Maney Publishing and International Institute for Conservation of Historic and Artistic Works are collaborating with

JSTOR to digitize, preserve and extend access to Studies in Conservation.

http://www.jstor.org
Maney Publishing
A German House Altar from the Sixteenth Century: CONSERVATION AND RESEARCH OF
REVERSE PAINTINGS ON GLASS
Author(s): Simone Bretz, Ursula Baumer, Heike Stege, Johannes von Miller and Dedo von
Kerssenbrock-Krosigk
Source: Studies in Conservation, Vol. 53, No. 4 (2008), pp. 209-224
Published by: on behalf of the Maney Publishing International Institute for Conservation of
Historic and Artistic Works
Stable URL: http://www.jstor.org/stable/27867045
Accessed: 18-08-2014 08:48 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
209
A German House Altar from the
Sixteenth
Century
CONSERVATION AND RESEARCH OF REVERSE PAINTINGS
ON GLASS
Simone
Bretz,
Ursula
Baumer,
Heike
Stege, Johannes
von Miller and Dedo von
Kerssenbrock-Krosigk
Prior to the conservation treatment
of
a late
sixteenth-century
South-German house altar with seven reverse
paintings
on
glass,
technical
and
scientific
research was conducted on the
glass substrate, binding media,
colourants and metals.
Microscopic
examination revealed the
particularities of
the so-called
amelierung technique
and the
glass panel manufacture.
For material
analysis, gas chromatography?mass
spectrometry,
amino acid
analysis, high performance liquid chromatography,
as well as micro
X-ray fluorescence spectrometry
and
scanning
electron
microscopy
with
energy dispersive X-ray microanalysis
were
applied.
The main
component of
the binder was
identified
as a heat-treated
pine
resin
(colophony), modified
with various
softening
resins. The red
paint layer
contains cochineal and small
amounts
of dragons
blood. The
original
blue and
green areas,
which contain smalt and
copper resinate,
have
undergone
considerable
colour
change.
The reverse
paintings
on
glass
were in a
poor
state
of preservation:
due to
high
moisture
(flooding
at the
museum)
and
material
degradation,
the
paint layers
were
severely delaminated, fractured
into hundreds
of
small
paint flakes
and
partially
attached to
silverfoil
and
paper backing
The detached
paint flakes
were
transferred
back to their
original positions
on the
glass
with a
fine
brush
and secured
using
a
hydrocarbon
resin. This was
followed by
the
application of microcrystalline
wax which achieved a brilliant saturation
of
the reverse
paintings
on
glass. However,
the discolouration remained in the blue and
green
areas. A
photograph of
the altar's central
panel after
the conservation treatment was
digitally manipulated
to recreate the
presumed original tonality.
INTRODUCTION
Almost 40
years
went
by
from the time of the
acquisition
of a small house altar
by
the
Corning
Museum of Glass
in 1959 until its
permanent display
in the museum's
galleries.
A conservation treatment
grant approved by
the
Getty
Foundation made it
possible
to exhibit this
masterpiece
of Renaissance reverse
painting
on
glass
(Figure 1).
The
poor preservation
condition and severe
damage
prompted
a
preliminary study
of the
technology
and
materials of the altar's reverse
paintings
on
glass.
This
delicate and
fascinating painting technique,
called
amelierung,
has been the
subject
of
technological
research,
which was
published by Ryser
and Bretz in several
monographs, papers
and contributions to exhibition
catalogues [1?6].
Scientific
analysis
of colourants and
binding
media of reverse
paintings
on
glass
has been
carried out
by
Hahn
[7]
and
by
Koller and Baumer
[8],
but more work is
required
to allow broader
comparisons
between
periods
or
regions
to be made.
Regarding
the actual state of the art in conservation on reverse
Received
February
2008
paintings
on
glass,
Bretz
published
an overview on
restoration treatments of the
past
20
years [9].
This
paper presents
a
summary
of
findings
on the
technique
and materials used for the
Corning
house
altar with
special emphasis
on the reverse
paintings
on
glass. Stereomicroscopy
was used to examine
closely
the
painting technique
and to document the state of
preservation
of the
paintings.
Various
spectroscopic
and
Chromatographie
methods were used for the
analysis
of
glass panels,
metals,
colourants and
binding
media..
The results of the
analysis
contributed to a better
understanding
of colour
changes
and allowed the
appropriate
choice of conservation materials to be
made. A
description
of the conservation/restoration
treatment and its difficulties is
given.
The
analytical
and
conservation efforts
yielded enough
information for a
virtual reconstruction of the
original appearance
of one
of the reverse
paintings
on
glass.
DESCRIPTION OF THE HOUSE ALTAR
The house altar from the
Corning
Museum of Glass
(acc.
no.
59.3.39)
is made of an
ebony-veneered
oak case
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
210 S.
BRETZ,
U.
BAUMER,
H.
STEGE,
J. VON MILLER AND D. VON KERSSENBROCK-KROSIGK
Figure
1 House altar with reverse
paintings
on
glass, Corning
Museum of
Glass
(59.3.39),
after conservation. ?
Corning
Museum of Glass, Corning
(NY).
and is decorated with seven reverse
painting
on
glass
panels (Figure 1).
It
originates
from South
Germany,
perhaps Nuremberg
or
Augsburg,
and
probably
dates
from around 1560-1580. The central
panel
shows
the
Crucifixion,
with St
Mary,
St
John,
and
probably
Mary Magdalene (Figure 2).
The
wings,
when
open,
depict
the
allegorical figures Pity
and
Charity, bearing
the
inscriptions
'MISERICORDI' and 'CHARITAS'
respectively.
The
wings,
when
closed,
display
St
Bridget
and St
Dorothy,
with the
inscriptions
'S. PRIGIDA'
and 'S. DOROTHEA'. The oval
panel
on the
pedestal
represents
the Seven Sorrows
of
the
Virgin
in a medallion:
Figure
2 Crucifixion after treatment. ? Simone Bretz, Munich.
the
Virgin Mary,
with seven swords
pointing
toward her.
The
panel
in the lower niche shows the
Virgin
as
Queen
of
Heaven,
holding
the Christ Child. The back of the altar
is flat. The dimensions of the house altar are:
height
49.6
cm; width with
wings
closed 19.5 cm;
depth
12.7 cm.
House altars
(i.e.,
altars used for
private devotion)
with reverse
paintings
on
glass
are not rare. An altar of
this kind in the
Bayerisches
Nationalmuseum in
Munich,
Germany (acc.
no. R
1104)
shows the Coronation
of
the
Virgin Mary,
surrounded
by
other scenes related to
Mary:
the Seven
Sorrows,
the
Piet?,
the
Flight
into
Egypt,
and the
Assumption [10, pp. 59-60,
fig. 14].
The artist followed
closely
a
print by Johannes
Sadeler after a
composition
by
Marten de Vos
[10, p. 58,
fig. 12].
The altar is 37 cm
high;
its
wings
are not movable. A second
comparable
altar,
showing
the
Assumption of
the
Virgin Mary
with
St
Jerome,
St
Ambrose,
St
Gregory
and St
Augustine,
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
A GERMAN HOUSE ALTAR FROM THE SIXTEENTH CENTURY 211
is in the Museo Civico in
Turin,
Italy (acc.
no. V.O. 48
2941),
and has a
height
of 39 cm
[11].
A third altar is
in the National Museum in
Cracow,
Poland
(acc.
no.
MNK
XIII-2324).The
case is made of
ebony-veneered
pinewood (height
44
cm),
and the central
panel
shows
the
Nativity,
with the Annunciation and the Adoration
of
the
Magi
on the side
panels.
These three altars also
originate
from South
Germany
and date from the late
sixteenth or
early
seventeenth
century.
Of these three
altars,
the Cracow altar most
closely
resembles the
Corning piece.
Both have
wings
that are
hinged
on the
columns in a
very sophisticated way. Also,
the
general
configuration,
with a drawer at the
bottom,
is the same.
TECHNICAL EXAMINATION
Glass manufacture
The
technique by
which the
glass panels
of the house
altar were made is as
yet
not
entirely
understood. The
glass
is
very glossy
and thin
(between
1.7 and 3
mm)
and
slightly
uneven, which makes it
unlikely
that the
panels
were cast and
polished.
The
presence
of
mostly
perfectly spherical
and
only
a few oval air bubbles in the
glass
indicate that the
panels
were made
by
the
cylinder
blowing process [5, pp. 189-190]. Judging
from the
length
of the oval
bubbles,
the
cylinder
must have been
short,
otherwise the bubbles would have been more
elongated
due to
lengthening
of the hot
glass cylinder
during blowing.
Once the
cylinder
had been cut
along
its
length
and allowed to
flatten,
the
panels
were cut to
size.
Amelierung technique
Well-executed reverse
paintings
on
glass
do not reveal
the
complexity
of their manufacture. Since the
design
is
applied
to the back of
glass panels,
the
painting
must be built
up
in reverse
order,
starting
with the
details and
foreground
and
working
'backwards'. This
technique
makes corrections
virtually impossible.
Reverse
paintings
on
glass
are known under the German
term
hinterglasmalerei
in
many
countries. If the
technique
makes use of
gold
and silver leaf
applied
to the
glass,
the
painting
is often referred to as verre
?glomis?.
However,
in
the case of the house
altar,
the German term
amelierung
(or gamalierung)
is more
appropriate.
The term is not
translatable but in old German
gamal
means
colourfully
decorated. This
speciality
had been introduced in
Nuremberg, Germany,
before 1532
by applying
metal
leaf and/or metal
powder (both gold
and/or
silver)
onto
the
glass panel,
after which the decoration was etched
with a needle or
stylus [5, pp. 196-201].
A second
layer
of colourful translucent
lacquers (lustre)
was then
applied
and covered with a
layer
of
silvery
foil
(silver
or
tin),
creating sparkle
and
luminosity.
In the case of the house
altar a thin silver foil was attached to
paper,
which was
then
glued
to the
edges
of the
glass panel.
In addition
to the
technique
of
amelierung,
the altar shows a
special
treatment of the flesh
layer, consisting
of silver
powder
and
pigments
that were
painted
on,
engraved
afterwards
and
highlighted by applying
black
paint.
Delamination of
paint layers
The most
typical disfigurement occurring
on reverse
paintings
on
glass,
which was also observed on the house
Figure
3 Crucifixion before treatment. ?
Corning
Museum of
Qlass,
Corning (NY).
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
212 S.
BRETZ,
U.
BAUMER,
H.
STEGE, J. VON MILLER AND D. VON KERSSENBROCK-KROSIGK
Figure
4 Detail of
Virgin
as Queen of Heaven, showing delamination,
cupping
and loss of
paint.
? Simone Bretz, Munich.
altar,
is the
separation
of
paint
from the
glass,
which
when seen from the front
appears
as
patches
of
greyish,
less saturated areas of
paint.
An estimated 80-90%
of the
paint
surface of the house altar was no
longer
securely
attached to the
glass
substrate
(Figures
3 and
4).
Hundreds of
small,
dislodged paint
flakes and
splinters
were loose and needed to be
reintegrated.Various
factors
like normal
ageing
mechanisms under the influence of
changing
climate conditions caused
brittleness,
flaking
and
resulting glass-like paint fragments. Degradation
processes
were further
aggravated
when the
Corning
Museum of Glass was inundated
during
a flood in 1972.
Lastly,
heat,
perhaps
from accelerated
drying
after the
flood,
had caused
softening
and
fusing
of some of the
paint layers.
The
glass
itself was found to be in
good
condition and showed no
signs
of
degradation
like
crizzling.
In
general,
a reverse
painting
on
glass comprises
a
non-porous glass
substrate and a
multi-layered paint
system.
Internal and external factors influence the
degradation.
A
cleaned,
degreased glass
surface and a
well-chosen
binding
medium are crucial factors for
good physical bonding
at the
paint/glass
interface. Poor
surface
wetting produces
voids that restrict contact and
trap
air. The level of adhesion of the
paint layer
to the
glass support depends
also on the
composition
of the
paint (colourants
and
binding media)
and its
drying
mechanism,
the
process
of film formation and the
method of
application.
The brittleness or
flexibility
of
the
organic coating
is in correlation with the thickness
of
application:
the thicker a
paint layer,
the more
prone
it
is to delamination. The
instability
can also be caused
by
a
warm environment that results in faster volatilization of
the solvent and softeners. With
changes
in
temperature
and,
as
consequence, changes
in relative
humidity,
the
paint
film takes
up
and releases moisture. The stress of
shrinking
and
expanding
will lead to hairline cracks and
voids within the
paint
matrix,
to
blistering
and
cupping,
and
ultimately
to the loss of
paint fragments.
MATERIAL ANALYSIS
Glass
To avoid
taking samples
from the
undamaged glass,
a
semi-quantitative,
and for some elements
qualitative
determination of the
glass composition
was undertaken
by energy dispersive
micro
X-ray
fluorescence
spectro
metry ^XRF).
Six of the seven
panels
were examined
after local
cleaning
with
ethanol/water,
but without
additional
polishing,
thus
taking
into account that values
obtained for
light
elements,
mainly
sodium to
silicon,
are
affected
by leaching processes.
The six
panels
showed a
very
similar
composition
with
only
small variations. In
general,
the
glass
is rich in
potassium (~
16-17 wt%
K20)
with medium calcium content
(~
8-10%
CaO)
and
only
traces of lead. In
addition,
the absence of bromine
in the
spectra proves
that ash from
halophytic,
soda-rich
plants
can be excluded as the
fluxing agent.
Using
a similar
spectrometer type,
M?ller was
able to show that traces of bromine can be
regarded
as
a characteristic marker for soda ash
glasses [12, p. 122].
The ratio of
K90/CaO
in the
glass panels
was found
to be around 1.7 to 2.0 and therefore at the
upper
limit
for so-called 'colourless wood ash
glasses',
a modified
typification suggested by
M?ller
[12] following
the
chemical classification of medieval
glass by Wedepohl
[13].
However,
a
partial
addition of
purified
wood ash
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
A GERMAN HOUSE ALTAR FROM THE SIXTEENTH CENTURY 213
(potash)
to the batch of
glass
used for the house altar
panels
cannot be excluded.
Very
low iron oxide content
(~
0.1?0.15%
Fe203)
indicates that
good quality
sand
was used as raw material. The
presence
of
manganese
dioxide
(~
0.5%
MnO) suggests
the addition
of'glass
makers'
soap' (Mn02)
as decolourizer.
Recent
analytical
research
by
M?ller on medieval
and
post-medieval glasses [12]
has shown that the
glass
composition
found for the house altar
panels
is the
most
typical
for
sixteenth-century
colourless
(vessel)
glass
from South
Germany.
However,
it could have
originated
elsewhere in Central
Europe,
for
example
Saxony,
Bohemia or Eastern Austria. An
import
of the
glass panels
from
regions producing mainly
sodium-rich
glass
at that
time, e.g. Italy
and also
Tyrol (Austria),
can be
ruled out
[12, especially pp. 194-220].
Binding
media
The technical literature on reverse
paintings
on
glass
mentions a wide
variety
of
possible binding agents,
such
as
drying oils,
natural resins and
gums [4, pp. 39?40, 14].
Yet
in-depth analyses
of
binding
media are rare. For the
identification of
binding
media of the house
altar,
five
small
samples (~
0.5
mg)
from the St
Bridget panel
were
used.
The
binding
media were first
analysed by gas
chromatography (GC)
and
gas chromatography?mass
spectrometry (GC-MS).
A
special
solvent extraction
scheme for
aged
and altered materials was used for
sample
preparation [15].
With the
help
of various
polar
solvents
complex
mixtures of
components
with different
polarity,
such as
waxes,
oils and resins can be
pre-separated,
thus
yielding
more
easily interpretable chromatograms.
The
solvent extracts were
directly injected
into the GC
or GC-MS without
prior
derivatization. Instead of
derivatization,
oxalic acid
(10% w/v)
was added to the
methanol extracts
(see Appendix).
The
transparent samples
of
red,
brownish-red and
green paint
were
easily
soluble in methanol and
gave
very
similar GC-MS results with
only slight
variations
in the
quantitative composition.
The main
compo
nents in these
samples
were
diterpenoid
resin acids of
the abietane
type (dehydroabietic acid), apart
from a
minor
variety
of oxidized and
dehydrogenated
diter
penoid
resin acids
(e.g. 7-oxo-dehydroabietic
acid,
7-oxo-dehydrodehydroabietic acid).
Resin acids of the
abietane
type
are the most abundant
components
of
diterpenoid
natural resins from the Pinaceae
family.
In
particular, colophony,
the resin of
pine
trees
(Pinns),
was
widely
used in art and craft. Due to isomerization
and
disproportionation
reactions
during
the distillation
process
of the
pine
resin,
abietic and
particularly
dehydroabietic
acid are formed from a
varying
number
of
diterpene
resin acids
[16].
On
long exposure
to air and
light
the
resulting relatively
stable
dehydroabietic
acid
will oxidize to form several oxidized
products [17, 18].
In contrast to these
ageing processes,
the
high
content of
oxidized and
dehydrogenated
resin acids in the
analysed
samples
is
only explicable by strong heating
of the resin
[19].
Research on resin and tar
production
revealed
that certain
types
of resins can be obtained
by melting
at
high temperatures.
The resins obtained after indirect
heating
of the resinous wood contain isomerized and
dehydrated products including dehydroabietic
acid and
dehydrodehydroabietic
acid
(abieta-tetraenoic acid)
as
well as oxidized
products
like
7-oxo-dehydroabietic
acid and
7-oxo-dehydrodehydroabietic
acid. The main
components
now
present
in the
Corning
altar
samples
are therefore the
degradation
and
heating products
of
former
pine
resin
(Figure
5,Table 1).
In
addition,
neutral
diterpenoid components
from
larch
turpentine (epimanool,
larixol,
larixylacetate, Figure
6)
could be identified
[20, 21].
Balsamic larch
turpentine
originates
from the Larix
species (e.g.
Larix decidua
Mill.),
which
grow
in the
Austrian-Tyrolean region.
Larch
turpentine belongs
to the
group
'fine
turpentine',
which
does not tend to
crystallize
and retains a
long-lasting
elasticity.
Even
today,
larch
turpentine (in
combination
with Canada
balsam) plays
a
special
role in the
glass
and
optical industry,
where it is used as
high-grade glue
for
glass prisms
and
glass panels [22].
Only
minor amounts of a third resin could be found
in the
triterpenoid region
of the
chromatogram (Figure
6).
Small amounts of noroleanenone and other related
oleanane structures were
identified,
which serve as
useful
compounds
in the identification of
aged
mastic
resin
[23].
Mastic is the resin of the mastic tree
(Pistacia
lentiscus
L.),
of the Pistachio
family.
Its natural distribution
area encloses the coastal
regions
of the Mediterranean
Sea,
but most mastic comes from the Greek island of
Chios.
After
nearly
400
years,
some volatile
compounds
were
still detectable in the
samples.
Due to the thickness of the
paint layers,
small amounts of
monoterpenoids camphor
and cineole are
preserved. Camphor
is
originally
contained in the bark of the
camphor
tree
(Cinnamomum
camphora)
and was added as softener
[24]. 1,8-cineole
(eucalyptol)
is no constituent of the resins
analysed
but
derives from the addition of an essential oil as solvent for
the resin mixture. A cineole-rich essential oil
(spike oil)
was
produced
from the lavender
subspecies
Lavendula
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
214 S.
BRET2,
U. BAUMER, H.
STEGE,
J. VON MILLER AND D. VON KERSSENBROCK-KROSIGK
Oxidation
DA
Dehydroabietic
acid 7-Oxo-dehydroabietic
acid
ODA
DDA
ODDA
Figure
5 Green
lacquer layer, gas chromatogram
of the methanol extract. The addition of oxalic acid leads to a
slightly changed polarity
of the
non-polar
separation
column and the resin acids could be
separated.
The structures of the most stable
pine
resin
compound (dehydroabietic acid)
and its oxidation
product (7-oxo-dehydroabietic acid)
are
given.
Abbreviations: C
=
camphor,
DA
=
dehydroabietic acid, DDA
=
dehydrodehydroabietic acid, ODA
=
7-oxo
dehydroabietic acid, ODDA
=
7-oxo-dehydrodehydroabietic
acid. ? Doerner Institut, Munich.
Table 1
Summary
of
binding
medium
components
identified with their
origin
and colourants of the
paints (Saint Bridget panel)
Components
identified
Origin
Red
sample
Red-brown
sample
'Green'
sample
'Blue'
samples
(now brown) (now greyish-black)
dehydroabietic
acid
dehydrodehydroabietic
acid
7-oxo-dehydroabietic
acid
7-oxo-dehydrodehydro
abietic acid
epimanool
larixyl
acetate
larixol
28-nor-olean-17-en-3-one
camphor
1,8-cineole (eucalyptol)
amino acids
(mainly
serine and
glutamic acid)
fatty
acids
(mainly palmitic
and stearic
acid)
colourants
heated and oxidized
pine
resin
larch
turpentine
mastic
camphor
spike
oil
aged egg protein
] (fat)
cochineal,
dragon's
blood
(+)
dragon's
blood copper
resinate smalt,
lamp
black
(thin,
black
layer)
Abbreviations: +++
=
main
component,
++
=
minor
component,
+
=
small amounts, (+)
=
trace amounts,
- =
not detected.
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
A GERMAN HOUSE ALTAR FROM THE SIXTEENTH CENTURY 215
Larjxyl
acetate
L2
DA
ODA
Figure
6 Red
lacquer layer, gas chromatogram
of the methanol extract. In the
pure methanol extract, without further derivatization
agents,
the neutral
diterpenoid
markers for larch
turpentine (L1
and
L2)
are
clearly
visible. The structure of the main
component
of larch
turpentine, larixylacetate,
is
given.
Abbreviations: C
=
camphor,
L1
=
epimanool,
L2
=
larixylacetate,
DA
=
dehydroabietic acid, ODA
=
7-oxo-dehydroabietic acid, M
=
noroleanenone
(mastic).
? Doerner Institut, Munich.
latifolia (Spike lavender) [25],
Hence,
spike
oil was used as
a solvent for this resin mixture
(Table 1).
Samples
taken from an area that was
originally
blue
were found to contain the
pigment
smalt
(see below)
and revealed similar resin mixtures
(larch turpentine,
oxidized
pine
resin, mastic),
but also small amounts of
saturated
fatty
acids in a ratio which is not characteristic
of
drying
oils but rather of
compounds
from
protein
aceous media. Hence the non-soluble residues of the
blue
samples
were further
investigated by
amino acid
analysis [26].
The amino acid
profiles
of the
hydrolysed
samples clearly
showed the
presence
of an
aged
and
degraded egg binding
medium. In coherence with the
fatty
acids detected in the solvent extracts, these markers
hint toward a 'fat'
egg binding
medium
(entire egg
or
egg yolk).
It is
possible
that
egg
was first used to
help
grind
the
pigment,
while the resin mixture served as
painting
medium.
Colourants and metals
Five colourants
(Table 1)
were identified in
red,
brownish-red, green
and blue
paint samples
from the
St
Bridget panel. Scanning
electron
microscopy/energy
dispersive X-ray microanalysis (SEM-EDX)
and
^,
high performance liquid chromatography (HPLC)
with
ultraviolet-visible
(UV-VIS)-detection
and GC-MS
were used for the
analysis.
No
samples
of the flesh tone
were available for
analysis.
The red lake contained carminic
acid,
which was
identified
by
its retention time and UV-VIS
spectrum
using
HPLC. Carminic acid is the main colourant of
cochineal,
one of the most valuable
organic
lakes used
for
paintings
in the late sixteenth
century [27, 28].
Due
to the minute
sample
size,
no minor
components
of the
lake could be
detected,
which left no means to determine
the
provenance
of the cochineal lake more
precisely.
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
216 S. BRETZ, U. BAUMER,
H.
STEGE,
J. VON MILLER AND D. VON KERSSENBROCK-KROSIGK
Small amounts of
aged
dracorhodin and characteristic
triterpenes
from
dragon's
blood,
a red
plant
resin,
here of
the
Daemonorops
draco
species,
were detected in red and
brownish-red
samples by
GC-MS
[28, pp. 142f., 29, 30].
This reddish-brown resinous
plant
exudation
originates
from Indonesia
(Sumatra/Borneo). Although dragon's
blood is
frequently
mentioned in historical sources, its
analytical
identification on sixteenth
century
art works
has
rarely
succeeded so far.
The bluish
paint layer
was found to contain smalt
as the main
pigment (it
also contains traces of
dragon's
blood,
presumably
from an
adjacent
red
paint). Significant
amounts of arsenic and
iron,
and traces of
bismuth,
copper,
nickel and zinc were detected
by
SEM-EDX and
E
They
are
typical
for smalt of the second half of
the sixteenth
century,
which was
mostly
extracted from
cobalt ores in the
Erzgebirge (Ore
Mountains between
Saxony
and
Bohemia) [31, 32]. Using smalt,
the artist of
the house altar
may
have tried to minimize the costs of
materials. The low refractive index of smalt
(1.49-1.53)
is close to that of the resinous
binding
medium that was
used,
and makes it an ideal translucent colourant for this
paint composition.
Whereas the use of smalt in easel
painting
is somewhat difficult because of
floating
and
dripping
of the wet
paint [33],
in reverse
paintings
on
glass
the blue
pigment advantageously
settles down on
the
glass
while the
paint
dries,
and thus becomes
highly
visible
through
the
glass.
The
greenish-brown paint layer
was found to have
a
high
content of
copper.
SEM-EDX
analysis
showed
that
copper
is
homogeneously
distributed
throughout
the
lacquer
without
showing
distinct
pigment particles.
Keeping
the results of the
binding
medium
analyses
in
mind,
this is one of the rare cases where
copper
resinate could
unambiguously
be identified in such a
glaze.
The colourant
may
have been
produced
in the
artists'
workshop by boiling
a
copper compound
such
as
verdigris
with the
resin(s)
as described in the De
Mayerne manuscript (c. 1620?1646, e.g.
fol. 31 and
31
[34, 35]),
which
explicitly
mentions the use of this
compound
on
gold
and silver foil as well as on
glass.
However,
it is also
possible
that the
copper
resinate
slowly
formed over the centuries as a reaction
product
of
verdigris
and the resinous medium. The
green paint layer
is
noticeably
thicker than the other
lacquers,
and it has
an even surface with no visible brushstrokes.
The silver area that
represents
the flesh was,
as men
tioned
earlier,
coated with a
black, opaque paint,
which
contains a
finely dispersed
black
pigment, presumably lamp
black.
Gold and silver foils were used to create the
amelierung
effect. Both metals were found to be of
high quality
with
only
about 5 wt%
Ag
in the
gold
and no other metal in
the silver in amounts that could be detected
by
SEM
EDX. A
heavily
corroded metal foil on the reverse side
of one of the
panels (Seven Sorrows),
which
protected
the
silver
layer,
was identified as
pure
tin foil
(thickness
about
7
).
6 size used to
apply
the
gold
and silver leaf on
the
glass
could not be identified because the
coating
was
extremely
thin.
DISCOLOURATION AND DIGITAL
RECONSTRUCTION
Although
the
paint layers
were
protected against
UVB
and UVC
light by
the
glass
substrate,
the
binding
medium was
aged.
It had become brittle and
changed
from the
original light yellow
to a
deeper yellow
or even
darker tone. The latter is most obvious in the
original
blue
paint layer (see below),
whereas the
optical
effect of
yellowing
is far less evident in the red
lacquer.
The red
paint
colour is in
comparably good condition,
although
some uneven
fading
has been observed in
transmitted
light (Figure 7).
The
light-fastness
of the two
organic
colourants identified is different: cochineal on
alum substrate is considered one of the
reasonably
stable
red lakes available in the sixteenth
century [36],
whereas
dragon's
blood is
usually
described as
being prone
to
fading.
Artificial
ageing
tests of this resin has confirmed
the low
light-fastness
but showed considerable difference
in
stability
between the various botanical sorts and
binding
media
[37].
The intense
green
areas
depicting
meadows
(Figures
2 and
3)
are
greenish
brown
today.
These areas
only
reveal their
original green
colour in transmitted
light.
The discolouration is most
likely
due to
yellowing
of
the medium and to a
greater
extent to the
browning
of
the
copper glaze.
This
impermanence
of
copper
resinate
and also
verdigris glazes
is a well-known
phenomenon
in
painting [38], yet
its discolouration mechanisms and
the
accelerating
factors are not
yet entirely
understood.
However,
the final formation of brown CuO is
suggested by
various authors as a
possible
cause for the
discolouration
[35, pp. 150f, 39].
Deterioration and colour
change
of the
formerly
blue
areas of the
sky
and the
background
of the medallions
(Figures
2 and
3)
are
comparably
severe. Smalt has
partially
discoloured due to the well-known
process
of
leaching
of
potassium
from the
pigment particles.
Interestingly,
the condition of the
binding
medium and
thus its effect on the
optical appearance
of the blue
paint
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
A GERMAN HOUSE ALTAR FROM THE SIXTEENTH CENTURY 217
Figure
7
Virgin
as Queen of Heaven in transmitted
light,
with
reintegrated
paint
flakes.
Fading
of carmine and
dragons'
blood can be seen. ? Simone
Bretz, Munich.
shows variations: in some areas the
binding
medium,
which was identified as a mixture of resins with
egg,
appears extremely
darkened or even black. Similar
darkening
is
typical
for oil
paint layers
that contain
degraded
smalt.
Saponification products
are mentioned
as a
possible
factor
[40], yet
the chemical
processes
in the
medium used here
may
be different. In other blue areas
of the
Corning
altar,
the
paint
is
totally opaque, giving
a
milky, light-blue appearance.
A cross-section revealed
strong degradation
of the binder in the back-scattered
electron
image
with several micro-cavities
probably
acting
as
light-scattering
sources
(Figure 8).
In order to have a better idea of the
original appear
ance of the house
altar,
given
the
severely
obscured
tonality
of the
glass paintings,
a
digital
reconstruction of
the
panel Crucifixion
was made.
Only
the
original green
and blue
parts,
which were found to be most
heavily
affected
by
colour
changes,
were
digitally
rendered.
Based on the visual colour estimation of the
remaining
green
and blue
spots
visible under
magnification
in
transmitted
light
and on the
knowledge
of the
original
colourants
through analysis, paint
trials with smalt of
dark
quality
and
copper resinate,
bound in a natural resin
mixture
(colophony,
mastic in
ethanol)
were
prepared.
Using Photoshop?
CS2,
the blue and
green
areas were
isolated and their colour values
selectively adjusted
until
they visually
matched the
paint samples provided.
Figures
2,
3 and 9
compare
the condition and
tonality
of the central
panel
of the house altar before and after
the conservation treatment and as a result of the
digital
reconstruction.
CONSERVATION
All seven
panels
with reverse
painting
on
glass
are set into
an oak frame and held in
place by ebony
veneer. The
construction is
simple,
but
finely
detailed. The
original
joints
were made with animal
glue.The glass panels
were
dismantled
by removing
the veneer frames
mechanically
without
any
use of solvents to avoid
possible damage
to
the
paint layers.
The surface of the
ebony
veneer was
cleaned with distilled water and white
spirit
afterwards
to remove accumulated
grime
and dirt.
In order to
gain
access to the decorative
layers
of the
reverse
paintings
on
glass
of the house
altar,
the
paper
covering
the reverse sides of the
glass panels
had to be
removed,
together
with the silver foil.
Unexpectedly, up
to 30% of the
paint
was found to be stuck to the foil
(Figure 10).
The
edges
of some of the flakes were not
sharp
and thus seemed to have softened due to heat. The
paint
transfer was
probably
caused
by exposure
to heat
when the house altar had to be dried after the flood at
Corning
in 1972. The most
time-consuming part
of the
conservation work was the transfer of
paint
back onto
the
glass.
Paint flakes were
trapped
under
lifting paint
and needed to be retrieved. Where
possible,
each flake
of
paint
was reattached in its
original position.
At the
beginning
it seemed that the decoration had suffered
extensive loss. After conservation
only
minor losses of
not more than 1-2% were
apparent (Figure 2).
Since there are
many
different
types
of reverse
paint
ing
on
glass,
each one
requires
an individual
approach
to
its treatment. Several conservation materials have been
tested within the last 20
years
to
identify
suitable materials
for the adhesion of
flaking paint
to a
glass
substrate,
such
as
acrylic
resin,
polyvinyl
acetate,
hydrocarbon
resin,
ketone
resin, cellulose,
natural
resin,
animal
glue
and wax
[9].
The
requirements
for the
consolidating
material are
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
218 S.
BRETZ,
U.
BAUMER,
H.
STEGE,
J. VON MILLER AND D. VON KERSSENBROCK-KROSIGK
(b)
Figure
8
(a) Light microscopical image
of a blue cross-section, 200*.
The
sample
was
extremely
brittle and difficult to
polish.
Residues of the
polishing
material in surface fractures render the
sample
whitish,
(b)
Back
scattered electron
image
of the blue
paint layer
at
higher magnification
showing
a
highly degraded binding
medium between the smalt
particles.
? Doerner Institut, Munich.
as follows: the
coating
has to
provide
excellent adhesion
to the
glass
and should have some
flexibility (adequate
transition
temperature, Tg).
It should not have
any
visible
impact
on the artefact. The refractive index should be
equal
to the
paint
medium to minimize reflection at the
interface. It must be
compatible
with
organic
materials
to avoid discolouration of the
original paint.
The
advantage
of solvent-based consolidant
systems
for reverse
paintings
on
glass
is their
ability
to bond
the delaminated
paint
in low concentrations as a
pre
Figure
9 Crucifixion, digitally
reconstructed colour
appearance.
? Andrew M. Fortune, Corning
Museum of Glass, Corning (NY).
liminary application.
Where loose
paint fragments
have
been
dislodged
and have to be
reintegrated
in the
paint layer,
resin solutions are
very
useful to make them
adhere to the
glass again.
The consolidation resin used
for this
purpose
on the house altar
paintings
was the
hydrogenated hydrocarbon
resin
Regalrez
1094? in a
20%
(w/v)
solution in white
spirit.
The
disadvantage
of
such a conservation material is that after the
evaporation
of the solvent a
high proportion
of voids become
visible after a
period
of time. In order to eliminate
this,
a
subsequent
treatment with an adhesive-like wax was
experimented
with on delaminated reverse
paintings
on
glass
and
proved
to be successful. In the case of the house
altar a sheet of silicone
paper
was coated with molten
TeCerowax 30445
(a microcrystalline hydrocarbon
wax)
and
placed
face down on the
damaged paint layer.
Pressure and heat
applied by
a heated
spatula
needed
to be
adjusted depending
on the level of
damage.
The
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
A GERMAN HOUSE ALTAR FROM THE SIXTEENTH CENTURY 219
Figure
10 Detail of St
Dorothy: (a)
reverse side
during conservation; (b) large parts
of the
paint layer
are attached to the silver foil and
paper.
? Simone Bretz, Munich.
softening
of the
original
resin medium
during
thermal
treatment enabled the
closing
of
gaps
between
paint
islands and the
flattening
of concave
flakes,
yielding
a
uniform
paint layer
which was not achievable
by using
a solvent-based consolidant alone. A brilliant saturation
of the
original paint
was the result. A thin wax
coating
was left on the
paint layer,
which can be reactivated
by
reheating
in case of further delamination from the
glass.
Finally,
the conserved
glass panels
were set into their
original positions
and the
ebony
veneer frames were
reattached with animal
glue.
Pieces of
missing
wood
were
replaced
with new
ebony
veneer.
CONCLUSION
The technical and scientific
research,
conducted
prior
to the conservation treatment of the reverse
paintings
on
glass
of the
Corning
altar
gave
valuable
insight
into
the still rather
little-investigated amelierung technique.
The colourless
glass panels,
which are not
significantly
altered,
are of
good quality
and
might
well be a
regional product.
The
binding
medium is a mixture of
various resins with
good
adhesion
properties,
which is
important
for its
application
to a
non-porous
substrate
like
glass. Quite unexpectedly,
no
drying
oils or
gums
were
found,
and
only
blue
samples
contained some
egg.
The most abundant
components
derive from
aged
and oxidized
pine
resin
(colophony),
which had most
likely
been heat-treated. The
processed pine
resin was
modified
by
the addition of different
softeners, i.e.,
larch
turpentine (also
known under the trade name 'Venetian
turpentine'),
mastic and
camphor.
In some
layers
remnants of
spike
oil were
preserved.
The identified resin
mixture is therefore
composed
of
regional
materials
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
220 S.
BRETZ,
U.
BAUMER,
H.
STEGE,
J. VON MILLER AND D. VON KERSSENBROCK-KROSIGK
(pine
resin,
larch
turpentine)
and
imported,
exotic resins
(mastic, camphor).
For the
colourants,
this situation is
much the same: the artist used
costly
and exotic
goods
(cochineal, dragon's
blood,
high-quality
noble
metals)
as well as
cheaper
and
locally
available materials
(smalt,
copper
resinate or
verdigris, lamp black).
The colour
changes
observed on the reverse
paintings
on
glass
most
heavily
affect the
formerly
blue and
green parts
of the
composition.
These are due to chemical reactions of
the
pigments
in
conjunction
with
degradation
of the
resinous medium.
Fading
was observed in the red
organic
colourants. It is
hoped
that future scientific research on
reverse
paintings
on
glass
will allow discussion of the
technical and material
particularities
of the
Corning
house altar in a broader context.
The
technique
and state of
preservation
of each
reverse
painting
on
glass
is
unique
and thus
requires
an individual
approach
for the choice of material and
treatment. The
speciality
of the
amelierung technique
which
only
uses
lacquers
enabled the conservation as
described. For the house altar
panels
the use of a
single
consolidant would have been
preferable.
However,
without
any pre-fixation
of the loose
paint particles
with the resin
(Regalrez 1094?,),
the
reintegration
of the
flakes back onto the
glass support
would not have been
possible.
The
fragments
would have been
displaced by
the treatment
using
molten wax
(TeCerowax 30445)
alone.
Only by combining
these two conservation
materials could a consolidated and saturated
amelierung
paint layer
be achieved.
APPENDIX: EXPERIMENTAL DETAILS OF
MATERIAL
ANALYSES;
MATERIALS
AND SUPPLIERS
Sample
treatment for
binding
media
analysis
Samples
were
pre-treated
for GC-MS
by step-wise
extraction with different solvents:
isooctane, methanol,
chloroform/methanol
(7:3 v/v),
oxalic acid in methanol
(10% w/v) (all
solvents of
gas chromatogaphic grade,
supplied by
Merck
chemicals,
and
anhydrous
oxalic
puriss by
Fluka
company).
The solvent extracts were
injected
into the GC-MS without derivatization.
Aged
resins are unstable even in common solvents and show
considerable
compositional changes
within a few
days;
therefore the extracts were
analysed
within a few hours.
The most
straightforward
identification of natural resins
is
by
non-acidic marker
compounds
that can be detected
in the underivatized methanolic extract. Resins are also
very
sensitive to unsuitable
preparation procedures,
and drastic
methylation methods,
in
particular
those
using strong
acid or basic
catalysts,
should be avoided.
Therefore,
anhydrous
oxalic acid
(5?10% w/v)
was
added to the methanolic extract in an additional
analysis
step.
In the
gas chromatogram
of these acidified solu
tions,
non-acidic
diterpenes
are
strongly
attenuated and
no
longer
detectable,
while
diterpenoid
acids elute with
fairly good response
factors and are
separated
well. In
that
step,
resins with a
high
amount of
diterpenoid
acids,
such as
pine
resin and
colophony,
can be identified. The
approach
described here does not
provide
a full
analysis
of all chemical
compounds
in a
sample.
Instead it is
optimized
to obtain
patterns
and marker
compounds
that enable the identification of the
specific types
of
natural resins that are used in art
[30].
Moreover,
similar
resins can be
distinguished
in
mixtures,
and conclusions
on treatments
during
the
manufacturing process (e.g.
thermal
treatments)
can be drawn even from
aged
samples.
Proteinaceous
binding
media were identified based
on their amino acid
composition.
In order to release
amino acids from
proteins,
the
sample
residues from
the solvent extraction described above were
hydrolysed
with
boiling
HCl
(6 )
for 24 hours under vacuum. If
necessary,
metal ions from
pigments
were
precipitated
with
8-hydroxy quinoline
at
pH
9 and removed
by
filtration with a C18
cartridge.
The amino acids were
collected
by
an acidic buffer and
separated
in an amino
acid
analyser (see below).
Gas
chromatography
Gas
chromatography
was
performed using
an
Agilent
GC 6890 with
autosampler equipped
with a DB-5ht
column
(-40
to
400?C,
15 m, 0.32 mm
ID,
0.1 film
thickness,
J
& W
company).
Measurement conditions
were the
following:
carrier
gas
helium 5.0
(purified
with
heated
high-capacity gas purifier, Supelco company),
constant-flow
mode,
1.7 mL
per minute;
split/splitless
injector: injection temperature
=
250?C,
splitless
mode,
1
.
injection volume,
0.5 minutes
splitless; purge
flow
36 mL
per minute;
oven
programme:
TI
=
55?C,
tl
=
1
minute,
Rl
=
11?C
per minute,
T2
=
150?C,
R2
=
10?C
per minute,
T3
=
360?C,
t3
=
5
minutes;
flame
ionization detector
(FID),
detector
temperature
=
360?C.
Gas
chromatography-mass spectrometry
GC-MS
analysis
was
performed using
a
gas
Chromato
graph
of the Hewlett-Packard
(now Agilent)
HP 5890
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
A GERMAN HOUSE ALTAR FROM THE SIXTEENTH CENTURY 221
series II
coupled
to a Hewlett-Packard
quadrupole
mass
spectrometer type 5989B,
MS
Engine,
in the
electron
impact (El) mode,
scan
range
40-500 m/z.The
Chromatograph
was
equipped
with a DB-5ht column
(5%-phenyl-methyl-polysiloxane),
30 m, 0.25 mm
ID,
0.1 film thickness
(J
&
W).
Measurement conditions
were as follows: carrier
gas
helium 5.0
(purified
with
heated
high-capacity gas purifier, Supelco company),
constant-flow
mode,
1.7 mL
per minute;
split/splitless
injector: injection temperature
=
250?C,
splitless
mode,
1?2
.
injection volume,
0.5 minutes
splitless; purge
flow 36 mL
per minute;
oven
programme:
Tl
=
55?C,
tl
=
1
minute,
Rl
=
14?C
per minute,T2
=
150?C,R2
=
10?C
per minute,
T3
=
360?C,
t3
=
5
minutes;
US
National Institute of Standards and
Technology (NIST)
library,
version 05.
Amino acid
analyses
For amino acid
analyses (AAA),
ion
exchange liquid
chromatography
with
post-column
derivatization
(reagent ninhydrin, supplied by
Merck
chemicals)
and
photometric absorption
detection
(570
nm, 440
nm)
was
applied using
a Biotronik LC 5001 with
autosampler
and
a
separation
column with cation
exchange
resin
(resin
type
BTC
2710,
210 mm
length,
Biotronik
company).
Measurement conditions were as follows:
hydrolysate
buffer
system
with sodium citrate buffers
(pH 2.2-13),
injection
volume 50 ., flow rate 30 mL
per
minute. For
a full
description
see
[41].
High performance liquid chromatography
The red lake was identified
using
a Hewlett-Packard
1050 HPLC
system consisting
of a 7125
injection
system (20
.
sample loop), gradient pump,
diode
array
detector 1100 with microcell and HP Chemstation.
UV-VIS
spectra
were measured between 200 and 600
nm; detection
wavelength
was set at 275
nm, reference
wavelength
at 550 nm
(band
with 4 and 100
nm).
For
separation
a C18 column
(column type HyperClone
5 u ODS
(C18)
100
x
2.00 mm 5
micron,
supplied
by Phenonemex)
was used. Eluents were distilled
water/0.1% trifluoroacetic acid
(A)
and acetonitril/0.1%
tetrafluoroacetic acid
(B).
Gradient elution A:B from
95:5 to 5:95 within 40
minutes,
then
kept
for 5 minutes.
Flow rate 0.8 mL
per
minute
(method
after
Halpine
[42]).The sample
was dissolved in oxalic acid/methanol
(10% w/v),
the reference
samples
in methanol/HCl
(1% v/v).
Preparation
of cross-sections
Paint cross-sections were embedded in Technovit? 2000
LC
(UV-hardening acrylic
resin,
supplied by
Heraeus
Kulzer)
under
water-jet
vacuum,
pre-ground
with Leco
Grit 400 and 1200
(Leco corporation)
and
polished
with
Micro-Mesh? cushioned abrasives
(supplied by
Dick
GmbH)
down to
grade
12000.
Light microscopy
Light
and fluorescence
microscopy
was carried out on an
Zeiss
Axioskop
with filter 'DIC for direct
light. Images
were taken with a
digital
AxioCAM MRc
microscope
camera and AxioVision 3.1 software
(Zeiss).
Scanning
electron
microscopy/X-ray microanalysis
SEM-EDX
analyses
were undertaken with a
Philips
(now
FEI
company)
XL 20 electron
microscope
with an EDAX
(now AMETEK) Si(Li)
detector and
Genesis software
(EDAX
/
AMETEK)
at the
following
measurement conditions: 25
kV,
30
A,
1000
cps,
100 s
live-time,
standardless ZAF
quantification.
Micro
X-ray
fluorescence
spectrometry
For
energy-dispersive
XRF,
a
portable
Mikro TAX
spectrometer (Bruker company)
with Mo-microfocus
tube
(30 W), poly capillary, peltier-cooled
SDD detector
and helium
purging
was made available
by
the Federal
Institute of Material
Testing
and
Research,
Berlin.
Measurement conditions: 45
kV,
600
A,
100 s live
time,
semi-quantitative
calibration with various standard
glasses
of
NIST,
The
Corning
Museum of Glass and
the Federal Institute for Material
Testing
and Research
(BAM).
Materials for conservation and
paint
trials
Smalt
(No. 10000), copper
resinate
(No. 12200)
and
Regalrez
1094?: Kremer
Pigmente, www.kremer-pigmente.de
TeCerowax 30445:Tromm
Wachs, www.wax-tromm.de
ACKNOWLEDGEMENTS
The authors would like to
acknowledge
the conservation
treatment
grant approved by
The
Getty
Foundation and
are
very grateful
for the
support.
The authors would like
to thank Cornelia
Tilenschi,
Karin
Lutzenberger
and
Irene
Fiedler,
Doerner Institut in Munich for
carrying
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
222 S. BRETZ,
U.
BAUMER,
H.
STEGE,
J. VON MILLER AND D. VON KERSSENBROCK-KROSIGK
out the
SEM/EDX,
HPLC and amino acid
analyses;
Oliver
Hahn,
Federal Institute for Material
Testing
and Research
(BAM)
in Berlin for his
support
with
the XRF measurements in
Munich;
and Andrew M.
Fortune,
assistant
photographer
and
digital image special
ist at The
Corning
Museum of Glass for the
digital
rendering
of the house altar. Sandra
Davison,
private
glass
conservator in
Thame, UK,
is
warmly
thanked for
valuable corrections and comments on the
manuscript.
REFERENCES
1
Ryser, F.,
Verzauberte Bilder. Die Kunst der Malerei hinter Glas
von der Antike bis zum 18.
Jahrhundert,
Klinkhardt &
Biermann,
Munich
(1991).
2
Ryser, F.,
Reverse
Paintings
on Glass:The
Ryser Collection,
ed. and
transi. R.
Eswarin,
The
Corning
Museum of
Glass, Corning
(1992).
3
Ryser, F., 'Hinterglasmalerei
und die Kunst des "Amelierens":
Ein
Begriff
im Wechsel der
Jahrhunderte',
Restauro 1
(1994)
26-31.
4
Ryser, F.,
'Die Kunst zu
amelieren',
in
1
Amalierte Stuck
uff
Glas/ Hinder Glas
gemalte
Historien und Gem?ld'.
Hinterglaskunst
von der Antike bis zur
Neuzeit,
ed. F.
Ryser
and B.
Salmen,
Schlo?museum Murnau
(1995)
25-46.
5
Bretz, S.,'Maltechnik
und Glastechnik in der
Hinterglasmalerei
1600 bis
1650',
in
Farbige
Kostbarkeiten aus Glas: Kabinettst?cke
der Z?rcher
Hinterglasmalerei
1600-1650,
ed. H. Lanz and L.
Seelig, Bayerisches
Nationalmuseum
M?nchen,
Schweizerisches
Landesmuseum
Z?rich,
Munich and Zurich
(1999)
181-219.
6
Bretz, S.,
and
Ryser, F.,'Kleines
Handbuch der
Hinterglasmalerei
(Petit
manuel de la
peinture
sous
verre)',
in Glanzlichter. Die
Kunst der
Hinterglasmalerei,
Mus?e Suisse du Vitrail
Romont,
Museum in der
Burg Zug (2000)
289-315.
7
Hahn, O., 'Untersuchungen
von Farbmitteln an
Hinterglas
bildern des 18./ 19.
Jahrhunderts',
in welche zuweilen
Kunstwerth
haben.', Hinterglasmalerei
in
S?dbayern
im 18. und 19.
Jahrhundert,
Schlo?museum,
Murnau
(2003)
107-110.
8
Koller,J.,
and
Baumer, U.,'Bindemittelanalysen
von
Hinterglas
bildern des 18.
Jahrhunderts',
in welche zuweilen Kunstwerth
haben.', Hinterglasmalerei
in
S?dbayern
im 18. und 19.
Jahrhundert,
Schlo?museum,
Murnau
(2003)
112-114.
9
Bretz, S.,'Information
on the
history, technology,
deterioration
and restoration of reverse
paintings
on
glass',
in Conservation and
Restoration
of Glass,
ed. S.
Davison, Butterworth-Heinemann,
Oxford
(2003)
341-343.
10
Volk, P., 'Hinterglasmalerei
aus dem
Bayerischen
National
museum',
Kunst &
Antiquit?ten
2
(1988)
52-63.
11
Pettenati, S.,
/ vetri dorati
graffiti
e i vetri
dipinti,
Museo
Civico,
Turin
(1978) 52-53,
no.
89,
plate
XIII,
95-98.
12
M?ller, ., Farbloses Glas im Wandel der Zeit:
Materialanalytische
Untersuchungen
an
farblosen
Gl?sern des 13. bis 17.
Jahrhunderts
mit
Hilfe
der Laser Induced Breakdown
Spectrometry
und der Mikro
R?ntgen?uoreszenzanalyse,
PhD
thesis,
Technische Universit?t
Berlin
(2006) 120-136,194-215.
13
Wedepohl,
.H.,
Glas in Antike und Mittelalter: Geschichte eines
Werkstoffes,
E. Schweizerbart'sche
Verlagsbuchhandlung N?gele
und
Obermiller, Stuttgart (2003).
14
Haff, C,
'Pigmente
und Bindemittel der
Hinterglasmalerei
nach 1600 und deren
Analysen',
in
Farbige
Kostbarkeiten aus Glas:
Kabinettst?cke der Z?rcher
Hinterglasmalerei
?600-?650,
ed. H.
Lanz and L.
Seelig, Bayerisches
Nationalmuseum
M?nchen,
Schweizerisches Landesmuseum
Z?rich,
Munich and Zurich
(1999)
220-242.
15
Koller, J.,
and
Baumer, U.,
'An
investigation
of the
"Lacquer
of the West": A methodical
survey',
in
Japanese
and
European
Lacquerware,
ed. M.
K?hlenthal,
Arbeitsheft 96 des
Bayerischen
Landesamtes f?r
Denkmalpflege, Lipp Verlag,
Munich
(1999)
339-348.
16
Sandermann, W,'Studien
?ber Harze,VII.
Mitteilung:
?ber die
Hitzeisomerisierung
der Harzs?uren und die
Zusammensetzung
technischer
Harze',
Fette und
Seifen
49
(1942)
578-585.
17
Pastorova, I.,
van den
Berg, K.J., Boon, J.J., andVerhoeven,J.W,
'Analysis
of oxidised
diterpenoid
acids
using thermally
assisted
methylation
with
TMAH',
fournal of Analytical
and
Applied
Pyrolysis
43
(1997)
41-57.
18
Ribechini, E.,
Study of Amorphous Organic
Residues in
Archaeological Findings by
Mass
Spectrometric Techniques,
PhD
thesis,
Universit? di Pisa
(2005) 5, 44-45,
61-66.
19
Koller, J., Baumer, U, Kaup,Y, Schmid, M.,
and
Weser, U,
'Effective mummification
compounds
used in
pharaonic
Egypt: Reactivity
on bone alkaline
Phosphatase', Zeitschrift f?r
Naturforschung 58b(5) (2003)
462-480.
20
Koller, J., Baumer, U., Grosser, D.,
and
Walch,
.,
'Turpentine,
larch
turpentine
and Venetian
turpentine',
in
Baroque
and
Rococo
Lacquers,
ed.
J.
Koller and K.
Walch,
Arbeitsheft 81 des
Bayerischen
Landesamtes f?r
Denkmalpflege, Lipp Verlag,
Munich
(1997)
359-378.
21
Scalarone, D., Lazzari, M.,
and
Chiantore,
.,
'Ageing
behaviour
and
pyrolytic
characterisation of
diterpenic
resins used as
art materials:
Colophony
and Venice
turpentine', Journal of
Analytical
and
Applied Pyrolysis
64
(2002)
345-361.
22
Hever, J., 'L?rchenharzung',
in Vom
Ri?
zum Rohharz
-
Das
Ende einer
forstlichen Nutzung
in der
ehemaligen
DDR, Staatliches
Naturhistorisches
Museum,
Braunschweig (1992)
80-87.
23
Koller, J., Baumer, U., Grosser, D.,
and
Schmid, E., 'Mastic',
in
Baroque
and Rococo
Lacquers,
ed.
J.
Koller and K.
Walch,
Arbeitsheft 81 des
Bayerischen
Landesamtes fur
Denkmalpflege,
Lipp Verlag,
Munich
(1997)
347-358.
24
K?hn, H.,
Erhaltung
und
Pflege
von
Kunstwerken,
3rd
edn,
Klinkhardt &
Biermann,
Munich
(2001)
363-364.
25
Braun, M., Verbesserung
der
Arzneibuchvorschriften
und ihre
Angleichung
and das
Europ?ische
Arzneibuch am
Beispiel
von
?therischen
?len,
Dissertation Naturwissenschaftliche Fakult?t
IV,
Universit?t
Regensburg (2002)
74-83.
26
Koller, J., Fiedler, I.,
and
Baumer, U.,
'Vermeers Maltechnik
-
eine Mischtechnik.
Untersuchung
der Bindemittel auf dem
Gem?lde Bei der
Kupplerin',
in
Johannes
Vermeer
-
Bei der
Kupplerin,
ed. U. Neidhardt and M.
Giebe,
Michael Sandstein
Verlag,
Dresden
(2004)
65-75.
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
A GERMAN HOUSE ALTAR FROM THE SIXTEENTH CENTURY 223
27
Schweppe H.,
and
Roosen-Runge, H., 'Carmine
-
cochineal
carmine and kermes
carmine',
in Artists'
Pigments:
A Handbook
of
Their
History
and
Characteristics,,Vol. 1,
ed. R.L.
Feller, Cambridge
University Press, Cambridge
and
Washington (1987)
255-283.
28
Eastaugh, N., Walsh, V.,
Chaplin,
T.,
and
Siddall, R., Pigment
Compendium:
A
Dictionary of
Historical
Pigments,
Elsevier
Butterworth-Heinemann,
Oxford
(2004)
118f. and further
references.
29
Baumer, U.,
and
Koller, J.,
'Drachenblut:
Herkunft, Analyse
und
Verwendung',
in abstract volume of the
Jahrestagung f?r
Arch?ometrie und
Denkmalpflege,
ed. U.
Sch?ssler,
R.
Fuchs,
Cologne, September (2001)
151-153.
30
Baumer, U., Dietemann, P.,
and
Koller, J.,
'Identification of
resinous materials on 16th and 17th
century reverse-glass
objects by gas chromatography/mass spectrometry',
International
Journal of
Mass
Spectrometry, published
online 30 October
2008,
doi:10.1016/j.ijms.2008.09.010,
in
press.
31
Stege, H.,'Out
of the blue? Considerations on the
early
use of
smalt as blue
pigment
in
European
easel
painting', Zeitschrift f?r
Kunsttechnologie
und
Konservierung
1
(2004)
137?139.
32
Gratuze, B., Soulier, I., Blet, M., andVallauri, L.,'De l'origine
du
cobalt: du verre ? la
c?ramique',
Revue d'Arch?om?trie 20
(1996)
77-94.
33 Van Eikema
Hommes, M.,
Changing
Pictures: Discoloration
in
15th?17th-Century
Oil
Paintings, Archetype Publications,
London
(2004)
20.
34 Van de
Graaf, J.A.,
Het De
Mayerne manuscript
als bron voor
de schildertechniek van de
Barok,
PhD
thesis,
Rijksuniversiteit
Utrecht,Verweij,Mijdrecht (1958)
174
(no. 68).
35
K?hn, H., 'Verdigris
and
copper resinate',
in Artists'
Pigments:
A Handbook
of
Their
History
and
Characteristics,Vol. 2,
ed. A.
Roy,
National
Gallery
of
Art, Washington (1993)
148-158.
36
Saunders, D.,
and
KirbyJ.,'Light
induced colour
changes
in red
and
yellow
lake
pigments',
National
Gallery
Technical Bulletin 15
(1994)
79-97.
37 Lewerentz K., Drachenblut:
Materialkunde, quellenkundliche
Studien zur maltechnischen
Verwendung
und
Analysem?glichkeiten
der verschiedenen Farbharze dieses
Namens,
Diploma
thesis,
University
of
Applied
Science
Cologne (2001) 75-83, Figure
5
in
Appendix.
38
Woudhuysen-Keller R.,
and
Woudhuysen
P.,
'Thoughts
on
the use of the
green glaze
called
"Copper
resinate" and its
colour
changes',
in
Looking Through Paintings,
ed. E.
Hermens,
A. Ouwerkerk and
.
Costaras,
de Prom and
Archetype
Publications,
Baarn and London
(1998)
133-146.
39 Van den
Berg, K.J.,
Eikema Hommes M.H. v., Groen, K.M.,
and
Berrie,
.
.,' copper green glazes
in
paintings',
in Art et
chimie,
la
couleur, ed.J. Goupy
and
J.-P. Mohen,
CNRS
Editions,
Paris
(2000)
18-21.
40
Spring, M., Higgitt, O,
and
Saunders, D, 'Investigation
of
pigment-medium
interaction
processes
in oil
paint containing
degraded smalt',
National
Gallery
Technical Bulletin 26
(2005)
56-70.
41
Fiedler, I.,
Anorganische
Biochemie und Kunst: Zur Stabilit?t
proteinhaltiger
Matrices in
Gegenwart anorganischer Pigmente,
PhD
thesis,
Eberhard-Karls-Universit?t
T?bingen (2001)
34-38.
42
Halpine,
S.,'An
improved dye
and lake
pigment analysis
method
for
high-performance liquid chromatography
and
diode-array
detector',
Studies in Conservation 41
(1996)
76-94.
AUTHORS
Simone Bretz is a
private
conservator based near
Munich,
Germany.
Since 1985 she has
specialized
in
the conservation and restoration of reverse
paintings
on
glass,
and has
acquired
a
unique knowledge
of the
painting technology
and
published
on the
subject.
She has worked for a number of
major
museums and
private
collections in
Europe
and the United States and
contributed her
expertise
to research
projects
both in
Germany
and abroad. Address:
Schmiedeweg 28,
82496
Oberau, Germany.
Email:
info@bretz-hinterglas.com
Ursula Baumer received a
degree
in
chemistry
from the
Chemieschule
Elhardt, Munich, Germany.
She
joined
the Doerner Institut in 1988 and is
working
as a research
chemist,
responsible
for
organic analysis.
She
specializes
in the
analysis
of ancient and modern
binding
media of
paintings
and works of
art,
archaeological
resins and
tars,
lacquers
and varnishes on furniture and altars. Address:
Doerner
Institut, Bayerische Staatsgem?ldesammlungen,
Barer Strasse
29,
80799
Munich, Germany.
Email:
baumer@doernerinstitut.
de
Heike Stege received her
diploma (1994)
and PhD
(1998) degree
in
chemistry
from the Technical
University
of
Berlin, Germany.
She has a
postgraduate degree
(habilitation
in
2005)
in the field of modern
analytical
techniques especially
suited to works of art made of
glass
and enamel. Since 2002 she has
specialized
in
pigment
research at the Doerner
Institut,
where she is
currently
head of the scientific
department.
Address: as Baumer.
Email:
stege@doernerinstitut.
de
Johannes
von Miller has been
working
as a
private
conservator of furniture and wooden
objects
for more
than 15
years.
He
specializes
in the research of historic
lacquer
and the reconstruction of historic surfaces of
wooden
objects.
Address:
Dorfplatz 9,
83707 Bad
Wiessee,
Germany. Email:johannes@vonmiller.
de
Dedo von Kerssenbrock-Krosigk is head of the
Glasmuseum Hentrich at the
Stiftung
museum kunst
palast,
D?sseldorf, Germany.
After
receiving
his doctorate
from Humboldt
University, Berlin, Germany
he worked
at the Br?han-Museum in
Berlin,
and from 2004 to 2008
as curator of
European glass
at the
Corning
Museum of
Glass in
Corning,
NY,
USA. Address:
Stiftung
museum
kunst
palast, Ehrenhof 4-5,
40479
D?sseldorf, Germany.
Email:
dedo.krosigk@museum-kunst-palast.de
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions
224 S.
BRETZ,
U.
BAUMER,
H.
STEGE,
J. VON MILLER AND D. VON KERSSENBROCK-KROSIGK
R?sum?
?
Lors de la
pr?paration
du traitement de conservation d'un autel
domestique
de la
fin
du XVIe si?cle du sud de
l'Allemagne,
au
Corning
Glass
Museum,
orn? de
peintures fix?es
sous
verre,
on a men? des ?tudes
techniques
et
analytiques
sur
le substrat de
verre,
les
liants,
les colorants et les ?l?ments
m?talliques.
Un examen au
microscope
a r?v?l? les
particularit?s
de la
technique appel?e amelierung
et de la
fa?on
dont le verre a ?t?
manufactur?.
Pour
l'analyse
des
mat?riaux,
la
Chromatographie
gazeuse coupl?e
? la
spectrom?trie
de
masse,
l'analyse
des acides
amin?s,
la
Chromatographie liquide
haute
performance,
la micro
spectrom?trie
de
fluorescence
des
rayons
X et la
microscopie ?lectronique
?
balayage coupl?e
? la
spectrom?trie
des
rayons
X ont
?t? utilis?es. Le
principal composant
des liants ?tait une r?sine de
pin chauff?e (colophane), modifi?e par l'ajout
de diverses
r?sines
assouplissantes.
La couche de
peinture rouge
contient de la cochenille et de
petites quantit?s
de
sang
de
dragon.
Les zones
originales
bleues et
vertes,
qui
contiennent du smalt et du r?sinate de
cuivre,
ont subi
d'importants changements
de couleur. Les
peintures fix?es
sous verre ?taient en
pi?tre
?tat. En raison de l'humidit? tr?s ?lev?e
(?
la suite d'une inondation au
mus?e)
et
des m?canismes de
vieillissement,
la couche de
peinture
?tait s?rieusement
soulev?e, fractur?e
en centaines de
petites
?cailles et
partiellement
coll?e ? la
feuille d'argent
et au
fond
en
papier.
Les
fragments
de
peinture
ont ?t? remis en
position
sur le verre
? l'aide d'un
fin pinceau
et adh?r?s ? l'aide d'une r?sine
hydrocarbure.
Ce traitement a ?t? suivi de
l'application
d'une cire
microcristalline
qui
a
permis
d'obtenir une saturation brillante des
peintures fix?es
sous verre.
Cependant,
le
changement
de
couleur subsistait dans les zones bleues et vertes.
Apr?s
le traitement on a
manipul?
des
images num?riques pour
restituer la
tonalit?
d'origine pr?sum?e.
Zusammenfassung
?
Zur
Vorbereitung f?r
die
Konservierung
eines s?ddeutschen Hausaltares mit
Hinterglaseinlagen
aus dem
sp?ten
I6.
Jahrhundert
wurden technische und
naturwissenschaftliche Untersuchungen
des
Glases,
der Binde- und Farbmittel
sowie der
Metallfolien durchgef?hrt.
Die
mikroskopische Untersuchung zeigte
die Besonderheiten der so
genannten Amelierungs
Technik und der
Herstellung
der
Glastafeln auf.
F?r die
Materialanalysen
wurden
Gaschromatographie
/
Massenspektrometrie,
Aminos?ureanalyse, Hochleistungsfi?ssigkeitschromatographie
sowie
Mikror?ntgenfluoreszenzspektrometrie
und Ras t?relektron?
nmikroskopie
mit
energiedispersiver R?ntgenmikroanalyse
verwendet. Als
Hauptbestandteil
des Bindemittels wurde ein thermisch
vorbehandeltes
Kiefernharz (Kolophonium) nachgewiesen,
das durch Zusatz verschiedener Weichmacherharze
modifiziert
wurde.
In den roten Malschichten wurden Cochenille und
geringe
Reste von Drachenblut
gefunden.
In ehemals blauen und
gr?nen
Bereichen,
die Smalte bzw.
Kupferresinat enthalten,
waren betr?chtliche
Farbver?nderungen festzustellen.
Die
Hinterglas tafeln
waren in schlechtem
Erhaltungszustand: Wegen
hoher
Luftfeuchte (?berflutung
des
Museums),
vermutlich
nachfolgender
W?rmebehandlung
und
Materialver?nderungen
war die
gesch?digte
Malschicht
gro?fl?chig
vom Glas
abgel?st,
in zahlreiche lose
Farbpartikel
zerbrochen und klebte an den
Hinterlagen
aus
Silberfolie
und
Papier.
Nach dem
Transfer
der
Malschichtfragmente
in ihre
originale
Position
auf
dem Glas und
Fixierung
mittels eines
Kohlenwasserstoffharzes und,
in einem zweiten
Schritt,
mit
einem mikrokristallinen
Wachs, zeigten
sich die
Hinterglastafeln
wieder strahlend.
Allerdings
blieben die
Farbver?nderungen
in den
blauen und
gr?nen
Partien irreversibel. Ein Photo nach der
Konservierung
der
Mitteltafel
des Hausaltars wurde
digital bearbeitet,
um die
angenommene urspr?ngliche Farbigkeit
zu visualisieren.
Resumen
?
Como
fase previa
a los
procesos
de restauraci?n de un altar de devoci?n
privada procedente
del sur de Alemania
del
siglo
XVI
tard?o, que incluye
siete
pinturas bajo vidrio,
se realiz? una
investigaci?n
t?cnica
y cient?fica
del substrato
vitreo,
de
los
aglutinantes,
metales
y
colorantes. El examen
microsc?pico inform?
sobre la t?cnica llamada
uamelierung" y
la
manufactura
de
los
paneles
de vidrio. Para los an?lisis de materiales se
emplearon
las
siguientes
t?cnicas:
cromatograf?a
de
gases-espectrometr?a
de
masas,
an?lisis de amino
?cidos, cromatograf?a l?quida
de alta
resoluci?n,
as? como micro
espectrometr?a por fluorescencia
de
rayos
X
y microscop?a
electr?nica de barrido combinada a microan?lisis
por energ?a dispersiva
de
rayos
X. El
principal componente
del
aglutinante
se
identific?
como resina de
pino precalentada (colofonia), modificada
con otras resinas
"plastificantes".
La
capa roja
de
pintura
contiene cochinilla
y peque?as
cantidades de
sangre
de
drag?n.
Las ?reas
que
en
principio
deb?an haber sido azules
y verdes, que
se
componen
de esmalte
y
resinato de
cobre,
han
sufrido
un dr?stico cambio crom?tico. Las
pinturas bajo
vidrio se
encontraban en un delicado estado de conservaci?n: debido a los altos niveles de humedad
(inundaciones
en el
museo) y
ala
degradaci?n
de los
materiales,
las
capas pict?ricas
se encontraban
gravemente exfoliadas, fracturadas
en cientos de
peque?as
escamas
y parcialmente
adheridas a la l?mina de
plata y
a la trasera de
papel.
Las
part?culas
de
policrom?a desprendidas
se
transfirieron
a sus localizaciones
originales
con un
pincel fino y aseguradas
mediante la
aplicaci?n
de una resina hidrocarbonada. A ello
sigui?
la
aplicaci?n
de cera micro-crisialina con la
que
se
consigui?
una buena saturaci?n de las
pinturas bajo
vidrio. Sin
embargo
se
mantuvo la decoloraci?n de las ?reas azules
y
verdes. Una
fotograf?a
de
despu?s
del tratamiento
fue manipulada digitalmente para
recrear el
probable
cromatismo
original.
STUDIES IN CONSERVATION 53
(2008)
PAGES 209-224
This content downloaded from 84.205.227.38 on Mon, 18 Aug 2014 08:48:57 UTC
All use subject to JSTOR Terms and Conditions