Identification and Removal of Nonoriginal Layers in the

16th Century Paintings of Funchal’s Cathedral Altarpiece.

Sofia Gomes(1), Carolina Ferreira(2,3), Glória Nascimento(2,3), Luís Piorro(1),

Ana Cardoso(3), António Candeias(1,3) and Mercês Lorena(1) *

(1) Conservation Department, Laboratório José de Figueiredo, Direção-Geral do Património Cultural, Rua das
Janelas Verdes 37, Lisboa, Portugal
(2) Research Centre for Studies in Final Arts, Faculdade de Belas Artes da Universidade de Lisboa, Largo da
Academia Nacional de Belas Artes, Lisboa, Portugal
(3) Chemistry Department, HERCULES laboratory, Évora University, Largo Marquês de Marialva 8, Évora,
Portugal

Abstract: This article is an outcome of the study, decisions and restoration work undertaken
during the integrated conservation project of the 16th century paintings of Funchal’s
Cathedral main chapel’s altarpiece, in Madeira Island, Portugal. After diagnostic research
has allowed a complete understanding of the original material of the 12 Portuguese-Flemish
paintings, it has been revealed the presence of other superimposed layers, whether actual
overpaints or the accumulation of materials occurring in time over the old ones.
Identification together with the cleaning and removal of these coatings was a delicate and
complex process. Conservators already suspected the existence of overpaints in some
paintings, a close observation of the painted surface showed differences in terms of color,
texture, and execution technique in specific areas. Chemical and stratigraphical research
together with radiography and historical research on primary documental sources provided
important clues about the paintings of the previous interventions and relevant and crucial
information for the decision making.

Keywords: Restoration; Conservation; Painting; Overpaint; Cleaning.

1. The paintings history

The Funchal altarpiece ordered by King Manuel I is the only Portuguese remaining
monumental altarpiece from the 16th century that is still in its original location. The 12
paintings of Funchal’s Cathedral main chapel’s altarpiece had a very unsatisfactory
appearance: the surface was very dark because of the many layers of grime, coatings, and
materials from the previous interventions. Obviously, the paintings had endured numerous
interventions. Since the end of the 16th century to the 19th century, the altarpiece was
submitted to “repaintings”, “renovations” and “washings” that resulted in changes and
additions to its original surface1.
Since the 17th century in Europe, varied and unexpected products and methods are described
in cleaning recipes in handbooks and painting treatises. Soap, bleach, egg white, lime water,
butter, ashes and even urine were some of the products used to clean the paintings at this
time2. Hazards of painting cleaning and the concern about the consequences of cleaning are

1
also mentioned in some of these treatises (Pacheco3), the use of acids and alkali mixtures in
antique restorations resulted in many overcleaned painting and irreparably damages.
Furthermore, with the purpose to reviving colors, the paintings were impregnated with
cooked oils, varnish layers and bitumen2, which formed a heterogeneous and darkened layer
preventing the display of the paintings’ original color.
In addition to these kinds of layers, overpainting was a common practice between 16th and
18th centuries within the Portuguese context, usually with the aim to adjust to different
stylistic trends, to promote iconographic alterations, or to disguise the degradation of the
paintings. These interventions, which were commissioned by the clergy and nobility and
trusted to painters, were a recurrent practice of this time and nearly always considered a
noble activity4.

2. Nonoriginal layers identification

All paintings, except four, which were restored for an exhibition in Lisbon in the 1940’s,
presented a very dark and thick coating of several materials accumulated during time and
from past treatments (Fig. 1). Like a black curtain, this superficial layer hides the visual
reading of the paintings, obscuring the aesthetic message. Through scientific analysis, it was
possible to identify, in this superficial layer, different kinds of materials such as animal glue,
different oils, resin, oxalates (degradation products resulting from biodeterioration), metal
carboxylates (degradation products), calcium carbonate, calcium sulfate, and silicates. Wax
was also detected, especially in the lower two central panels, which had greater proximity to
the altarpiece candles. Also, relevant was the identification of styrene, and naphthalene,
which indicate the possible use of a bituminous material from a past treatment mainly applied
to the Virgins drapery, covering the modeling. This dark layer of grime, dirt, and old
degraded organic varnishes was reducing the paintings to a kind of monochrome surface,
distorting the original colors.
Different types of total or partial overpaints were identified, with the application of
nonoriginal color layers, motivated either by the paintings’ degradation or by the need to
adapt to different stylistic trends and iconographic representations. Obvious differences of
texture and technical quality in both original and nonoriginal layers were visible to the naked
eye. Overpaint was applied on top of paint losses, directly on the wooden support and losses
of the overpainted layer showed the original underlying surface.
Three paintings showed overpaints (Fig. 2): a partial one in the Descent from the Cross
painting, located in the third level of the altarpiece, motivated by an iconographic change;
and two total overpaints that, while keeping with the theme, show composition and detail
changes, in the Last Supper and the Mass of Saint Gregory paintings, located in the first
level, drove by the paintings’ degradation and/or by the need to adapt to different stylistic
trends. These two types of overpaints were probably made in different times and/or by
different painters.
The paintings’ original compositions were revealed by radiography, one of the examination
and imaging techniques used to explore the multiple layers of the paintings. This technique is
capable of recording hidden layers lying below the visible image and it was an important
examination in the identification of the overpaints. The image obtained is a result of the
interaction between the X-rays and the paintings’ materials and depends not only on the
physical thickness of the materials, but also on their atomic number5. Pigments, such lead
white, the most common white pigment used until the 19th century, are capable of absorbing
X-rays owing to the lead’s high atomic number. In contrast, pigments of low atomic number,

2
such as carbon black, do not absorb the X-rays strongly6. Lead white and carbon black
pigments were identified in most of the microsamples, mixed with other colors to create
effects of light and shadow on the paintings. As a result, light areas on the paintings are
mostly light on the radiographs owing to the presence of lead white, producing a readable
image and giving additional information that were not visible in visible light.
The radiographic images of these three paintings have revealed many changes hidden below
the upper paint layers. While keeping the theme, the radiographic image of the Last supper
painting shows objects hidden and changes on some of the figures’ positions when compared
with the normal light image, proving that this painting was reworked and overpainted (Fig.
3). The X-ray also provided information about the support, on how panels with multiple
planks were joined by a series of wooden dowels and showed dark patches on the painting
surface, with sharply defined edges, which correspond to paint losses.
Additionally, paint layers and pigments identification also provided important clues. For
instance, Prussian blue (Fe4[Fe(CN)6]3), an artificially manufactured pigment available to
artists after 1724, was identified in nonoriginal layers of the three paintings7. The FTIR
spectra (Fig. 4) shows a very strong C=N absorption at about 2089 cm-1, a characteristic
absorption peak of Prussian blue. Only the presence of this pigment gave a possible
indication of the date of the intervention, and the other pigments identified in the overpaints
were in common use at this time. Also, the cross-section of the microsamples had shown that
the overpaints were applied over a dark organic layer, probably a varnish, which separates it
from the original one (Fig. 4).

3. Decision making
In contemporary conservation, there are no general rules on how to make such decisions
when handling later additions to the pictorial artwork and these must be considered case by
case8. The removal of such layers requires reflection and discussion in order to establish their
importance and must be justified by historical and aesthetic criteria since this procedure
greatly modifies the appearance of the artworks irreversibly. Most often, the justification for
the removal of an overpaint layer is to recover the painting to a condition closer to its original
state9.
The information gathered through radiographic examation and the identification of the paint
layers’ composition together with the cleaning tests that exposed areas of the original paint
layer provided relevant information, which supported the decision making. Also, considering
the religious function of the paintings and their place of display as well as acknowledging the
reasons that led to these alterations, it was obvious that the painting’s current appearance was
a very inaccurate representation of the artist’s intent and that the viewer was being distracted
by the surface grime, dirt, and old varnishes, that had no longer their aesthetic and protective
functions, and deceived by the overpaints, which were falsifying the aesthetic message of the
paintings. With the conviction that these coatings were actually detrimental to the paintings
and that the chromatic and stylistic changes and new figurations that altered the original
paintings were depriving from their pictorial text without adding anything to it, but altering
its content. Conservators, chemists, art historians, clergy and cultural heritage authorities
decided to clean and remove the overpaints as they were acting negatively and that the value
present in the overpaints could be sufficiently preserved through a photographic
reproduction. A total cleaning would expose the true state of the paintings and would
improve access to information about the paintings. The restoration work aimed to restore the
potential unity of the work of art, as Brandi writes this is the “methodological moment of

3
recognition of the work of art in its physical consistency and its dual aesthetic and historical
polarity, with a view to its transmission to the future”10.

4. Overpaint removal
Visible to the naked eye, the Descent from the Cross painting showed shapes and textured
brush strokes of an underlying figure that did not match the visible painting surface. A cross-
section from the area where the overpaint, representing the floor, superimposes the lying
Virgin’s mantle shows a thin brown nonoriginal paint layer on top of a thick blue original
paint layer (Fig. 5), proving that the standing Virgin Mary was not an original.
With radiography, it was possible to further determine the extent of the overpaint and the
existence of hidden figures. A clear radiographic image of two new women was obtained due
to the density differences of both black thin nonoriginal upper paint layer and white under
paint layer (Fig. 6). Overpaint was covering the two women depicted – a suffering Virgin
Mary lying on the ground held by Mary Salome, a large portion of this painting was
overpainted in order to change the scenes’s iconography. The mechanical removal of a thin
black layer composed of pitch in the Descent from the Cross painting revealed the two
women figures (Figure 6).
This iconography change follows the Council of Trent standards and the imposed Stabat
Mater model (the Virgin Mother of God standing up, controlling her pain)4. The change in
the taste and sensibility of the times rejected the dramatic and sorrowful image of the Virgin
Mary. This was a common practice after the 17th century when many paintings were
destroyed, concealed, replaced or overpainted in churches throughout the country9. The poor
technical quality of the coarse overpaint contrasted with the rest of the original paint surface
and therefore it was decided to remove it (Fig. 7).
On the Last Supper and the Mass of Saint Gregory paintings, conservators have already
suspected the existence of a total overpaints, keeping the original iconographic theme with
some detail changes. These overpaints were probably made as an update to taste of the time
and to cover the paint surface damages. According to art historians the changes occurred
between the 17th and 18th centuries. They point out that the nature of the total overpaint is
similar to that found in the vast majority of the Portuguese painting from the reigns of King
Manuel and King João III11. Their location on the first level of the altarpiece made them more
accessible and vulnerable to damages, such as candle burns and wax deposits as well as paint
losses. Additionally, differences in aesthetic and material characteristics were also evident.
During the overpaint removal process, the poor condition of the original paint surface became
increasingly obvious. Large losses and several candle burns of the paint layer had caused
irreversible damage, which became visible after the removal of overpaint (Fig. 8). This
procedure also revealed details of the original composition such objects or different positions
of the original figures.
In both the paintings, the overpaint colors were, for the most part, similar to the original ones
but darker and thinner and were separated from the original ones by a dark organic layer,
probably an old varnish (Fig. 9). However, with the removal of the nonoriginal layers in the
Last Supper painting, the Christ’s purple cloth became blue, the figure with brown cloth
became white, and the dark blue sky revealed a light green. Additionally, in the Mass of Saint
Gregory painting, the green baldachin became white and the yellow/red dark areas from the
modeling of the priest’s clothing, which were opaque and flat got volume and dimension and
became blue.

4
The paintings gained considerably with the removal of the overpaints: the modeling of the
three-dimensional clothing became evident as the technical quality of the original paint layer
was revealed and the original composition was restored to its original vibrant and lighter
colors (Figs. 10 and 11).

5. Conclusion
The integrated conservation and restoration project presented contributed to the History of
Art through the technical study of the paintings, its materials, and manufacturing processes,
as well as for the safeguard of this masterpiece by the identification of the alteration and
degradation processes. The decisions of cleaning and removing overpaints were taken with
the aim that with restoration the paintings could gain a better state physically and
aesthetically.
The cleaning of the black layer of surface grime and the removal of the overpaints brought to
light substantial differences, their removal from the colorful paintings were dramatic and the
artist’s original palette and composition are visible again. Furthermore, the previous research
of the technique and materials of the paintings was a valuable input to the understanding of
the pictorial set and to the decision of the removal of overpaints, making the restoration work
far better informed.
Cleaning, together with the removal of the overpaint additions, must allow not only the
recovery of a presumed ‘original’ aspect of color, but also a reconstruction of the correct
relationships between colors and tones, recovering the legibility of the work of art while
respecting both the aesthetic and the historic characteristics. With this in mind, it was obvious
that not doing nothing was not an option. Considering the 12 paintings as a ‘total work of art’
(Gesamtkunstwerk concept) where each painting are combined together into a single, and
unified ensemble, leaving these additions would misrepresent and conceal the visual reading
as a whole.

1. Rodrigues R. A Pintura Proto-Barroca e Barroca no Arquipélago da Madeira entre 1646 e

1750. A eficácia da imagem, texto policopiado, tese doutoral apesentada à Universidade da
Madeira, 2012, coorientada por Vitor Serrão e Isabel Santa Clara Gomes Pestana.
2. Miguel AMM. Historia de la conservación y la restauración desde la antigüedad hasta el
siglo XX. Madrid: Editorial Tecnos; 2002.
3. Pacheco, F. Arte de la pintura. Madrid : Instituto de Valencia de Don Juan, 1956.
4 Serrão, V. «Renovar», «repintar», «retocar»: estratégias do pintor-restaurador em Portugal,
do século XVI ao XIX. Razões ideológicas do iconoclasta destruidor e da iconofilia
conservadora, ou o conceito de «restauro utilitarista» versus «restauro científico». Conservar
Património. Nº 3-4; Dezembro 2006. p. 53-71.
5 MacBeth, R. The technical examination and documentation of easel paintings. In: Stoner
JH, Rushfield R, editors. Conservation of easel painting. Oxon: Routledge; 2012. p. 291-305.
6 Hassal, C. Paintings. In: Lang J, Middleton, A, editors. Radiography of Cultural Material.
Oxford: Butterworth-Heinemann; 1997. P. 98-116

5
7. Gettens RJ, Stout GL. Painting Materials. A short encyclopedia. New York: Dover
Publications, Inc.; 1966.
8. Goltz M, Stoner JH. Consideration on removing or retaining overpainted additions and
alterations. In: Stoner JH, Rushfield R, editors. Conservation of easel painting. Oxon:
Routledge; 2012. p. 147-158.
9. Alves AN. De Santa Catarina a… São Vicente… Questões em torno do levantamento de
repintes na pintura portuguesa. As preparações na pintura portuguesa dos séculos XV e XVI.
In: Serrão V, Antunes V, Seruya AI, coordinators. As preparações na pintura portuguesa.
Séculos XV e XVI – Actas do Colóquio Internacional. Faculdade de Letras da Universidade
de Lisboa; Agosto 2013. p.225-234.
10. Brandi C. Teoria do Restauro. Amadora: Edições Orion; 2006.
11. Projecto integrado de conservação e restauro do retábulo da capela-mor do Funchal. Acta
da Reunião dos Historiadores – Repintes. Lisboa, 15 Julho de 2013.

6
Figures

Fig. 1. The Resurrection painting in normal light with a very dark layer on surface, which
was obscuring the visual reading. (a) N2 microsample location in the painting. (b) N2 cross-
section: 1 ground; 2 yellow original paint; 3 dark grime surface layer.

Fig. 2. Paintings with overpaints. (a) The Last Supper, (b) the Mass of Saint Gregory
paintings with a total overpaint, and (c) the Descent from the Cross painting with a partial
overpaint.

7
Fig. 3. The Last supper painting in normal light and details of the painting in X-ray. (a)
Radiographic image shows the changes in the figures’ hands position, the dark patches of
paint losses where paint and ground layers have flacked away and the vertical planks joined
by a wooden dowel. (b and c) The objects hidden by the overpaint.

Fig. 4. Identification of the Prussian blue in the Last Supper painting. (a) B4 microsample
location in the painting; (b) FTIR spectra of the B4 microsample upper layer number 4 from
the Last supper sky. A characteristic IR absorption (2089 cm-1) band [Fe(C=N)6]3-; (c) B4
cross-section: 1 ground; 2 original layers; 3 dark organic layer; 4 blue overpaint. The

8
overpaint layer is over a dark organic layer, probably an earlier varnish from a previous
restoration.

Fig. 5. The Descent from the Cross painting in normal light. (a) Detail, during the overpaint
removal and M21 microsample location; (b) M21 cross-section: 1 ground; 2 original paint
layers; 3 dark organic layer; 4 light brown overpaint; 5 dark grime surface layer.

Fig. 6. The Descent from the Cross painting details. (a) Overpaint, nonoriginal Virgin Mary;
(b) two new figures revealed by radiography; (c) during the removal of overpaint.

9
Fig. 7. The Descent from the Cross painting in normal light. (a) Before and (b) after cleaning
and the removal of overpaint.

Fig. 8. Detail of the Mass of Saint Gregory painting. (a) Overpainted book; (b) candle burn,
after the removal of overpaint.

10
Fig. 9. The Mass of Saint Gregory painting in normal light and location of the C10
microsample. (a) Detail during the removal of the overpaint showing the several layers
superimposed to the original one: 2 original layers; 3 dark organic layer; 4 overpaint layer; 5
dark surface grime. (b) C10 cross-section: 1 ground; 3 dark organic layer; 4 overpaint layer; 5
dark grime surface layer.

Fig. 10. The Last Supper painting in normal light. (a) Overpainted painting; (b) after cleaning
and the removal of overpaint.

11
Fig. 11. The Mass of Saint Gregory painting in normal light. (a) Overpainted painting; (b)
after the removal of overpaint.

12