand
hoped that pigment analyses on a wider scale will give remains uncertain but could be a mixture of egg and
insight into these matters. drying oil. In general the classification of a mixed
technique is based on visual assessment (U. Plahter
Variations in the composition of cadmium yellows
1994).
most likely escaped the awareness of artists. However,
7. Heller 1984: 114; L.E. Plahter 1994: 121; 1999: 126;
Eibner in 1908 had already published some experi- Lange 2005: 247−8.
ments on cadmium yellows produced by the indirect 8. L.E. Plahter 1974: 114. The X-radiograph showed that
wet method.114 In this case he used cadmium oxa- Munch painted The Sick Child over one or perhaps
late as a substrate and found that the product faded two other compositions. Large amounts of pigment
within two or three months. New findings in Munch’s on the canvas have made it difficult to identify the
paintings have added to the insight on how pigments underlying motif. Aslaksby (2009a: 63, ill. 5) has,
however, made some attempt to trace details that may
were manufactured and how they may have changed
belong to the underlying composition.
over time. Thus new possibilities have been opened
9. L.E. Plahter 1994: 20−22; 1999: 120−21.
up for dating and an improved understanding of pos- 10. L.E. Plahter 1974: 113; 1999: 116. Like The Sick
sible precautions that can be taken for future care. Child, Spring was also painted over an underlying
Future caretakers of Munch’s paintings will also composition.
need to respond to the conclusions of Stein’s recent 11. Thiis 1933.
evaluation of the significance of the ‘kill-or-cure’ tech- 12. Eggum 1991: 196, 201; Jacobsen 2008; Aslaksby 2009:
nique. If the weathering, including bird droppings 137−8. Photograph UWP 13239 (Norwegian Museum
and deposits of dirt, are indeed unintentional, discus- of Science Technology) was taken winter 1892/93.
13. Jacobsen 2008: 201−2.
sion must focus on their actual significance and on
14. Eggum 1987: 32
any physical risks that they may pose to the long-term 15. ‘Jeg oppdaget ogsaa at mine egne øienhaar hadde
survival of the paintings. It is hoped that research- virket med I billedindtrykket. –Jeg antydet dem derfor
ers will continue to build upon the findings outlined som skygger over billedet’: Munch 1929?: 9.
in these studies of Munch’s complex and expansive 16. www.emunch.no/english.xhtml (accessed May 2015).
oeuvre, both to assist in the preservation of his paint- 17. Eggum 1978.
ings and to enhance our understanding of one of 18. L.E. Plahter 1992; Bjerke 2006: 78−9.
Norway’s greatest painters. 19. Aslaksby (2009a: 58–9) referred to Eggum and his
misinterpretation of the historical paper copy and its
clarification when Jacobsen made a new copy from
the recovered glass plate. Clarification based on the
technical implications had, however, already been
Acknowledgements published by Plahter in 1992: L.E. Plahter 1992: 90–94.
20. Jacobsen 2008: 201−2. Photograph UWP 13239, see
We are greatly indebted to Dr Jilleen Nadolny, principal note 11 above.
research associate, Art Access & Research, London for 21. Jæger 1886.
her editorial assistance. We also thank the paintings 22. Aslaksby 2009a: 64, ill. 6.
conservators Thierry Ford, National Museum and Inger 23. Singer et al. 2010: 274−6.
Grimstad, Munch Museum for carrying out measurements 24. Further studies on these issues are required.
on artworks. 25. Aslaksby 1999: 143.
31

26. Woll 2009, 1: 146–7, 382–3; 2: 780–81; 4: 1472. 50. Analysed using the INCA EDX program.
27. Recent unpublished research by Unn Plahter. 51. Singer et al. 2010: 280. Singer also noted the presence
28. Skedsmo 1985: 189. of sulphate and ascribed it to a barium sulphate
29. Woll 2009, no. 130, The Sick Child. The painting was which in fact was observed using PLM. As a sulphate
exhibited in Paris, Bing, nr. 3 Jeune Malade. of barium could not be detected using either XRD
30. Skedsmo 1985. Ustvedt and Aslaksby (2009: 49, n. 66) or EDX in this paper we have concluded that the
refer to a letter from Munch while in Warnemünde sulphate essentially may be assigned to a cadmium
to Thiel 20 July1907 in which Munch wrote ‘I have sulphate rather than a barium sulphate.
received The Sick Child and will soon start thereupon.’ 52. Van der Snickt et al. 2009.
31. Woll 1996: 22. 53. Leone 2003: 34–9, esp. 37; Leone et al. 2005: 811.
32. Ibid., 24. 54. Mirjam Liu carried out the first series of analyses.
33. Ibid., 25. Unn Plahter was responsible for the second series and
34. Woll 2008: 86−7; Bjerke 2006: 65; Eggum in Aslaksby the interpretation of these data.
1999: 139−40. 55. Mass et al., in this vol.
35. Lithograph Woll G84, H: 468 × W: 343 mm; Gerd 56. Buchner 1887b.
Woll, pers. comm. 57. Church 1915: 166−7.
36. Lithograph, Woll G748 H: 552 × W: 430 mm . 58. Eibner 1908a: 1515; 1908b: 1550; U. Plahter and
37. With thanks to paintings conservators Inger Topalova-Casadiego 2011: 250.
Grimstad at the Munch Museum and Thierry Ford 59. Mass et al., in this vol.
at the National Museum for carrying out the internal 60. FTIR analyses of a broader range of cadmium-based
measurements. yellows in Munch’s paintings from the early decades
38. Woll 2008: 88. will throw light on the type of yellows that were
39. Clarke 2006: 44. available in the early decades of the 20th century and
40. Aslaksby 2009: 137. how methods for pigment production developed and
41. U. Plahter 1975. spread in these years.
42. XRF analyses were made using the portable handheld 61. UV induced infrared fluorescence photographs
Thermo Scientific Niton XL3t900 with fundamental obtained from The Scream (MM) were produced by
parameter mining mode calibration. FTIR Mass and her team in Oslo June 2013. The method is
spectroscopic analyses were carried with a Perkin described in the papers by Mass et al., in this vol.
Elmer spectrum v5.0.1. Samples were scanned 32 62. Topalova-Casadiego 2009a.
times over the wavenumber range 4000–450 cm–1 at 63. See Table 2.
a resolution of 4 cm–1. XRD analyses were performed 64. Frøysaker 2007: 246.
using the Debye Scherrer camera, Cu Kα radiation. 65. Frøysaker et al. 2011: 59−61; 2013; Keune et al. 2007.
EDX analyses in the SEM – referred to simply as 66. Frøysaker 2007: 248; Frøysaker and Liu 2009: 48−9.
EDX analyses – were made using a JEOL-JSM 840 67. In the early 1970s paintings conservator Svein Wiik
microscope equipped with Oxford Instruments INCA at the Museum of Cultural History, University of
EDX and Link A10000 analysis system. Oslo, conducted a project on the restoration of the
43. U. Plahter and Topalova-Casadiego 2011: 247−52. Aula paintings. In this connection paint samples for
44. Singer et al. 2010: 274−92. analyses were taken which are stored in the museum
45. Mass et al., in this vol. laboratory for restoration of the paintings. Light green
46. U. Plahter 1999. The paintings investigated included: sample 13 from Men Turned towards the Sun taken to
The Climbing Human /Klatrende mennesker M1021 the right of the thigh of the man at the top and purple
(1927); The Human Mountain: Sphinx M 801 (1927); sample 5 from a beam a little to the left and above the
Kneeling Female Figure M1020 (1927–29); The Human sun were analysed by Unn Plahter in 2009.
Mountain: Kneeling Nude MM 804 (1927); Workers on 68. Frøysaker and Liu 2009.
the Building Site. Draft for a Decoration of Oslo City 69. Ibid., 56.
Hall MM 977 (1929–33); and Portrait of Kai Wefring 70. Tøjner 2003.
M 461 (1934). 71. Ydstie 2008b.
47. In addition to d-spacings for otavite, hexagonal and 72. Topalova-Casadiego 2008.
cubic CdS there are some unidentified d-spacings 73. Hougen 2001.
in the powder pattern for The Scream (MM) not 74. Brenna 1978: 218.
observed in the pattern of The Scream (NM). 75. Ydstie 2008b: 80; Lange 2005. Lange’s article gives
48. Fiedler and Bayard 1986: 77; Leone et al. 2005: 803, a résumé of the debate on the dating of The Scream
810−12. (MM).
49. Analysed using LINK A10000 program. 76. Recent unpublished study carried out by Unn Plahter.
32

77. Bjerke 2008: 13−57. In Munch’s previous portrayals since they show no sign of outdoor exposure. Tine
of Anxiety the position of the main vertical and the Frøysaker, Conservation Studies, IAKH, University
horizon line match rather closely with the placement of Oslo, pers. comm.., 15 January 2015.
of these lines in the two painted versions of The 109. Stein 2014: 108.
Scream. 110. Singer et al. 2010: 274.
78. Bjerke 2008: 17−18, Ill 6. Fra Ljabruchausseen 111. Biljana Topalova-Casadiego, Munch Museum,
1889−90. pers. comm., 10 June 2013. MOLAB EU-ARTECH,
79. Reasons for rejecting the sketch on the reverse side University in Perugia (Eu-ARTECH – Access,
have been discussed by Lange 2005: 244−5. Research and Technology for the conservation of the
80. Bjerke 2008: 13−57. This phenomena can be recognised European Cultural Heritage).
in Munch’s various versions of The Scream, in sketches 112. See also Van der Snickt et al. 2009.
of Anxiety 1891/92 and in the paintings of Anxiety 113. SEM-EDX analyses of a yellow in Women Turned
(1892) and Portrait of Nietsche (1906), both in Thielsa towards the Sun suggest that in this case CdS has
Galleriet, Stockholm, as well as many more. been applied and photo-degradation has so far not
81. Woll 2009, no. 130. been observed.
82. L.E. Plahter 1994: 85–7; Aslaksby 2009a: 57–8; 2009b: 114. Eibner 1908b: 1550.
135–6.
83. Jæger 1886; Munch 1929?: 9.
84. Aslaksby 2009b: 136–7.
85. L.E. Plahter 1974: 103–8.
86. Aslaksby 2009a: 62, 139.
References
87. Plahter 2000: 97.
88. Singer et al. 2010. Aslaksby, T.E. (1999) ‘The Sick Child: a closer look’, in
89. Topalova-Casadiego 2004. Conservare necesse est- Festskrift til Leif Einar Plahter på
90. Frøysaker and Liu 2009: 54−6. hans 70 års dag, E.S. Skaug, K. Kollandsrud and F. Hanssen-
91. U. Plahter and Topalova-Casadiego 2011: 248. Bauer (eds), 139–43. Oslo: Nordisk Konservatorforbund,
92. Topalova-Casadiego 2012. Den Norske Seksjon (IIC Nordic Group).
93. pXRF, the handheld Thermo Scientific Niton XL3t900. Aslaksby, T.E. (2009a) ‘Det syke barn NG.M.00839.
Plahter 2009. Et maleri i forvandling - teknikker og behandling ‘,
94. Stein 2011a,b, 2012. in Edvard Munch: The syke barn. Historien om et
95. While Stein thus far has concentrated on the Ekely mesterverk, Ø. Ustvedt and T.E. Aslaksby (eds), 57‒79.
Collection, her ongoing work will include the Oslo: Nasjonalmuseet.
investigation of other works by Munch such as the Aslaksby, T.E. (2009b) ‘The Sick Child NG.M.00839. A
paintings he sold during his lifetime to find out if any painting in transformation: techniques and treatment’,
show evidence of being deliberately weathered. in Edvard Munch: The Sick Child. The Story of a
96. Stenersen 1945: 73. Masterpiece, Ø. Ustvedt and T.E. Aslaksby (eds),
97. The term referred to the radical steps taken to cure 135‒51. Oslo: Nasjonalmuseet.
a horse (here a painting) of an illness. The ‘cure’ was Bjerke, Ø.S. (2006) ‘Edvard Munch, The Sick Child: form
expected either to kill or to save the animal. and content’, in Edvard Munch: An Anthology, E.
98. Thurmann-Moe 1987; Buchhart 2007; Tøjner 2003. Mørstad (ed.), 65−86. Oslo: Unipub.
99. Stein 2014. Sigurd Willoch was an art historian Bjerke, Ø.S. (2008) ‘Skrik som del av den kunsthistoriske
and later director of Oslo’s National Gallery (1946– kanon’, in Skrik Munch-Museet, I. Ydstie (ed.), 13−55.
1973). He acted as registrar and appraiser in winding Bergen: Vigmostad & Bjørke.
up the Ekely estate. Dieter Buchhart is an Austrian Brenna, A. (1978) ‘Edvard Munchs arbeiderbilder 1909–
conservator and curator (Buchhart 2007). 1915’, Kunst og Kultur 61: 197−220.
100. Thurmann-Moe 1987. Buchhart, D. (2007) ‘Edvard Munch: signs of modern art.
101. Tøjner 2003: 39–40. The duality of a material-based modernity’, in Edvard
102. Buchhart 2007. Munch: Signs of Modern Art, 11–23. Basel: Fondation
103. Stein 2011a. Beyeler.
104. Stein 2011b. Buchner, G. (1887a) ‘Ueber Schwefelcadmium und uber
105. Stein 2012. die verschiedenen Cadmium Farben des Handels’,
106. Ibid., 442. Chemischer-Zeitung 11(72): 1087–9.
107. Woll 2009, nos. 345 and 1411. Buchner, G. (1887b) ‘Ueber Schwefelcadmium und uber
108. This observation does not apply to the monumental die verschiedenen Cadmium Farben des Handels’,
Munch paintings in the Oslo University Aula Chemischer-Zeitung 11(72): 1107–9.
33

Clarke, J.A. (2006) ‘Originality and repetition in Edvard Jæger, H. (1886) ‘Udstillingen. 1. Første Værelse’, Dagen,
Munch’s The Sick Child’, in Edvard Munch: An ‘Exibition I’ 20 October 1886.
Anthology, E. Mørstad (ed.), 43–63. Oslo: Unipub. Keune, K. and Boon, J.J. (2007) ‘Analytical imaging studies
Church, A.H. (1915) The Chemistry of Paints and Paintings, of cross-sections of paintings affected by lead soap
4th edn, revised and enlarged. London: Seeley, Service aggregate formation’, Studies in Conservation 52(3):
& Co Ltd. 161‒76.
Eggum, A. (1978) ‘The theme of death’, in Edvard Munch Lange, M.I. (2005) ‘Skrik etc. nok en gang. Kontrafaktisk
Symbols and Images, 143−53. Washington DC: National kunsthistorie’, Kunst og Kultur 88: 240−53.
Gallery of Art. Leone, L. (2003) Deterioration of Cadmium Sulphide Yellow
Eggum, A. (1987) Munch og fotografi. Oslo: Gyldendal. Pigments, Diploma in Easel Paintings Conservation,
Eggum, A. (1991) Munch og Frankrike. Oslo: Courtauld Institute of Art, London.
Munch-museet. Leone, B., Burnstock, A., Jones, C., Hallebeek, P., Boon, J.
Eibner, A. (1908a) ‘Cadmiumgelb, Cadmiumorange und and Keune, K. (2005) ‘The deterioration of cadmium
Cadmiumrot I’, Farben Zeitung 13(43): 1511–16. sulphide yellow artist’s pigments’, in Preprints of
Eibner, A. (1908b) ‘Cadmiumgelb, Cadmiumorange und ICOM-CC 14th Triennial Meeting, I. Verger (ed.), vol.
Cadmiumrot I’, Farben Zeitung 13(44): 1549–52. 2, 803–12, London: James & James.
Fiedler, I. and Bayard, M.A. (1986) ‘Cadmium yellows Liu, M., Mengshoel, K. and Frøysaker, T. (forthcoming
oranges and reds’, in Artists’ Pigments: A Handbook of 2015) ‘The surface of The Sun: materials and techniques
their History and Characteristics, R.L. Feller (ed.), vol. in Edvard Munch’s central Aula painting’, Painting
1, 65–108. Cambridge: Cambridge University Press. Techniques, History, Materials and Studio Practice,
Frøysaker, T. (2007) ‘The paintings of Edvard Munch Proceedings of the 5th International Symposium at the
in the Assembly Hall of Oslo University: their Rijksmuseum Amsterdam 2013.
treatment history and the Aula-project’, Restauro 4: MOLAB Access Service. User Report. A technical study
246‒57, 266. of paintings by Edvard Munch: “Puberty” (1894),
Frøysaker, T. and Liu, M. (2009) ‘Four (of eleven) “Angst” (1894) and “Vampire” (1895). Materials,
unvarnished oil paintings on canvas by Edvard Munch Technique, Conservation, Future Care. Available
in the Aula of Oslo University: preliminary notes on at: www.eu-artech.org/files/MOLAB_5/REPORT_
their materials, techniques and original appearances’, ANGST_2009.pdf (accessed June 2013).
Restauro 1: 44–62. Munch, E. (1929?) Lifsfriesens tilblivelse. Kristiania.
Frøysaker, T., Miliani, C. and Liu, M. (2011) ‘Non-invasive Plahter L.E. (1974) ‘Det syke barn og Vår. En
evaluation of cleaning tests performed on “Chemistry” røntgenundersøkelse av to Munchbilder’, Kunst og
(1909–1916): a large unvarnished oil painting on canvas Kultur 57: 103−15.
by Edvard Munch’, Restauro 4: 53–63. Plahter, L.E. (1992) ‘Det syke barn - spekulasjoner og fakta
Frøysaker, T., Mengshoel, K. and Moltubakk Kempton, H. om maleriets opprinnelig utseende’, Kunst og Kultur 72:
(2012) ‘Edvard Munch’s Aula Frieze: conservation and 85−99.
change in public attitudes’, in IIC Vienna Conference, The Plahter, L.E. (1994) Munch under overflaten. Beneath the
Decorative, Conservation and the Applied Arts, 359–60. Surface of Edvard Munch. Oslo: Nasjonalgalleriet.
Frøysaker, T., Liu, M., and Milani, C. (2013) ‘Extended Plahter, L.E. (1999) ‘Beneath the surface of Edvard Munch:
abstract – non-invasive assessments of cleaning tests technical examinations of four paintings by Edvard
on an unvarnished oil painting on canvas by Edvard Munch’, in Conservare necesse est: For Leif Einar Plahter
Munch’ in New Insights into the Cleaning of Paintings. on his 70th birthday, E.S. Skaug, K. Kollandsrud
Proceedings of the Cleaning 2010 International and F. Hanssen-Bauer (eds), 111−27. Oslo: Nordisk
Conference Universidad Politécnica de Valencia and Konservatorforbund, Den Norske Seksjon (IIC Nordic
Museum Conservation Institute, M.F. Mecklenburg, A.E. Group).
Charola, and R.J. Koestler (eds), 119–23. Washington Plahter, L.E. (2000) Norsk Signaturleksikon. Malere født før
DC: Smithsonian Institution Scholarly Press. 1920. Oslo: Nasjonalgalleriet.
Heller, R. (1984) Munch: His Life and Work. London: John Plahter, U. (1975) Analyser av malingprøver fra tre malerier
Murray. av Munch. Utført mai 1974 (unpublished report dated
Hougen, P. (2001) ‘“Skrik”: Mennene i bakgrunden er ulike’, 13 March 1975).
Aftenposten 14 March 2001. Plahter, U. (1994) Analyser av malingprøver fra maleri av
Jacobsen, L. (2008) ‘Edvard Munchs fotoalbum fra Atelier Munch: Døden i sykeværelset (unpublished report dated
Marskalk, Berlin 1892–93’ and ‘“Det Syke Barn”’, in 6 January 1994).
Munch blir ‘Munch’: Kunstneriske strategier 1880– Plahter, U. (1999) Analyser av maling fra Munchmalerier
1892, I. Ydstie and M.B. Guleng (eds), 199−205. Oslo: (datert til 1920 årene) (unpublished report dated 11
Munch-Museet. January 1999).
34

Plahter, U. (2009) Pigment analyses of Munch’s Puberty Topalova-Casadiego, B. (2008) ‘De to malte versoner av
(unpublished report dated 9 June 2009). Skrik’, in Skrik Munch-Museet’, I. Ystie (ed.), 87−99.
Plahter, U. and Topalova-Casadiego, B. (2011) ‘The Bergen: Vigmostad & Bjørke.
Scream by Edvard Munch: painting techniques and Topalova-Casadiego, B. (2009a) ‘Technical aspects of
colouring materials’, in Studying Old Master Paintings: Edvard Munch’s paintings’, in Edvard Munch: Complete
Technology and Practice, M. Spring (ed.), 244–52, Paintings, G. Woll (ed.), vol. 2, 425–57, London:
London: Archetype Publications in association with Thames and Hudson.
the National Gallery. Topalova-Casadiego, B. (2009b) ‘Edvard Munch’s Puberty,
Singer, B., Aslaksby, T.E., Topalova-Casadiego, B. and 1894, analysed by Eu-Artech MoLab’, Norske Konserves
Tveit, E.S. (2010) ‘Investigation of materials used by 2: 27.
Edvard Munch’, Studies in Conservation 55(4): 274−92. Topalova-Casadiego, B. (2012) ‘Painterly aspects of
Skedsmo, T. (1985) ‘Tautrekking om Det syke barn’, Kunst Puberty: the results of recent studies’, in Puberty, I.
og Kultur 68: 184−95. Ydstie (ed.), 66–84. Oslo: Munch Museum.
Stein, M. (2011a) ‘Edvard Munch paintings with bird Tøjner, P.E. (2003) Munch: In his Own Words. Munich:
droppings: analysis of the Ekely Collection at the Prestel Verlag.
Munch Museum’, Zeitschrift für Kunsttechnologie und Ustvedt, Ø. and Aslaksby, T. (eds) (2009) Edvard Munch
Konservierung 25(1): 93–102. Det Syke Barn/The Sick Child. Oslo: Nasjonalmuseet.
Stein, M. (2011b) ‘Edvard Munch’s paintings with Van der Snickt, G., Dik, J., Cotte, M., Janssens, K.,
water stains: analysis of the Ekely Collection at the Jaroszewica, J., De Nolf, W., Groenewegen, J. and Van
Munch Museum’, Zeitschrift für Kunsttechnologie und der Loeff, L. (2009) ‘Characterization of a degraded
Konservierung 25(2): 273–84. cadmium yellow (cadmium sulphide) pigment in an
Stein, M. (2012) ‘Edvard Munch paintings with weathered oil painting by means of synchrotron radiation based
paint surface: analysis of the Ekely Collection at the X-ray techniques’, Analytical Chemistry 87: 2600−10.
Munch Museum’, Zeitschrift für Kunsttechnologie und Woll, G. (1996) Edvard Munch Grafikk fra 1896. Oslo:
Konservierung 26(2): 435–47. Munch Museum.
Stein, M. (2014) ‘Edvard Munch and the ‘kill-or-cure’ Woll, G. (2008) ‘Bruk og gjenbruk av Munchs tidligste
remedy: analysis of the Ekely Collection at the malerier’, in Munch blir “Munch”, Kunstneriske strategier
Munch Museum’, Zeitschrift für Kunsttechnologie und 1880–1892, I. Ydstie and M.B. Guleng (eds), 85–103.
Konservierung 28(2): 279–97. Oslo: Munch-Museet.
Stenersen, R. (1945) Edvard Munch: Nærbilde av et geni. Woll, G. (2009) Edvard Munch: Complete Paintings, 4 vols.
Oslo: Sem & Stenersen. London: Thames and Hudson.
Thiis, J. (1933) Edvard Munch og hans samtid. Slekten, livet Ydstie, I. (ed.) (2008a) Skrik Munch-Museet. Bergen:
og kunsten, geniet. Oslo: Gyldendal Norsk Forlag. Vigmostad & Bjørke.
Thurmann-Moe, J. (1987) ‘“Rosskur” und Firnis bei Edvard Ydstie, I. (2008b) ‘Datering av Munch-museets Skrik’, in
Munch’, in Das 19. Jahrhundert und die Restaurierung: Skrik Munch-Museet, I. Ydstie (ed.), 77−87. Bergen:
Beiträge zur Malerei, Maltechnik und Konservierung, H. Vigmostad & Bjørke.
Althöfer (ed.), 112–21. Munich: Callwey.
Topalova-Casadiego, B. (2004) ‘Edvard Munchs Løsrivelse,
1896 – aspekter ved maleteknikken’, Norske Konserves
2: 13–14.
35


Munch S Paintings

Uploaded by

Copyright:

Available Formats

Munch S Paintings

Uploaded by

Document Information

Original Title

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

Munch S Paintings

Uploaded by

Copyright:

Available Formats

Introduction

EdMunch-01-Plahter-v6.indd 1 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 3 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 4 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 5 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 6 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 7 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 8 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 9 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 10 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 11 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 12 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 13 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 14 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 15 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 16 09/10/2015 14:29

As mentioned above, pXRF analyses of thin, off-

EdMunch-01-Plahter-v6.indd 17 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 18 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 19 09/10/2015 14:29

The Scream: perspective studies76

EdMunch-01-Plahter-v6.indd 20 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 21 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 22 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 23 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 24 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 25 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 26 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 27 09/10/2015 14:29

Previous title: Portrait of the

EdMunch-01-Plahter-v6.indd 30 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 31 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 32 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 33 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 34 09/10/2015 14:29

EdMunch-01-Plahter-v6.indd 35 09/10/2015 14:29

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.