NATIONAL

GALLERY
TECHNICAL
BULLETIN
30th anniversary volume, 2009

National Gallery Company

London

Distributed by
Yale University Press
This volume of the Technical Bulletin has been funded by the American Friends of the National Gallery,
London with a generous donation from Mrs Charles Wrightsman

Studying Old Master Paintings – Technology and Practice: The National Gallery Technical Bulletin 30th
Anniversary Conference is supported by The Elizabeth Cayzer Charitable Trust

Series editor Ashok Roy

© National Gallery Company Limited 2009

All rights reserved. No part of this publication may be

transmitted in any form or by any means, electronic or
mechanical, including photocopy, recording, or any
information storage and retrieval system, without the
prior permission in writing of the publisher.

First published in Great Britain in 2009 by

National Gallery Company Limited
St Vincent House, 30 Orange Street
London wc2h 7hh

www.nationalgallery.org.uk

British Library Cataloguing in Publication Data

A catalogue record for this journal is available from
the British Library

isbn 978 1 85709 420 6

issn 0140 7430
525051

Project manager Jan Green

Editor Diana Davies
Editorial assistance Giselle Osborne
Designer Heather Bowen
Picture research Giulia Ariete
Production Jane Hyne and Penny Le Tissier
Repro by Alta Image, London

Printed in Italy by Conti Tipocolor

front cover
Details from Aelbert Cuyp, The Large Dort (plate 1, page 71);
Jean-Baptiste-Camille Corot, The Roman Campagna, with the
Claudian Aqueduct (plate 1, page 89); Sassetta, Sansepolcro
Altarpiece (plate 1, page 8; plate 3, page 10); Diego Velázquez,
Christ after the Flagellation contemplated by the Christian Soul
(plate 1, page 53); Jean-Baptiste-Camille Corot Villeneuve-lès-
Avignon (plate 4, page 91), Sebastiano del Piombo, The Raising
of Lazarus (plate 1, page 27)

title page
Sebastiano del Piombo, The Raising of Lazarus, detail
Photographic credits

All photographs reproduced in this Bulletin are

© The National Gallery, London unless credited
otherwise below.

THE HAGUE Museum Mesdag © Photo SCALA,

Florence: p. 94, pl. 8

INDIANAPOLIS © Indianapolis Museum of Art,

Indiana: p. 91, pl. 4

LONDON © The Trustees of The British Museum.

All rights reserved: p. 74, pls. 4 and 5. © V&A Images /
Victoria and Albert Museum, London: p. 29, pl. 2

MADRID Monasterio de la Encarnación, Madrid ©

Patrimonio Nacional, Madrid: p. 66, pl. 20. © Museo
Nacional del Prado, Madrid: p. 57, pl. 10; p. 58, pl. 13;
p. 66, pl. 21; p. 68, pl. 23; p. 69, pl. 25

NEW YORK Wildenstein & Co., Inc New York ©

Private Collection, U.S.A. Photograph Courtesy of the
Wildenstein Archives: p. 100, pl. 14

PARIS © Bibliothèque Nationale de France, Paris:

p. 102, figs. 5 and 6. © Collection Frits Lugt, Institut
Néerlandais, Paris: p. 74, pl. 3. Musée du Louvre, Paris
© RMN, Paris / Photo Jean-Gilles Berizzi: p. 79, pl. 12.
Musée du Louvre, Paris © RMN, Paris / Photo René-
Gabriel Ojéda: p. 94, pl. 9

SAN LORENZO DE EL ESCORIAL Monasterio

de San Lorenzo de El Escorial © Patrimonio Nacional,
Madrid: p. 68, pl. 22

SAO PAULO © MASP, Museu de Arte de São Paulo

Assis Chateaubriand, São Paulo, Brazil / Photo João L.
Musa: p. 56, pl. 8

VATICAN CITY, ROME Vatican Museums,Vatican

City © akg-images / Nimatallah: p. 45, pl. 38

WASHINGTON, DC © Image courtesy of the Board

of Trustees, National Gallery of Art, Washington, DC:
p. 101, pl. 21

Private Collection: p. 106, pl. 27

 Sebastiano del Piombo’s Raising of Lazarus:
A History of Change

jill dunkerton and helen howard

Commission and early history for Lazarus – still extant – and probably also for the

T he huge panel painting showing the raising of

Lazarus from the dead, as recounted in the Gospel
of Saint John (plate 1), was commissioned by Cardinal
figure of Christ.4
The progress of the altarpiece can be tracked from
Sellaio’s and Sebastiano’s letters.5 On 26 September
Giulio de’ Medici from Sebastiano del Piombo, prob- 1517 Sellaio informed Michelangelo that Sebastiano
ably towards the end of 1516.1 Shortly before, it appears had stopped work on other commissions to concentrate
that the cardinal had ordered from Raphael, Sebastiano’s on the altarpiece. In January 1518 Michelangelo
great rival, a painting of the Transfiguration, which was himself was briefly in Rome and saw the Lazarus. In
to be on a panel of the same dimensions. Both were July Sebastiano reported to Michelangelo that he was
destined to be set beneath the tall gothic arches of the delaying completion because he did not want Raphael
cathedral of St Just in Giulio’s bishopric of Narbonne. to see it until Raphael had finished his own painting,
Letters sent to Michelangelo (who had left Rome apparently not even begun at this stage. In addition there
in December 1516 for Florence and Carrara) by his was an issue about the frame, which Sebastiano wished
friend and assistant Leonardo Sellaio, as well as some to have made in Rome, whereas Raphael was trying
from Sebastiano himself, are an important source of to influence the cardinal to have it made in Narbonne,
information for the competition between the two perhaps to avoid a public confrontation between the
painters (at least from the point of view of the Sebastiano altarpieces. Nevertheless, in the spring of 1519 the Raising
party). The letters also include occasional references to of Lazarus was placed on view, probably in Sebastiano’s
practical details of the making of the altarpiece. We workshop; according to Sellaio ‘everyone was stunned’
learn from a letter from Sellaio dated 19 January 1517 (‘ogni uomo resta balordo’). By December the painting
that Sebastiano was responsible for arranging for the had been varnished6 and the huge panel transported
carpentry of the panel and that he had received funds for to the Vatican for a more formal presentation, where,
that purpose.2 It seems that the panel was constructed according to the Venetian diarist, Marcantonio Michiel,
from very long boards joined vertically, like those for it was much praised by all, including the Pope.7 This
Raphael’s Transfiguration (see plate 38), rather than in was followed by wrangling over Sebastiano’s large bill.
the horizontal arrangement that would be expected in a Meanwhile Raphael had at last made progress on The
Venetian panel of comparable dimensions. Whether the Transfiguration, the monumental scale and colours of
boards were of the usual poplar or a more unusual species, the foreground figures clearly influenced by the work
such as the cherry which has been reported as the wood of his Venetian rival.8 Eventually, in April 1520, little
for The Transfiguration,3 can no longer be known. more than a week after Raphael’s death, both panels
In the same letter Sellaio told Michelangelo that were brought to the Vatican (we know from Sebastiano
he believed that Raphael was holding back on the himself that his panel had to be transported again,
execution of his altarpiece in order to avoid direct and had not simply remained there9) and exhibited
comparison with Sebastiano, and, by implication, for together. According to Vasari the two works received
fear of the borrowing of his inventions. Sebastiano equal praise.10
did indeed take inspiration from recent projects of The Transfiguration remained in Rome, set up on the
Raphael’s, for example the Tapestry Cartoons and The high altar of San Pietro in Montorio, and eventually,
Way to Calvary (‘Lo Spasimo di Sicilia’) (Madrid, Prado), following a spell in Paris between 1797 and 1816, where
but in designing the Raising of Lazarus he showed it was restored but mercifully with minimal intervention
himself ready for the challenge of arranging a grand- to the panel,11 it was placed in the Pinacoteca of the
scale narrative with attendant crowds while constrained Vatican. The Raising of Lazarus, on the other hand,
by the vertical format of the altarpiece. In this he was followed a more precarious course over the centuries.
famously aided by Michelangelo who supplied drawings It is not known exactly when it left Rome; nor are any

26 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

plate 1 Sebastiano del Piombo, The Raising of Lazarus (NG 1), 1517–19. Oil on synthetic panel, transferred from original panel, 381 × 289.6 cm.

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 27

Jill Dunkerton and Helen Howard

details known about its transport to Narbonne, which maker (and, it would seem, the inventor of the cradle),
was presumably by sea, since the French city was still a was probably better than most, carving away the wood
port at the time. It seems that Sebastiano had his way of the panels from the paint layers using planes and
over the making of the frame in Rome for the lower chisels (still the preferred method in those exceptional
section of a frame, of highly sophisticated design with cases when transfer is unavoidable20). This technique
gilded ornament against a blue ground, still survives on was certainly safer than that of his near contemporary
the altar in St Just which now holds the copy of The Robert Picault who had a ‘secret’ method, which seems
Raising of Lazarus made in the eighteenth century by to have involved separation of the paint from the panel
Carl van Loo.12 The presence of Sebastiano’s altarpiece by breaking down the ground layer through prolonged
in Narbonne in the sixteenth century is confirmed by exposure to nitric acid vapours; this allowed him to
reflections of the composition in French painting of the display intact the original wooden support alongside the
period and in 1599 it was the subject of an appreciative transferred painting.21 The showmanship that was part
description by a Swiss physician, Félix Platter, who of the process of transfer meant that its consequences
mentioned its great value and much-copied status.13 were recognised by connoisseurs of painting such as
The painting remained on the itinerary for other Richard Payne Knight, who seems to have known
visitors to the city until 1722 when it was acquired by The Raising of Lazarus before its transfer. He claimed
Philippe, Duc D’Orléans, regent of France, in exchange many years later that ‘those who have only seen it since
for a grant for the repair of the cathedral and the copy that fatal operation of cutting away the pannel [sic] on
by van Loo still on the altar.14 The original panel was which it was painted, and gluing cloth to the back of
moved to Paris, where it joined the duke’s magnificent the colour in its place, can form but very imperfect
collection in the Palais-Royal.15 notions of what it was before.’ 22
There can be little doubt that this drastic and
Transfer and re-transfer dangerous intervention was totally unnecessary. The
The Raising of Lazarus is described as ‘peint sur bois’ distribution of damage to the painting shows that,
in Du Bois de Saint Gelais’s 1727 catalogue of the given its great size, the panel had remained remarkably
pictures at Palais-Royal; he also records the colours of stable, with evidence for the opening up of only one
several draperies, among them the ‘jaune clair’ of the of the vertical joins, that running through the standing
kneeling Magdalen. Since this is a colour that would figure of Martha and the right leg of Lazarus. As well
be indistinguishable from a darker yellow, or indeed as fashion, a reason for its transfer might well have
from white, if the painting were covered with a heavily been the size and weight of the panel – a report of
discoloured varnish, it has to be assumed that the 1749 concerning the proposed transfer of the two most
painting was reasonably clean and visible at the time.16 famous Raphaels from the French Royal Collection,
In the 1770s, when the palace and its collection were Saint Michael and The Holy Family of François I, which
in the hands of Louis-Philippe D’Orléans, grandson of were regularly moved between the royal apartments
Philippe, a systematic programme of restoration of the at Versailles and the picture stores, observed that ‘sont
paintings was undertaken. This included the transfer peints sur bois, ce qui, joint à leurs cadres, les rend
to canvas of many of the paintings on panel, including d’un poids prodigeux et par consequent très difficiles
The Raising of Lazarus. According to National Gallery à manier, ou les transporter’.23 Given the readiness of
records this was carried out, or at least begun, by Sebastiano and the cardinal to have the panel moved
‘Haquin’ in 1771.17 The treatment is likely to have been back and forth from the Vatican in the early sixteenth
protracted. The Haquin referred to must have been century, it would be a strange irony if this were one of
Jean-Louis Hacquin (before 1726–1783) rather than his the reasons for the transfer three centuries later.
son, François-Toussaint (1756–1832).18 Particularly after Payne Knight was not alone in his criticism of the
1775 Jean-Louis was also responsible for the transfer procedure and consequences of transfer 24 but the fact
and lining of a great many paintings from the French that a painting had been transferred from its original
Royal Collection, now in the Louvre.19 It was believed support seems to have had little effect on its value. In 1793,
that by transferring the paint layers from unstable and following a sequence of sales and changes of ownership,
perhaps worm-eaten wooden panels to new canvas The Raising of Lazarus came to London with other
supports the future preservation of the works would be Italian paintings from the Orléans collection, eventually
ensured. In practice, a great deal of damage was caused being put up for sale in 1798, when it was bought by
to the paintings, involving at best a complete alteration the insurance underwriter John Julius Angerstein for
to the paint texture and at worst, the loss of large areas the considerable sum of 3,500 guineas; this was a higher
of the picture surface. Hacquin, originally a cabinet valuation than that of many now celebrated paintings

28 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

by Titian and Raphael from the same collection and of the old varnish was removed (suggesting that the
was surpassed only by Annibale Carracci’s The Dead previous cleaning was no more than a partial thinning),
Christ Mourned (‘The Three Maries’) (NG 2923).25 In followed by a light revarnishing. At this point the
part because of the association with Michelangelo, the restorers involved, Morrill and Holder, decided that the
altarpiece caused a sensation and much debate among altarpiece showed no signs of having been painted on
artists; its most vocal admirer was Benjamin West,26 who panel and that it must always have been on a canvas.31 In
in about 1820 is supposed to have been responsible November 1939, when the painting had been evacuated
for the restoration and repainting of parts of Lazarus’s to Penrhyn Castle in Wales, a ‘sizeable’ flake loss was
damaged right leg.27 West’s intervention was recorded noted and the following year areas of the surface were
in the Manuscript Catalogue of the newly founded covered with facing paper in order to secure loose paint.
National Gallery, centred on the core collection of 38 Further attempts to secure the flaking paint were made
pictures from the Angerstein collection which were in 1941 and again in 1951 but with little success.
acquired for the nation in 1824. The importance of Following the establishment of a Conservation
Sebastiano’s altarpiece was recognised by its being Department at the Gallery in 1946 the stabilisation of
assigned the first number in the new catalogue. The Raising of Lazarus became a priority. Eventually in
Over the next few decades several entries were 1958 it was decided that it should undergo a radical
made in the Manuscript Catalogue relating to the structural intervention with the aim of reducing the
painting’s condition. On only one occasion is the large amount of glue that was present as a result of the
support mentioned, in 1837, when it needed treatment application of several canvases to the reverse following
for infestation by insects; these apparently fed on the transfer. Raking-light photographs taken at the time
the glue of the lining adhesives and were probably illustrate the alarming extent to which the contraction
either flour or biscuit beetles – this outbreak at the of the glue was causing compression and lifting of the
National Gallery occasioned a short report in the paint film (FIG. 1). When treatment began, the intention
Observer of 19 September 1841. In general, there was was to remove three of the lining canvases, leaving a
greater preoccupation with the surface appearance and last canvas in place. The plan was that this would be
especially the varnish layers. Already by 1798 it was stretched out and the paint flakes secured once there
observed that many of the Orléans pictures appeared was sufficient space to reattach them. Unfortunately,
to have become ‘dirty, or more sunk in their colours’,28
although the Sebastiano appears surprisingly bright
and richly coloured in the watercolour by Frederick
Mackenzie of Angerstein’s pictures hanging in his Pall
Mall house (plate 2). Nevertheless, it needed varnishing
in 1834, 1852 and 1867. Two letters sent in 1865 by Sir
Charles Eastlake to the Keeper, Ralph Wornum, express
concern about the sunk and opaque condition of the
varnish, but Eastlake was emphatic that ‘no cleaning, in
the picture cleaner’s sense of the term, should on any
account take place’; instead ‘Pinti [Raffaelle Pinti, the
London-based Italian restorer most trusted by Eastlake]
should endeavour to tone down what is prominent and
crude, and in short to harmonize the whole.’29 In 1881
following an enquiry among the Trustees, ‘assisted by
artists and others’ who included the restorers Bentley,
Dyer and Pinti,30 it was agreed that the painting should
be cleaned and revarnished by Dyer, although the extent
of the cleaning is not known. Following this cleaning
the frame was fitted with an enormous sheet of plate
glass in order to protect the paint surface from the dirt
and pollution of nineteenth-century London.
In the twentieth century it was the structural
condition of the painting that caused the greater plate 2 Detail of a watercolour by Frederick Mackenzie showing
concern. In 1929 it was treated using a mixture of glue The Raising of Lazarus at Angerstein’s Pall Mall House (London,
and rye flour for ‘a large number of small blisters’. Some Victoria and Albert Museum).

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 29

Jill Dunkerton and Helen Howard

ground or ‘enduit de transposition’ – usually lead white

and oil – as used in later transfers, including those by
his son, François-Toussaint. In the case of The Raising of
Lazarus, however, there is no mention of there having
been any silk, but there is a new pinkish-brown ground,
present in all the paint samples that include the full
layer structure and consisting mainly of red earth, lead
white and a carbon black pigment.33 The gesso ground
was removed completely but in some of the samples a
thin layer of an unpigmented material, probably glue,
can be seen between the new ground and the original
paint layers. This was probably applied as part of the
transfer rather than being the remains of an application
FIG. 1 Raking-light detail taken in 1958 of the buildings in the of glue to seal the original gesso ground. At present
background of NG 1. too few of Hacquin’s transfers have been examined to
know whether he had a standard practice.The works by
Domenichino and Le Sueur referred to above already
had red-brown oil-based grounds (which were not
removed), and so Hacquin may not have seen the need
to supplement them with a second ground. With the
Sebastiano, on the other hand, the removal of the gesso
necessitated a new ground. An alternative and less likely
possibility (which would contradict Payne Knight’s
admittedly distant memories of the operation) is that
Hacquin’s transfer was not the first, and that the painting
had already been transferred by Picault who worked on
Orléans pictures in the 1750s.34 He did sometimes use a
red-brown ‘enduit de transposition’35 and the premature
FIG. 2 Macro detail taken following removal of the eighteenth- failure of some of his transfers meant that they had to
century transfer canvases, showing splinters of the original wood panel. be reattached to new supports only a few years later.36
Some record of such a major undertaking might be
expected, however, and the survival of a few splinters of
the wood of Sebastiano’s panel (FIG. 2) is more indicative
once the first three canvases had been removed, it was of a panel removal by mechanical means.
discovered that the fourth layer was not a canvas textile, The alarming discovery made in 1958 that the first
but instead consisted of sheets of paper. These were transfer layer was paper, and not canvas as expected,
badly decayed and in many areas had separated from meant that the surviving original layers were held
the paint and ground. together only by a tissue paper facing, applied using
It could not have been known at the time, but a mastic and wax adhesive. As the treatment report37
the use of paper to back a transferred paint film is candidly records, this was adequate for the planned
characteristic of a transfer by Hacquin. In his transfers operation but not for a total re-transfer. There was
of paintings by Domenichino and Eustache Le Sueur, no possibility of turning the painting over in order to
now in the Louvre,32 Hacquin used sheets of paper attach a more substantial facing and the original layers
cut from old printed books or manuscript texts. In the were found to be sensitive to water-based adhesives
report on the treatment of The Raising of Lazarus there (‘wrinkling and breaking into minute fragments’) and
is no mention of any text on the paper layer, although to the amount of heat needed to melt a solid wax-based
it may have been so stained and decayed that this was adhesive; eventually it was decided to brush on multiple
overlooked. In the other transfers by Hacquin to have thin layers of warm wax-resin dissolved in white spirit,
been investigated, two layers of fine silk (sometimes embedding a layer of inert terylene net fabric within
printed with a pattern) have been found embedded the layers as they solidified. Although the discoloured
in the mixture of glue and flour paste which lies varnishes had yet to be removed from the paint surface,
immediately behind the paint film, together with a the appearance of the picture was considered to be darker
residue of the original ground; he did not add a new than intended as a result of lack of reflectance from

30 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

fig. 3 The Raising of Lazarus,

photographed in 1967 after re-transfer
and cleaning, before restoration.

the ground and so titanium white (titanium dioxide) place before the introduction of lightweight and stable
was added to the wax cement. This bright white layer panels made from glass fibre with aluminium honeycomb
appears in some of the paint samples (for example plate cores.38 The painting is mounted, therefore, on a support
3). These were taken only after completion of the re- constructed with ‘sundeala’ composite board outer faces
transfer of the paint film and so the presence of the and a core of paper honeycomb.39 In spite of its wooden
red-brown ‘enduit de transposition’, which negates any edges and an internal wooden framework this panel
reflective properties of the new white ground, could is now showing signs of instability, with a tendency to
not have been known – presumably its brown colour flex and twist when the painting is moved, an operation
meant that previously it was taken to be a discoloured which is therefore avoided as far as possible.
old glue layer. With the paint film secure, removal of the old layers
Once the wax and titanium white layers had been of varnish could proceed. Judging by the extent of the
built up to a sufficient thickness the paint film could then discoloration visible in the patches of varnish that still
be mounted on a new solid support.This had previously remained on Martha’s dress and in the area of Lazarus’s
been coated with wax-resin allowing a bond to be shroud and the forearm of the man supporting him
achieved by ironing with a thermostatically controlled when the painting was photographed before retouching
iron to soften the wax-resin layers which then fused as in 1967 (FI FIG. 3), several layers of the notorious ‘gallery
they cooled. Although it is unlikely that these methods varnish’ (a mixture of mastic and drying oil) were
and materials would be used nowadays, the treatment present; Dyer’s cleaning in 1881 can therefore have
can be judged a success in that there has been no further involved no more than a partial varnish removal. Six
flaking of the paint layers. Unfortunately, the work took months of retouching then followed, but, considering

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 31

Jill Dunkerton and Helen Howard

FIG. 4 Digital infrared reflectogram

detail of Saint Peter.

all that the painting has been through, the amount of loss Painting technique
is less than might be expected.The many small scattered The condition of the painting following cleaning in
losses can be attributed to flaking, while certain patterns 1967, with losses located in all the principal colour
of damage atypical of panel paintings, for example the areas, allowed for the taking of an unusual number of
jagged lines through the group of bystanders in the paint samples. At the time some of these were mounted
background on the left, can be attributed to accidents in as cross-sections, which were used to identify the
the original transfer process. The delicate condition of range of pigments employed.40 No detailed study of
the painting meant that some of the older and relatively Sebastiano’s technique was undertaken. As is always the
insoluble restoration, including the repainting down case, however, the samples, in the form of cross-sections
the join that ran through the figures of Lazarus and and unmounted fragments of paint, were labelled and
Martha, was not removed in the most recent cleaning; stored, and it is this archive that some forty years later
where it was very discoloured it was covered by new provides the basis of the present study.41 The old cross-
retouching. The paint layers of some of the figures are sections have been re-polished and re-photographed and
damaged by abrasion, particularly in the lower part of many new cross-sections made from the unmounted
the picture (for instance the figures supporting Lazarus). samples. Fragments of unmounted sample have also
In common with other paintings of its age, it is likely been examined by Fourier transform infrared (FTIR)
to have been cleaned wholly or partially on several microscopy and analysed by gas chromatography–mass
occasions before its recorded conservation history; spectrometry (GC–MS), supplementing results of
the distribution of the damage suggests that these first gas–chromatography originally published in 1976.42
cleanings (including the removal of the varnish that we Pigments and other inorganic materials have been
know from the documents was applied by Sebastiano) identified by energy dispersive X-ray analysis (EDX) in
are likely to have taken place while it was still on the the scanning electron microscope (SEM) and Raman
altar in Narbonne. microspectroscopy (RAMAN).

32 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

Underdrawing, imprimitura and flesh painting

The size and fragility of the support means that
X-radiography with the equipment at present in use at
the National Gallery is not feasible; for similar reasons
it has so far only been possible to record an infrared
reflectogram image of a test area (a detail is shown in
FIG. 4).43 A full infrared study is likely to shed light
on Michelangelo’s contributions to the design of the
altarpiece in much the same way as with Sebastiano’s
Pietà, painted in 1512–15 (Viterbo, Museo Civico).44 The
character of the underdrawing detected on The Raising
of Lazarus is similar to that in the Pietà, but it appears
generally looser and with more revision, for instance plate 3 Cross-section of a sample from the green shadowed area of
the fingers of the right hand of the apostle in red and the Magdalen’s dress. The green layer contains a copper green pigment,
lead-tin yellow, lead white, yellow earth and a few large carbon black
green in the group kneeling or crouching on the left
particles. The original mid-brownish grey imprimitura of lead white, a
were drawn extending into the area now covered by coarse carbon black, brown earth, and a little lead-tin yellow is visible
Christ’s mantle. The underdrawing was executed with a under the green paint. All layers below this were added as part of the
brush, apparently over the broadly brushed imprimitura, various transfer processes. The red ground layer is presumed to have
hence the broken quality of the drawn lines, for instance been applied by Hacquin in the late eighteenth century, while the
around Christ’s right knee. Particles of charcoal appear titanium white ground was added to the reverse in 1958.

between the imprimitura and paint layers in some cross-

sections (for example plates 25 and 32) and are probably
connected with the underdrawing.
The broad horizontal brushstrokes used to apply
the imprimitura over the gesso are clearly visible in the
infrared image because of the black pigment present
in the layer. It might be thought that the marked
difference in tone between the upper and lower parts
of the detail illustrated (a difference that extends
further across the painting) was caused by a difference
in composition of the priming, as appears to be the
case with the Pietà, where the upper part has a white plate 4 Cross-section of a sample from the lightest blue of Christ’s
or very pale preparation and the lower part, comprising mantle. The light blue layer is composed of ultramarine combined
the foreground and the dead Christ, has a mid-grey with lead white over a pink underpaint of red lead with lead white.
Beneath this is the original brownish-grey imprimitura, over Hacquin’s
underlayer.45 In The Raising of Lazarus, however, cross-
red-brown ground. The colour of the imprimitura is similar to that
sections of samples from below and above this division shown in plate 3 though the two areas appear different in infrared.
(plates 3 and 4) confirm that the composition of the
priming is essentially the same, consisting of a medium-
rich mixture of lead white, a coarse carbon black brownish-grey colour, still light enough for lines of
pigment, a brown earth pigment and a little lead-tin black underdrawing to be clearly visible to the painter.
yellow; therefore the differences apparent in infrared Over the course of his career Sebastiano seems to have
must be the result either of variations in thickness favoured progressively darker preparations: The Daughter
of the application or the mixing of a new batch of of Herodias (NG 2493), painted in 1510, shortly before
priming mixture – highly likely for a painting of this he left Venice for Rome, has a very pale grey priming of
size – which contained a slightly lower proportion of lead white with a small amount of a fine carbon black,
carbon black pigment. The different explanations for probably lamp black,46 while from the 1520s onwards
what appears to be the same phenomenon in the two he tended to work on very dark grey painting surfaces,
paintings are valuable reminders of the difficulties in including unprimed slates – indeed it seems that he can
interpreting infrared images. be credited with the invention of this technique.47
Assessing the colour of an imprimitura on the basis Given that the priming is a mid-grey colour, this
of the pigments present and their appearance in cross- raises the question of whether the loss of the original
sections is not easy, but in the case of The Raising gesso and its replacement with the red-brown ‘enduit de
of Lazarus it is likely to have been a light to mid- transposition’ makes any optical difference to the paint

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 33

Jill Dunkerton and Helen Howard

layers, let alone the brilliant white ground added in the

re-transfer.48 Evidence from The Virgin and Child with
Saint Joseph and Saint John the Baptist and a Donor (plate
5), of 1517, when work was beginning on The Raising
of Lazarus, suggests that Sebastiano did not expect the
reflective properties of a white gesso ground to play any
part in the final effect. When he had a vertical plank
of wood added to the right edge to extend the panel
during the course of painting, he did not bother to
apply gesso to this addition, preparing it instead with
the same brownish-grey priming that he had used for
the imprimitura brushed over the gesso of the rest of the
panel.49 The priming (plate 6) is similar in composition
to that of The Raising of Lazarus, consisting of a medium-
rich mixture of lead white, black and a little lead-tin
yellow, and so in the infrared image (FIG. 5) it shows as
very dark where the gesso is absent.
plate 5 Sebastiano del Piombo, The Virgin and Child with Saint Joseph
and Saint John the Baptist and a Donor (NG 1450), 1517. Oil on panel,
Changes with time as a result of Sebastiano’s own
97.8 × 106.7 cm. choice of a non-reflective painting surface are evident in
more thinly and directly painted parts of the altarpiece,
where the inevitable increase with age in transparency
of paint mixtures containing little lead white means
that they no longer cover sufficiently the grey-brown
priming. As a result the darker colours appear to
merge with the priming, becoming ‘sunk’, and detail
can no longer be distinguished. The structure of the
plate 6 Cross-section of a sample taken from the extension to the rocky outcrop crowned with bushes and trees that rises
right edge of The Virgin and Child with Saint Joseph and Saint John immediately behind Lazarus and his attendants is now
the Baptist and a Donor, showing a priming of similar composition to barely legible, and indeed it can easily be mistaken for
that used in The Raising of Lazarus. The brownish-grey layer, which
an enormous tree. In areas that have been abraded or
is medium-rich and contains lead white, black and a little lead-tin
yellow, is especially thick since there is no gesso on the extension.
where the paint has broken up as a result of the transfer,
for example some of the heads on the right, there is a
similar loss of volume and detail.
It might also be thought that the colour of the
imprimitura is responsible for some of the very dark
flesh tones, especially of the male figures. However,
a sample from the highlight on the muscle along
the top of Lazarus’s left leg (plates 7 and 8) shows a
sequence of brown undermodelling layers, some of
them superimposed wet-in-wet, which contain brown,
yellow and red earth pigments with very little lead
white. These layers are sufficiently thick and opaque
to conceal completely the colour of the priming. The
highlight was then applied with a single layer of pale
yellow flesh tint based on lead white with yellow and
brown earth pigments. Since Lazarus has only just been
raised from the dead his skin is sallow, but a sample from
the bronzed arm of the muscular young man on the right
who supports him confirms that other flesh tones were
built up on the same system, but with a healthy colour
FIG. 5 The Virgin and Child with Saint Joseph and Saint John the Baptist achieved by the addition of red earth to the final layer.
and a Donor, infrared reflectogram detail.
There is likely to have been a loss of covering power in
the uppermost layers of the flesh tints, especially where

34 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

plate 7 The Raising of

Lazarus, detail.

they become thinner in the transitions from highlight

to shadow, which can result in abrupt juxtapositions
of light and shade, for example the sharp edge formed
along what should be the rounded muscles of Lazarus’s
upper arm.50 Nevertheless, the wide tonal range and
dramatic chiaroscuro in the modelling of flesh tints is
clearly intended, even if exaggerated by time.
Sebastiano’s adoption of chiaroscuro modelling
systems 51 and strongly directional lighting is, however,
far from consistent, even in his rendering of the many
heads and hands. The profile head of the kneeling
Magdalen, whose head ought to be in shadow, is pale
plate 8 Cross-section of a sample from the top of Lazarus’s left
and brightly illuminated, for obvious dramatic and
thigh. The highlight of the pale yellow skin-colour consists of lead
design purposes. Similarly, the fine idealised profile of white combined with yellow and brown earth pigments. This was
Saint John the Evangelist, the young standing apostle applied over a sequence of brown undermodelling layers of varying
on the left who gestures towards Christ and whose combinations of red, brown and yellow earth pigments with a little
gospel is the source for the miracle, stands out from the lead white.
swarthier heads around him.52
When it came to the painting of the many draperies,
Sebastiano seems to have been equally prepared to
abandon consistency of method, employing colour
modelling techniques that he brought from Venice in
combination with new approaches learnt in Rome.
Account has to be taken of the ways in which some
of the pigments have changed; nevertheless, it seems

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 35

Jill Dunkerton and Helen Howard

plate 9 The Raising of Lazarus, detail.

possible that Sebastiano set out to proclaim his Venetian of lead white (see plate 4). Since ultramarine in oil – in
origins by showing the Roman public (and Raphael) this case confirmed as linseed oil54 – has poor covering
the greatest and most subtly varied range of colours power when used without lead white, some form of
ever seen in a single painting. As a Venetian, he was also underlayer was needed, especially if its brilliance was
well placed to source an extensive range of the highest to be preserved when there was a grey painting surface.
quality pigments, even if, as is likely, he had to arrange The mantle therefore was underpainted with a pale pink
for them to be sent from his home city.53 mixture of red lake and lead white, apparently blocked
in without any preliminary modelling of the folds. The
Blue and purple draperies pink colour is clearly intended to intensify the rich
One pigment that was almost certainly procured via purple-blue of the lapis and Sebastiano used the same
Venice is the ultramarine blue of Christ’s mantle (plate technique for the Virgin’s mantle in The Virgin and Child
9).The superior quality of the lapis is best demonstrated with Saint Joseph and Saint John the Baptist and a Donor
in the shadows where it is used unadulterated with white (plate 5).55 The use of red lake to underpaint areas of
(plate 10); the highlights are modelled by the addition blue can be seen on the works of later Venetian painters,

36 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

plate 10 Cross-section of a sample from the deepest blue shadow

of Christ’s mantle. The uppermost layer consists of high-quality plate 12 Cross-section of a sample from Saint Peter’s pale greenish
ultramarine with just a little lead white. This has been applied over a lilac robe. At the base of the sample, the brownish-grey imprimitura is
pink layer of red lake with lead white. clearly visible. Over this is a thick layer of azurite combined with lead
white and a carbon black pigment. At the surface there is a thin layer
of lead white with red lake, now much faded, and a few tiny particles
of ultramarine.

plate 11 Cross-section of a sample from the shadow of Martha’s deep

mauve sleeve, showing a layer of red lake combined with ultramarine
and a little lead white over a strongly coloured undermodelling of
ultramarine combined with red earth.

plate 13 Cross-section of a sample from the highlight of Martha’s

sleeve, showing a pale pink undermodelling; over this are two layers of
ultramarine, red lake and lead white combined in varying quantities.
The red lake pigment in the uppermost layer has faded.

especially Veronese,56 but in early sixteenth-century there has been some loss of colour from Saint Peter’s
Venice it was more usual to underpaint ultramarine draperies, Martha’s violet dress at the centre of the
with the greener blue azurite. Pink underpaintings painting was probably always more intense in colour
for ultramarine have also been found in some Roman and it has a very different layer structure. Here relatively
works by Raphael, such as The Holy Family of François substantial layers of ultramarine, red lake and lead
I,57 but the possibility that he learnt the technique from white were applied over a strongly coloured purple-
Sebastiano is supported by its discovery in the Virgin’s red underpainting, now visible in many areas as a result
mantle in The Adoration of the Shepherds (Cambridge, of abrasion to the upper paint layers. Unlike the pink
Fitzwilliam Museum), a work which is thought to have underlayer for Christ’s mantle, this drapery appears to
been painted either immediately before Sebastiano left have been modelled at this underpainting stage, for a
Venice or shortly after his arrival in Rome.58 sample from the shadowed side of the sleeve shows
The soft greenish-lavender colour of the robe a deep mauve containing red earth and ultramarine
of Saint Peter kneeling in the lower left corner is, (plate 11), whereas one from the highlight has a very
however, underpainted with a layer of azurite and lead pale pink (plate 13). Some of the structure and volume
white (plate 12). Scumbled over it, often covering only of this drapery has been lost as a result of blanching of
parts of the underlayer, is a thin layer of lead white with the upper paint layer containing ultramarine and red
red lake, now much faded, and a very little ultramarine. lake; this is especially marked in the area below Christ’s
Essentially this is the same technique as that used in the outstretched hand. Rather surprisingly, given the
lilac draperies with a cool slightly metallic sheen which conservation history of the painting, the ultramarine
are characteristic of Sebastiano’s Venetian predecessors, of Christ’s mantle is relatively little affected with only a
Giovanni Bellini and Cima da Conegliano.59 Even if few small patches of slightly grey pigment.

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 37

Jill Dunkerton and Helen Howard

plate 17 Cross-section of a sample from the foliage of a bush

growing out of the lower part of the rocky outcrop. The uppermost
dark brown paint layers consist of black and red earth pigments. These
layers have been applied over the lighter green employed for the river
bank which incorporates a copper green, lead-tin yellow, yellow earth
and lead white.

The landscape
More ultramarine appears in the sky, where it is used
in the classic Venetian manner over an underpainting of
plate 14 The Raising of Lazarus, detail. azurite and lead white (plates 14 and 15). In both layers
there is a small amount of red lake, resulting in a slight
purple cast, especially in the lighter area above the clouds.
However, the streaks of a dark greenish blue towards
the horizon give the impression of a nocturnal scene60
– at odds with the rest of the landscape which appears
illuminated with shafts of late afternoon sunlight. This
paint (which registers as black in an infrared photograph
taken before the treatment in 1958) consists of coarsely
ground azurite, mixed with only a small amount of lead
plate 15 Cross-section of a sample from the mid-blue sky above the white and rich in binding medium (plate 16); clearly
clouds. Here ultramarine combined with a little lead white and red it was intended to be an intense deep blue – and the
lake has been applied over azurite with lead white, red lake and a few individual pigment particles retain their colour – but
tiny particles of other mineral red pigments.
the paint now appears dark as a result of a reaction
between pigment and medium which has caused the
oil to discolour.
The grassy banks of the river in the background
are painted with muted green mixtures comprising a
copper green, lead-tin yellow, yellow earth and lead
white, while the clumps of grass and small plants in
the foreground are painted with much brighter greens
based on a copper green and lead-tin yellow. It might
be thought that the foliage of the trees and bushes
that grow out the rocks was once green but is now
plate 16 Cross-section of a sample from the dark band of sky below discoloured to brown. However, an autumnal setting
the clouds showing azurite with very little lead white in a matrix of was clearly intended, exactly as in certain paintings by
discoloured medium.
Titian from around this time;61 the leaves seen against
the sky were always a rich red brown and in a sample
from the foliage of the bush growing out of the lower
part of the rocky outcrop, taken at a point where it goes
over the light green nearer bank of the river, there is no

38 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

plate 18 Cross-section of a sample from the brilliant green drapery plate 20 Cross-section of a sample from the highlight of the bright
of the woman holding her cloak to her face, behind and to the right green sleeve of the apostle crouching to the left of Christ. The acid
of Martha. Here an opaque layer of copper green combined with green has been produced by a layer of lead-tin yellow applied wet-in-
lead white and a little black is modelled with a translucent copper wet over copper green combined with lead-tin yellow, lead white and
green glaze. The brown layer (earth pigment combined with carbon a few large dark earth particles. The underlying dark green layer of
black and lead white) beneath the green may be connected with verdigris and earth pigments was clearly dry before the upper layers
pentimenti in this area. were added.

plate 21 Cross-section of a sample from the cloak of the elderly

apostle with hands raised behind Christ’s left shoulder. The moss-
plate 19 Cross-section of a sample from Saint John’s sleeve. The cool green colour of the drapery has been produced by applying a layer of
blue-green colour has been achieved by applying a thin scumble of dull yellow earth combined with copper green over the more intense
ultramarine with a little lead white over several layers of verdigris green underpaint of copper green with lead white.
with lead white in oil. A few black particles (perhaps underdrawing)
are visible on top of the imprimitura.

discoloured copper green, only a mixture of black and in the left background and on the woman who holds
red earth pigments (plate 17). her cloak to her face, behind and just to the right of
Martha (see plate 9). A sample from the latter (plate
Green draperies 18) confirms that this is a true Venetian green, with
The many green draperies distributed across the an opaque underlayer of copper green and lead white
composition generally contain the same pigments as in with a little black, modelled with translucent glazes of
the green foliage but combined and layered in several copper green, applied even over the highlights. Where
different ways with a remarkable variety of effects. thinly applied, the glazes are now somewhat rubbed,
The distribution of these various greens confirms that but they appear to have retained much of their original
Sebastiano had little interest in using colour to make his intensity of colour. The brown paint layer under the
figures recede in space, any more than he had in their green layers in the sample can probably be explained by
logical positioning. A deep saturated colour is as likely the evident pentimenti in this area (see plate 23). The
to appear in the draperies of a background figure as on shadowed part of the robe of the young apostle on the
one in the foreground, creating a tension – or some left, immediately below Saint John, is also richly glazed,
would say imbalance – between the painting surface but where the dramatic lighting picks out his shoulder
and the implied recession of the arc of figures for which and cuff Sebastiano applied bold highlights of lead-tin
neither restorers nor the effects of time can really be yellow, painted wet-in-wet over the still soft underlayer
blamed.62 The richest and deepest green appears on the and modified for the mid-tone with a thin green glaze
cloak of the Pharisee on the right of the group of three (plate 20). The paler rather cold blue green of Saint

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 39

Jill Dunkerton and Helen Howard

John’s robe immediately above might seem to be the

result of loss of glazes in a past cleaning; the bluish
tinge, however, was clearly intended since in the sample
the sequence of green layers was completed with a thin
scumble of ultramarine and lead white (plate 19). Yet
another variation occurs in the softer moss green of
the elderly apostle with his hands raised, to the right of
Christ: here the brilliance of the copper green and lead
white underlayers is suppressed by the addition of a dull
yellow earth to the final layer (plate 21).
An even more muted green features in the jerkin of
the man who bends over Lazarus as he lifts him from
the tomb (plate 22). In the sample illustrated (plate
23) the principal paint layer consists of a pale mauve
colour, containing azurite, lead white and a little red
lake (more red lake is evident in a second sample).
Whether this is part of the build-up of colour or
whether it is related to an alteration is not clear. Even
with the naked eye it can be seen that Sebastiano made
considerable changes in this area, including, it would
appear, to the shoulder of this figure. In adjusting the
plate 22 The Raising of Lazarus, detail. design he may also have revised the colour distribution,
especially since the figure of Martha to the left is also
wearing a lilac-coloured drapery. That he was prepared
to change the colour of a drapery is confirmed by a
sample from the light green sleeve of the bystander
wearing a pinkish-mauve turban; underneath the three
progressively paler layers of verdigris and lead white is a
rich purple, containing azurite, red lake and lead white
(plate 24).
Returning to the figure supporting Lazarus, the
plate 23 Cross-section of a sample from the dull green tunic of the eventual dull bluish-green colour of his jerkin was
man leaning over Lazarus. The layer at the surface includes lead white applied very thinly, and apparently when the underlying
with small particles of ultramarine, translucent yellowish copper-rich
mauve colour had been dry for a considerable time;
inclusions and a few shiny dark particles of galena (lead sulphide).
The lower layers, perhaps associated with pentimenti, consist of
in the sample the crack in the paint affects only the
azurite combined with lead white over a pale mauve of azurite, red upper layer and the boundary between layers is clearly
lake and lead white. defined. The composition of the upper layer is unusual,
containing lead white combined with small particles of
natural ultramarine, translucent yellowish copper-rich
inclusions and a few shiny dark particles of galena (lead
sulphide). Although translucent yellow-brown particles
containing copper have been noted in other paintings,
the identity and nomenclature of the pigment has not
yet been securely established.63 Sparkling black galena
(iron sulphide) has been identified in a number of
other Italian easel paintings from the late fifteenth and
plate 24 Cross-section of a sample from the sleeve of the man with early sixteenth centuries, including on two altarpieces
a turban at top right. The lowest layer contains azurite combined at the National Gallery, both north Italian: The Virgin
with red lake and lead white. Over this are two layers of verdigris and Child Enthroned between a Soldier Saint and Saint
combined with lead white and, for the highlights, a thin scumble
John the Baptist (‘La Pala Strozzi’) (NG 1119) begun
of lead-tin yellow and lead white.
by Gianfrancesco Maineri and completed, probably
in 1499 by Lorenzo Costa, where galena was used for
the soldier’s armour, and The Circumcision (NG 803)

40 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

by Marco Marziale, dated to 1500. Here galena was

used with stibnite, another grey black pigment, mixed
together in the grey decorative border of the red cloak
of the boy kneeling in the foreground of the painting.64
Whether painters recognised the differences between
these various grey and black pigments, and what they
asked for when they bought them from their suppliers
is not yet known.
The final variation on the colour green occurs in plate 25 Cross-section of a sample from a highlight of Mary
the extraordinary acid yellow-green of the kneeling Magdalene’s robe. The acid-yellow highlight contains lead-tin yellow
Magdalen’s dress; this is, in fact, a cangiante fabric, with type I (confirmed by Raman spectroscopy) and lead white. This has
been applied over a pale lime-green underpaint of lead-tin yellow,
green shadows containing copper green, lead-tin
copper green, yellow earth and a little lead white.
yellow, lead white, yellow earth and some large black
particles, which may be responsible for the sour edge
to the green (see plate 3), and highlights of pure lead-
tin yellow over a pale lime green, consisting of lead-
tin yellow with a little copper green and yellow earth
(plate 25). Green and yellow cangiante fabrics appear
on Michelangelo’s Sistine ceiling,65 but with almost
no blending at juxtapositions between highlight and
shadow, which results in clean pure colour. In adapting
this colour combination to the oil medium, with its
potential for more blended transitions, Sebastiano has
produced a lime green colour which appears novel to
panel painting.66 plate 26 Cross-section of a sample from the deep yellow collar of the
apostle kneeling behind Saint Peter. The yellow paint layer consists of
a golden-yellow earth pigment combined with a little lead-tin yellow
Yellow and orange draperies and lead white. This has been applied over a pale mauve containing
Nowhere on the altarpiece did Sebastiano paint a true azurite with red lake, red earth, vermilion and lead white.
yellow drapery. The headdress and cloak of the woman
at the back of the group behind Martha have the same
green cast as the Magdalen’s dress, while other yellow
areas tend towards gold or pink.The small area of yellow
on the apostle kneeling immediately behind Saint Peter
in the lower left corner (see plate 9) contains mainly a
golden yellow earth, with only a little lead-tin yellow
(plate 26). This was painted over a lilac layer, of the
same composition as elsewhere on the painting – the
change is another example of Sebastiano’s concern
to achieve the widest possible distribution of colours plate 27 Cross-section of a sample from the highlight of the yellow
across the design. cloak of the man covering his nose on the far right. A layer of lead-
The brightest highlights of the cloak of the elderly tin yellow lies over an undermodelling of yellow earth with a little
man at the right edge in the middle background, who red earth. Below, is a dark copper-rich layer, and another yellow earth
shields himself from the stench of the dead Lazarus, based paint layer on top of the imprimitura.

are painted with pure lead-tin yellow, but over an

undermodelling of yellow earth, which becomes
pinker towards the shadows, probably because of the
addition of a red earth (plate 27). In the paint sample
the layers of yellow earth are interrupted by a thin dark
layer, either black or possibly a very dark green since
it contains copper. This layer is not easily explained
but the sample point is close to the area affected by
revisions to the design. In both hue and tonal range the
drapery of this bystander is now similar to the cloak

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 41

Jill Dunkerton and Helen Howard

plate 30 Cross-section of a sample from Saint Peter’s deep red-

orange cloak. The uppermost paint layer contains realgar, now
partially transformed to its yellow polymorph, pararealgar. This was
applied over a red-orange underpaint of red earth combined with
lead-tin yellow and lead white.

plate 28 The Raising of Lazarus, detail showing Saint John the

Evangelist.

plate 31 Cross-section of a sample from the decayed orange-brown

cloak of the apostle leaning over Christ’s shoulder showing arsenic
plate 29 Cross-section of a sample from Saint John’s pale orange
sulphide pigments (realgar and orpiment or pararealgar) combined
robe showing two layers of red lead, with some red lake in the lower
with red and yellow earth and glazed with red lake.
layer. Semi-translucent white lead soaps are present throughout the
paint layers but at the surface the alteration is so complete that no red
lead particles remain.

of Saint John the Evangelist on the opposite side of Saint Peter (plate 30); this was laid in with red earth
the composition (plate 28). This was not always so, combined with lead-tin yellow and lead white and
however, for the colour of the latter is considerably then finished in the mid-tones with the orange arsenic
altered. Originally it must have been a bright reddish- trisulphide mineral, realgar, widely used in Venice in
orange colour, containing red lead with some red lake, the early years of the sixteenth century, but probably
especially in the shadows, but, as the cross-section something of a novelty in Rome. In the sample both
(plate 29) shows, the red lead has reacted with the realgar and its yellow polymorph pararealgar are
linseed oil medium forming translucent lead soap present. Pararealgar is a naturally occurring mineral, but
inclusions which are white.67 These inclusions are is also produced by a light-induced transformation of
present throughout the layer structure, but at the top realgar.69 The breakdown of parts of the paint film on
surface the deterioration of the red lead is so complete this drapery suggests that the pararealgar found here
that it now appears as a very pale pinkish-orange is related to the deterioration of realgar rather than
highlight.68 Although Sebastiano clearly wished to being a deliberate addition of a golden yellow pigment.
draw attention to this figure by clothing him in bright Moreover, in another sample from a lighter area the
colours, these unintended highlights now compete yellow arsenic trisulphide mineral, orpiment, is present.
with the strong side-lighting of Christ and also of the The darker orange-brown drapery of the apostle
apostle below Saint John. peering over Christ’s right shoulder also contains an
Some alteration has inevitably occurred to the arsenic sulphide pigment (plate 31), but in the sample
orange cloak typically assigned by Venetian painters to it occurs mixed with red earth rather than over it. The

42 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

damaged condition of parts of this drapery indicates

that realgar is probably also present. Where there is a
final glaze of red lake, which gives this drapery a richer
deeper colour than the orange of Saint Peter’s, the paint
is generally better preserved.

Red and pink draperies

With one exception, all the areas of red in the altarpiece
incline towards orange; the samples show that Sebastiano
usually added an orange-red earth to his vermilion plate 32 Cross-section of a sample from Martha’s orange-red cloak.
for the lighter areas of red, for example the cloaks of The orange-red paint layer consists of vermilion and red earth
Martha (plate 32) and the apostle at the left below combined with lead white. Immediately below this, at the right side
of the paint sample, is what appears to be a trace of underdrawing in
Saint John (plate 33), as well as the tunic of the young carbon black. This fine black layer has been applied over lead white.
man who supports Lazarus on the right.Vermilion and A trace of the grey-brown imprimitura is present at the base of the
white alone appear only in the Magdalen’s cloak in sample. It is possible that a revision was made to the underdrawing in
the foreground. In the sample from Martha’s drapery this area over a fresh layer of lead white.
(plate 32) there is a layer of lead white immediately
above the brownish-grey imprimitura (present only as
a trace, perhaps as a result of the transfer process). A
few black particles of possible underdrawing appear
over the lead white. There is no obvious reason why
this figure should have been assigned a lighter, more
reflective preparation, and so it may be that the sample
point coincides with an area where the underdrawing
was revised and the lead white was used to cancel the
previous design. The salmon-pink highlight of the
plate 33 Cross-section of a sample from the pale terracotta-red of
apostle’s cloak (plate 33) also includes yellow earth but the cloak of the apostle below Saint John. The bright orange-red
the composition of a sample from the depths of a fold underpaint contains vermilion combined with lead white. Over this,
(plate 34) is rather unexpected, in that it consists of a the terracotta-pink highlight is composed of vermilion, lead white,
single layer of purple made from red lake combined earth pigments and a little lead-tin yellow.
with vermilion, lead white and a mineral blue pigment.
In contrast to this direct technique of modelling,
Martha’s drapery was completed in the shadows with
glazes of red lake over the underlayer of vermilion
and red lake (plate 35). Sebastiano reserved the cool
pink of pure red lake for the robe of Christ, where it
balances the purple blue of his mantle, the only area
of the painting where the natural ultramarine pigment
is used in its fully saturated form. Analysis of the red plate 34 Cross-section of a sample from the purple shading of the
lake dyestuff has identified it as kermes with a little cloak shown in plate 33. The uppermost layer of the sample consists
madder. The madder is only in the lower layer; it may of red lake combined with vermilion and a blue mineral pigment and
have been a deliberate choice by the painter, but it is a few large dark inclusions applied directly over the imprimitura.
also possible that it is present as a result of the inclusion
of shearings of cloth dyed with madder together with
the kermes-dyed textiles from which the greater part
of the pigment was extracted.70 In the final layers of
the pink robe (plates 36 and 37) a little ultramarine
was added to the red lake to enhance its purple tinge
and to link its colour to that of the ultramarine mantle
(itself underpainted with red lake). In addition, particles
of powdered glass (see Appendix, pp. 48–51) have been
identified in the red lake paints, and also in a few other
samples, for example that from Lazarus’s leg (plate 8); it is

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 43

Jill Dunkerton and Helen Howard

now evident that the addition of powdered manganese-

containing glass to increase the drying rate of slow-
drying pigments, such as lakes when used in an oil
medium, was a common practice throughout Europe
in this period, and was certainly known to Raphael.71
On The Raising of Lazarus where there is a substantial
amount of red lake in the paint layer, for example the
glazed shadows, the colour is still rich and intense. In
the lighter areas, on the other hand, where the lake is
mixed with large amounts of lead white, it can be seen
in the cross-section (plate 37) that there is some loss
plate 35 Cross-section of a sample from Martha’s cloak, showing of colour at the upper surface. This fading exaggerates
a layer of vermilion combined with red lake over the imprimitura. the contrast between highlight and shadow but does
This has been glazed with red lake to produce the deep shadow of not greatly diminish the splendour of Christ’s garments.
the drapery.
Sebastiano was still following the tradition of using the
two most costly pigments, ultramarine and kermes,
to distinguish the most important figure; moreover,
he set apart Christ and his apostles (and therefore a
large part of the picture surface) by not giving them
any white draperies, whereas creamy-white fabrics,
painted with the relish that one would expect from a
Venetian, are widely distributed across Lazarus’s side of
the altarpiece.
When Raphael eventually made progress on The
Transfiguration (plate 38) he chose to echo Sebastiano’s
Christ by using the same colours for the draperies of
plate 36 Cross-section of a sample from the shadow of Christ’s robe, the woman who kneels to the right of centre in the
showing kermes lake combined with a little madder and a little lead
foreground (an important figure in the design, if not
white and then glazed with the same combination of lake pigments,
with the addition of few particles of ultramarine and ground glass, the narrative). Although Raphael never allowed colour
probably added as a siccative (see Appendix, plate 39). to disrupt or compete with the design, it is clear from
the extended palette used for his altarpiece, including
acid greens and many shades of red, orange and brown,
that he was responding to the challenge set by the
Venetian. It is not surprising that he was impressed;
what is perhaps more remarkable is how, in spite of
its conservation history and the many changes to its
appearance – some not apparent until this study and
some perhaps less important than previously thought
– The Raising of Lazarus can still inspire admiration
for its magnificence, just as it did in early sixteenth-
plate 37 Cross-section of a sample from a highlight of Christ’s century Rome.
robe, showing red lake was combined with lead white and a little
ultramarine. There is significant fading of the red lake at the upper Acknowledgements
surface of the sample.
We are very grateful to Ann Massing for discussion of
French transfer techniques in the eighteenth century
and for allowing us to consult the manuscript of her
forthcoming book on the history of restoration in
France, and to Renate Woudhuysen-Keller for sharing
information about Sebastiano’s Adoration of the Shepherds
(Cambridge, Fitzwilliam Museum), which she is
currently restoring at the Hamilton Kerr Institute. We
would also like to thank Janet Ambers, Science Group,
Department of Conservation and Scientific Research,

44 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

Notes
1 For the commission see M. Hirst, Sebastiano del Piombo, Oxford 1981, p. 66.
2 Letter dated 19 January 1517 from Leonardo Sellaio in Rome to
Michelangelo in Carrara: à avuti danari per fare e’ legname’. P. Barocchi and
R. Ristori eds., Il Carteggio di Michelangelo, Florence 1965,Vol. I, p. 243, cited
in Hirst 1981, op.cit., p. 66, note 2.
3 F. Mancinelli, ‘La “Trasfigurazione” e la “Pala di Monteluce”: Considerazioni
sulla loro tecnica esecutiva alla luce dei recenti restauri’ in J. Shearman and
M. B. Hall eds., The Princeton Raphael Symposium. Science in the Service of Art
History, Princeton University Press 1990, pp. 149–60, esp. pp. 149–50.
4 For Michelangelo’s part in the design and his drawings see Hirst 1981 (cited
in note 1), pp. 69–71.
5 These are cited with references in the account of the altarpiece given in
Hirst 1981 (cited in note 1), pp. 66–75.
6 Letter dated 10 December 1519 from Leonardo Sellaio in Rome to
Michelangelo in Florence:‘Bastiano…à vernichato la [ta]vola, che è mirabile’.
P. Barocchi and R. Ristori eds, Il Carteggio di Michelangelo, Florence 1967,Vol.
II, p. 205, cited in Hirst 1981 (cited in note 1), p. 69, note 19.
7 See Hirst 1981 (cited in note 1), p. 69.
8 For Raphael’s response to Sebastiano in the colouring of the lower half of
The Transfiguration see M.B. Hall, Color and Meaning. Practice and Theory in
Renaissance Painting, Cambridge University Press 1992, pp. 134–5.
9 Letter dated 12 April 1520 from Sebastiano del Piombo in Rome to
Michelangelo in Florence: ‘avisovi come hozi io ho portato la mia tavola
un’altra volta a Palazo, con quella che ha facto Rafaello, et non ho havuto
vergogna.’ Barocchi and Ristori 1967 (cited in note 2), p. 227.
10 The passage from Vasari is quoted in full in Hirst 1981 (cited in note 1), p.
67, note 7.
11 Having been brought to Paris in 1797 as the most prized of all the
Napoleonic spoils, The Transfiguration was cleaned and restored in 1802.
Although thinning of the panel and possible transfer by François-Toussaint
Hacquin (son of Jean-Louis) had been recommended by the Conseil
d’Administration, caution prevailed and instead the panel was braced by iron
cross bars. See A. Massing, Towards a History of Painting Restoration: the origins
plate 38 Raphael, The Transfiguration, 1516–20,. Oil on panel, 410 × of the profession in France, forthcoming 2009, Chapter 8, ‘François-Toussaint
279 cm. Vatican Museums,Vatican City inv. 40333. Hacquin, painting restorer for the New Régime’. The three horizontal
battens of chestnut inserted in dovetail grooves described by Mancinelli in
Shearman and Hall 1990 (cited in note 3), p. 150, may well be original to
the panel. The two narrower battens visible in the X-radiograph (plate 332),
British Museum, for Raman microspectroscopy of inserted between these are perhaps F-T. Hacquin’s additions. The panel has
since undergone extensive structural treatment.
samples; and at the National Gallery, Marika Spring 12 C. Gardner von Teuffel, ‘Sebastiano del Piombo, Raphael and Narbonne:
and Rachel Morrison for work on the samples, new evidence’, Burlington Magazine, 126, No. 981, 1984, pp. 265–6.
13 C. Gardner von Teuffel, ‘An early description of Sebastiano’s Raising of
Rachel Billinge for infrared reflectography, and Lazarus at Narbonne’, Burlington Magazine, 129, No. 1008, 1987, pp. 185–6.
Martin Wyld, Carol Plazzotta and Nicholas Penny for 14 The Regent had been negotiating for the acquisition since 1706. See F.
Mardrus, ‘Le Régent, mécène et collectionneur’ in Le Palais Royal, exh.cat.,
helpful comments and references. In this issue of the Musée Carnavalet, Paris 1988, pp. 79–115, esp. p. 98.
Technical Bulletin which celebrates the thirtieth year of 15 For an account of the Orléans Collection, see N. Penny, National Gallery
its publication we especially want to acknowledge the Catalogues: The Sixteenth Century Italian Paintings, Vol. II, London 2008,
Appendix of Collectors’ Biographies pp. 461–70. See also Mardrus 1988
contribution of Joyce Plesters. Even though she was (cited in note 14).
not able to develop this material for publication, she 16 [L-F. Du Bois de Saint-Gelais], Description des tableaux du Palais-Royal, Paris
1727, pp. 448 ff. He also notes:‘J. C. habillé de rouge avec une draperie bleuë’
recognised the value of taking a wide range of samples and Saint Peter with ‘une robe d’un bleu clair avec une draperie jaune par
while the painting was undergoing restoration and dessus’.
17 National Gallery Manuscript Catalogue, transcribed into the Conservation
the sample points were available – ‘samples for a rainy Record for NG 1. The source for this date is not clear.
day’ as she used to call them. Forty years later the day 18 For biographical details of Hacquin father and son, see Massing, forthcoming
(cited in note 11),and also F. Barrès, ‘Les peintures transposes du Musée du
finally came; we hope that we have done justice to Louvre, étude des techniques de transpositions en France, de 1750 jusqu’à
the material. la fin du XIXème siècle’, ICOM Committee for Conservation 14th Triennial
Meeting The Hague, 12–16 September 2005, Preprints Vol. II, pp. 1001–8.
19 See Massing, forthcoming (cited in note 11), chapter 4, ‘Jean-Louis Hacquin
and François-Toussaint Hacquin. Painting restoration under the Ancien
Régime’.
20 M. Wyld and J. Dunkerton, ‘The Transfer of Cima’s “The Incredulity of S.
Thomas”’, National Gallery Technical Bulletin, 9, 1985, pp. 38-59.
21 Massing, forthcoming (cited in note 11), chapter 2, ‘Robert Picault and
the transfer procedure in France’, especially the section on Picault’s ‘secret’
. technique.
22 R.P. Knight (anon. at time of publication), ‘Northcote’s “Life of Reynolds”’,
Edinburgh Review, XLVI, September 1814, pp. 263–92, esp. pp. 283–4. In a
rather confused (and all too familiar) diatribe Knight claimed that glue from
the transfer had penetrated the surface and ‘diffused over it that general hue
of dim and opaque brown [more probably a discoloured old varnish]…

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 45

Jill Dunkerton and Helen Howard

which the operator then endeavoured to remove by the usual process of Musée du Louvre) had to be re-transferred by Hacquin père in 1776 and
scouring away all the delicate and transparent glazing, by which the various 1780 respectively. See Barrès 2005 (cited in note 18), p. 1002.
splendour of the various parts had been tuned and harmonized together 37 Treatment Report by A.W. Lucas, October 1967 (National Gallery
into one brilliant mass’, now reduced to ‘monotony and crudeness’. We are Conservation Record for NG 1).
grateful to Nicholas Penny for drawing our attention to this account 38 See Wyld and Dunkerton 1985 (cited in note 20), pp. 54–8.
23 A. Conti, Storia del Restauro e della Conservazione delle Opere d’Arte, Electa, 39 A.W. Lucas, ‘The Transfer of Easel Paintings’ in G. Thomson ed., Recent
Florence 1988, p. 126; translated by H. Glanville as A History of the Restoration Advances in Conservation: Contributions to the IIC Rome Conference, London
and Conservation of Works of Art, Oxford 2007, p. 147. 1963, pp. 165–8.
24 Initially criticism of Picault – and his eventual fall from favour - was as 40 This information was supplied by Joyce Plesters to Michael Hirst. See Hirst
much about the outrageous sums that he charged for his transfers as for 1981 (cited in note 1), p.73.
the consequences (see Massing, forthcoming, cited in note 11, chapter 2), 41 Only three new samples were taken in order to resolve specific questions
but by the end of the century a report on the conservation of Raphael’s that arose as result of the examination of the archived samples.
Saint Michael and Saint Margaret (both in the Louvre) noted that their 42 See J.S. Mills and R. White, ‘The Gas Chromatographic Examination of
deteriorating condition was caused by the reactive nature of the layers of Paint Media. Some Examples of Medium Identification in Paintings by
glue applied in the transfer, and regretted the loss of the smooth surface Fatty Acid Analysis’, Conservation and Restoration of Pictorial Art, London
characteristic of panel paintings, since they were now imprinted with the 1976, eds N. Brommelle and P. Smith, pp. 72–7, esp. p. 75.
texture of the transfer canvases; Conti 1988, p. 140, and 2007, p. 167 (both 43 This preliminary examination by infrared was carried out by Rachel
cited in note 23). Billinge with the painting hanging in the Gallery. The SIRIS prototype
25 Penny 2008 (cited in note 15), p. 467. digital infrared camera was used.
26 J. Pomeroy, ‘“To comprehend the excellencies of that performance”’ 44 See C. Barbieri ed., Notturno Sublime. Sebastiano e Michelangelo nella Pietà
Sebastiano del Piombo’s Raising of Lazarus comes to England’, The British di Viterbo, exh. cat. Museo Civico, Viterbo 2004, esp. pp. 32–7, 89–105, and
Art Journal,Vol. II, no. 3, Spring/Summer 2001, pp.63–8. R. Bellucci and C. Frosinini, ‘Il processo di elaborazione dell’immagine in
27 The restorer John Seguier, in his evidence to the House of Commons Sebastiano del Piombo: “La Pietà” e “La Flagellazione di Viterbo”’ in C.
Select Committee on the National Gallery, 1853, p. 43, reported: ‘The Barbieri, E. Parlato and S. Rinaldi eds, ‘La Pietà’ di Sebastiano a Viterbo: storia
Sebastiano del Piombo was so much damaged when Mr. Angerstein bought e tecniche a confronto, postprints of study day at Università della Tuscia,Viterbo,
it, that Mr. West repaired it considerably’. Landseer, in his 1834 catalogue 10 June 2005, Rome 2009, pp. 148–69, esp. pp. 156–64.
of the collection, described West’s intervention more extravagantly: ‘the 45 See B. Marocchini, ‘La Pietà di Viterbo: la tecnica di esecuzione’ in Barbieri
Lazarus of Michael Angelo was not only dead, but was fast sinking into ed. 2004 (cited in note 44), pp. 89–93, esp. pp. 90–1 and p. 95 (for the
the ground, when Mr. West was empowered to stand before it, and again cross-sections, although the location of the sample points is not clearly
command Lazarus to “come forth”’. He commended West for his ‘dexterous described).
and successful restorations’ (J. Landseer, Descriptive, Explanatory, and Critical 46 See J. Dunkerton and M. Spring,‘The Development of Painting on Coloured
Catalogue of Fifty of the Earliest Pictures Contained in the National Gallery of Surfaces in Sixteenth-century Italy’, Painting Techniques: History, Materials and
Great Britain, London 1834, p. 111). Studio Practice, Contributions to the Dublin Congress of the International
28 Penny 2008 (cited in note 15), p. 467. Institute for Conservation, 7-11 September 1998, A. Roy and P. Smith eds,
29 Letter from Sir Charles Eastlake in Milan, 20 October 1865, to Ralph pp. 120–30, esp. p. 128. See also J. Dunkerton,‘Tra Venezia e Roma: il disegno
Wornum, National Gallery Archive. Pinti was paid 19 guineas (£19 19s.) for preparatorio nei dipinti londinesi di Sebastiano’ in Barbieri, Parlato and
his work on the Sebastiano. Rinaldi eds 2009 (cited in note 44), pp. 170–85, esp. pp. 172–3.
30 Minutes of the Trustees of the National Gallery, 8 May 1881. National 47 J. Dunkerton, S. Foister and N. Penny, Dürer to Veronese: Sixteenth-Century
Gallery Archive. Painting in the National Gallery, New Haven and London 1999, pp. 274–
31 ‘During these operations it became clear that the statement that the picture 5, and A. Cerasuolo, ‘I dipinti di Sebastiano del Piombo del Museo di
had been transferred from wood was inaccurate. 1) The apparent cracks are Capodimonte. Note sulla tecnica’, in Barbieri, Parlato and Rinaldi eds, 2009
the joins (longitudinal) of four strips of canvas. 2) The grain of the original (cited in note 44), pp. 128–47.
canvas is visible in the paint and there is no trace of wood grain; the latter 48 Hirst 1981 (cited in note 1), p. 72, note 38, suggests that the addition of a
could not have been completely removed by transfer. 3) The canvas visible white ground has resulted in an imbalance in the tonal transitions. Other
under the relining canvas is of the 16th cent. coarse Italian type. Probably explanations for this are presented in this article. Moreover, back in 1834,
for “transferred” should be read “relined”.’ National Gallery Manuscript Landseer commented on Sebastiano’s colour distribution, observing that
Catalogue transcribed into Conservation Record for NG 1. ‘the women who are muffling their olfactory nerves from the diminution
32 See Barrès 2005 (cited in note 18), pp. 1004–5, and also E. Martin, J. Bret of their dimensions…we are taught to think of are at some distance; yet the
and C. Naffah, ‘Le décor du petit cloître de la Chartreuse de Paris peint par lights on their head gear are as bright as the white drapery around the head
Le Sueur. Etude technique et historique des restaurations’, Techne, no. 1, of Lazarus’. Landseer 1834 (cited in note 27), p. 116.
1994, pp. 85–102, esp. pp. 89–91. 49 See Dunkerton in Barbieri, Parlato and Rinaldi eds 2009 (cited in note 44),
33 The composition of this ‘enduit de transposition’ is not unlike that of oil pp. 181–3.
grounds that were still being applied to canvases for painting. For example, 50 The similar sharp edge to modelling of his right shin, however, is the result
Le Comte de Vaudreuil (NG 4253), painted by François-Hubert Drouais in of repainting in the course of past restorations.
1758, has a double ground of the type often seen on seventeenth-century 51 For Sebastiano’s use of chiaroscuro on panels and walls see Hall 1992 (cited
French canvases, with a first layer of orange red, ‘composed of a silicaceous in note 8), pp. 134–5 and 138–41.
natural red earth’ and an upper layer of light to mid-brownish grey, ‘made up 52 Hirst 1981 (cited in note 1), p. 72, draws attention to the contrasting flesh
of lead white, some calcium carbonate, wood charcoal (splintery particles) tones.
and a small amount of yellow earth’ (report by Ashok Roy, 23 January 2003, 53 On 22 September 1518, Beltrame Costabili in Rome wrote a letter of
in Scientific Department file). introduction to Alfonso d’Este in Ferrara on behalf of one of Raphael’s
34 Picault’s transfer in 1751 of a Raphael Madonna for the Orléans Collection garzoni, who was on his way to Venice, ‘mandato da epso Raphael, credo
was highly praised; of the three Raphael paintings from the collection that per comperere colori’ [sent by the said Raphael, to buy colours, I believe];
retain their attribution to Raphael, The Bridgewater Madonna, The Holy Family J. Shearman, Raphael in Early Modern Sources (1483–1602), New Haven and
with a Palm (both now Edinburgh, National Galleries of Scotland), and The London 2003,Vol. I, p. 373. Given the date, it is possible that these pigments
Macintosh Madonna (NG 2069), the transfer of the first was attributed by were purchased for The Transfiguration.
Waagen to Hacquin. The painting transferred by Picault would therefore 54 See Mills and White 1976 (cited in note 42), p. 75. GC–MS analysis by
have to be either The Holy Family or The Macintosh Madonna (although Rachel Morrison of two further samples, from Saint John’s green sleeve and
the transfer of the latter, now a ruin, could hardly be considered worthy of from the dull green tunic of the figure leaning over Lazarus, found heat-
admiration). See Conti 1988, p. 128 and p. 231, and 2007, p. 150 and p. 177, bodied linseed oil in both.
note 12 (both cited in note 23, although here the translation confuses the 55 See Dunkerton in Barbieri, Parlato and Rinaldi eds 2009 (cited in note 44),
issue by conflating the two Bridgewater pictures). pp. 183–4.
35 For example he replaced the gesso ground of Andrea del Sarto’s Charity 56 For example, The Consecration of Saint Nicholas (NG 26); see N. Penny and
(Paris, Louvre) with an inappropriate red-brown ground; see G. Emile-Mâle, M. Spring, ‘Veronese’s Paintings in the National Gallery. Technique and
‘Histoire rapide de la restauration des peintures du Louvre. 1ère partie: Des Materials: Part I’, National Gallery Technical Bulletin, 16 , 1995, pp. 4–29, esp. p.
origines jusqu’à 1848’, Coré, no. 3, October 1997, pp. 52–7, esp. p. 55. 13.
36 For example Raphael’s Saint Michael and del Sarto’s Charity (both Paris, 57 See S. Béguin, Nouvelles analyses résultantes de l’étude et de la restauration des

46 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

Raphaël du Louvre, in Shearman and Hall eds 1990 (cited in note 3), pp. Conservation Institute Newsletter 22, 1988). The P phase was not detected in
39–55, and esp. the appendix, pp. 54–5. samples from the Raising of Lazarus, but the very deteriorated nature of
58 This very badly damaged painting – also a victim of a French transfer while the paint layer and the presence of pararealgar and realgar in the samples
in the Orléans collection – is at present undergoing treatment by Renate suggest light-induced alteration rather than the use of mineral pararealgar
Woudhuysen-Keller at the Hamilton Kerr Institute. by the artist. Pararealgar has also been identified in the Holy Family and
59 For example the young apostle to the left of Christ in Cima’s The Incredulity Saints, tentatively ascribed to the school of Titian and now in Winnipeg
of Saint Thomas (NG 816); see J. Dunkerton and A. Roy, ‘The Technique and Art Gallery (M.C. Corbeil and K. Helwig, ‘An occurrence of pararealgar as
Restoration of Cima’s “The Incredulity of S. Thomas”’, National Gallery an original or altered artists’ pigment’, Studies in Conservation, 40, 1995, pp.
Technical Bulletin, 10, 1986, pp. 4–27, esp. p. 14 and plate 2d. 133–8).
60 So misleading is this darkening of the azurite that Hall 1992 (cited in note 70 The HPLC analysis of the red lake was carried out by Jo Kirby. The madder
8), p. 134, actually describes The Raising of Lazarus as ‘a night scene’. was present in two of the three samples, but it was not possible to determine
61 See J. Dunkerton and M. Spring, ‘The Technique and Materials of Titian’s for certain that there was a deliberate layering of lakes as in Lorenzo Lotto’s
Early Paintings in the National Gallery, London’, in Titian, Jacopo Pesaro being two versions of A Virgin and Child with Saints Jerome and Nicholas of Tolentino
presented by Pope Alexander VI to Saint Peter,Vol. 3, no.1 of Restoration, journal (see J. Dunkerton, N. Penny and A. Roy, ‘Two Paintings by Lorenzo Lotto in
of the Koninklijk Museum voor Schone Kunsten, Antwerp 2003, pp. 9–21, the National Gallery’, National Gallery Technical Bulletin, 19, 1998, pp. 52–63,
esp. p. 17. esp. pp. 55–6) or whether the lake was made from a mixture of dyestuffs (see
62 Conti writes of ‘a serious imbalance due to poor cleaning’ (see Conti 2007 J. Kirby,‘The Identification of Red Lake Pigment Dyestuffs and a Discussion
[cited in note 23], p. 166 – this passage is not in the first edition of 1988). It of their Use’, National Gallery Technical Bulletin, 17, 1996, pp. 56–80, esp. pp.
is questionable whether the painting was ever harmonious in its colouring, 67–8).
or indeed was intended to be so. Moreover, its conservation history and the 71 Spring has identified soda-lime glass in Italian paintings and high-lime or
many alterations to the painting materials – some not previously identified mixed alkali glass in paintings from Germany and the Netherlands. See M.
– make it unlikely that a so-called balanced cleaning could have been Spring, ‘Raphael’s Materials: some new discoveries and their context within
achieved. early sixteenth-century painting’, (cited in note 64).
63 A yellow-brown pigment containing copper has been noted on a painting by
Quinten Massys at the National Gallery; see J. Dunkerton, ‘The Technique
and Restoration of “The Virgin and Child with Four Angels” by Quinten
Massys’, National Gallery Technical Bulletin, 29, 2008, pp. 60–75, esp. p. 70 and
p. 75, notes 25 and 26 (supplied by Marika Spring).
64 See M. Spring, R. Grout and R. White, ‘“Black earths”: A Study of Unusual
Black and Dark Grey Pigments used by Artists in the Sixteenth Century’,
National Gallery Technical Bulletin, 24, 2003, pp. 96–113, esp. pp. 107–10; and
M. Spring, ‘Raphael’s Materials: some new discoveries and their context
within early sixteenth-century painting’, in Raphael’s painting technique:
working practices before Rome. Proceedings of the Eu-ARTEX+CH workshop,
National Gallery, London, 11 November 2004, eds A. Roy and M. Spring,
Nardini Editore, 2007, pp. 77–86.
65 For example, on the prophet Daniel and on the Delphic Sibyl.
66 The novelty of this colour is commented upon by M. Hall,‘La “Resurrezione
di Lazzaro” di Sebastiano e la sfida al colore di Raffaello’ in Barbieri, Parlato
and Rinaldi eds, 2009 (cited in note 44), pp. 26–41, esp. p. 38. In this paper
Hall also argues that the extraordinary palette of The Raising of Lazarus is
to some extent a response to paintings by Raphael such as the ‘Spasimo di
Sicilia’ (Madrid, Museo del Prado).
67 For a discussion of the reactions between red lead (and lead-tin yellow)
and the fatty acids present in oil binders, see: C. Higgitt, M. Spring and D.
Saunders, ‘Pigment–medium interactions in oil paint films containing red
lead or lead-tin yellow’, National Gallery Technical Bulletin, 24, 2003, pp. 75–
95; and M. Spring and C. Higgitt, ‘Analyses reconsidered: the importance
of the pigment content of paint in the interpretation of the results of the
examination of binding media’, in Medieval Painting in Northern Europe:
Techniques, Analysis, Art History; Studies in Commemoration of the 70th Birthday
of Unn Plahter, ed. J. Nadolny, with K. Kollandsrud, M.L. Sauerberg and
T. Trøysaker, Archetype Publications Ltd, London 2006, pp. 223–9, esp. pp.
223–5.
68 The possibility that this alteration had already occurred by 1727 is suggested
by Du Bois de Saint-Gelais’s description of ‘S. Jean qui a une robe verte &
par dessus une draperie jaune’ (Du Bois de Saint-Gelais 1727, cited in note
16). Although he also described Saint Peter’s yellow-orange robe as ‘jaune’,
unaltered red lead would be more likely to be described as orange or red
rather than yellow.
69 The presence of realgar and pararealgar were confirmed by Raman
spectroscopy at the British Museum. We are grateful to Janet Ambers for
undertaking this analysis. Realgar may alter to pararealgar by a light-induced
transformation. Pararealgar is less dense and of greater volume than realgar
and initially forms on the surface of mineral samples as a thin layer or
nodules and then, on reaching a critical thickness, cracks and spalls forming
a powdery orange-yellow material. Recent studies have also shown that the
alteration process involves the formation of an intermediate phase, phase P,
which appears to be a precursor to pararealgar (D.L. Douglass and C. Shing
‘The light-induced alteration of realgar to pararealgar’, American Mineralogist,
77, 1992, pp. 1266–74, and A.C. Roberts, H.G. Ansell and M. Bonardi,
‘Pararealgar, a new polymorph of AsS, from British Columbia’, Canadian
Mineralogist 18, 1980, pp. 525–7). The P phase has been identified by XRD
on a polychromed sculpture dated to c. 700, confirming that, in this instance,
the artist originally employed orange-red realgar and that partial alteration
to yellow pararealgar has occurred (M.C. Corbeil, ‘The P file’, Canadian

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 47

Jill Dunkerton and Helen Howard

Appendix with a little zinc. Elemental analysis of the yellowish-green translucent material
For the deep crimson shadow of Christ’s robe (see plate 36), an opaque layer of itself confirmed the presence of copper, lead and sulphur.The overall moss-green
red lake combined with lead white was glazed with a further layer of translucent appearance of this paint is likely to be due to a combination of the small number
red lake pigment. Examination of the paint cross-section in ultraviolet light of tiny blue ultramarine particles with the yellowish translucent material and a
(plate 39) revealed the presence of translucent inclusions with glassy fracture few grey particles of galena (lead sulphide) combined with lead white.
in both paint layers, though the particles are particularly clearly visible in the The backscatter scanning electron image (fig. 8) of the sample taken from
pale pink underpaint. Several of these particles were separated from the paint the area of dark green-blue sky (see plate 16) shows angular particles of the
layer by micromanipulation, mounted as dispersions in Meltmount (which has mineral blue pigment azurite, with a very few inclusions of lead white (which
a refractive index of 1.662) and observed in plane polarised transmitted light. appear bright white due to their high atomic number) embedded in the oil
These particles display the distinctive conchoidal fracture and stress lines which binding medium. The ratio of pigment to binder is low and it is the darkening
are characteristic of broken glass (plate 40). Soda-lime glass has been identified of the medium which produces the dark greenish-blue appearance of the paint
in a number of works by Raphael (see M. Spring, ‘Raphael’s Materials: some layer.
new discoveries and their context within early sixteenth-century painting’, in
Raphael’s painting technique: working practices before Rome. Proceedings of the Eu-
ARTEX+CH workshop, National Gallery, London, 11 November 2004, eds A.
Roy and M. Spring, Nardini Editore, 2007, pp. 77–86) and in paintings from all
over Italy and throughout Europe at this period. It seems likely that the powdered
glass was added as a drier to the oil paint with the manganese acting as a siccative.
A particularly large elongated glass particle (20μm in length) is visible in the
deep rich glaze of Martha’s drapery (see plate 35) when the sample is viewed in
ultraviolet light (plate 41).
In the backscatter scanning electron image (fig. 6) of the sample taken from
the moss-green jerkin of the figure leaning over Lazarus (see plate 23), one of the
copper-containing semi-translucent yellowish-green areas is visible in the centre
of the image as a large mid-grey area. The lead white matrix appears white (due
to its high atomic number). Of particular interest here are the elongated particles
found within the translucent copper-rich yellowish-green areas which were
also found to contain lead and sulphur, sometimes with a little zinc, which are
clearly visible in the backscatter electron image below (fig. 7). Spot analysis of the
highlighted area shown in the insert confirmed the presence of lead and sulphur

FIG. 6 Backscatter scanning electron image of the sample from the

dull green tunic of the man leaning over Lazarus (plate 23).

plate 39 Cross-section of a sample from the shadow of Christ’s robe

(plate 36) photographed in ultraviolet light.

FIG. 7 Backscatter scanning electron image of another area of the

sample from the dull green tunic of the man leaning over Lazarus
(plate 23), with highlighted insert.

plate 40 Particle of glass from glaze layer in plate 39, photographed

in transmitted semi-polarised light.

plate 41 Cross-section of a sample from Martha’s cloak (plate 35) FIG. 8 Backscatter scanning electron image of the sample from the
photographed in ultraviolet light. dark band of sky below the clouds (plate 16).

48 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

Table of pigments and paint stratigraphy

(priming and transfer layers not included)

Blue and Purple Red Orange Yellow and Brown Green

Lightest blue of Christ’s Deep terracotta pink of cloak Pale orange-pink of cloak of Highlight of yellow robe of Green shadowed area of
cloak. of the apostle below Saint the apostle in profile at the Mary Magdalene’s robe. yellow Mary Magdalene’s
plate 4 John (area of highlight). left edge of the painting. plate 25 robe.
• ultramarine combined plate 33 plate 29 • lead-tin yellow type I plate 3
with lead white • lead white combined • red lead. This layer combined with lead • green paint layer
• pale pink layer of lead with vermilion, yellow exhibits significant white incorporating copper
white combined with and red earth pigments lightening, particular at • lead-tin yellow green, lead-tin yellow,
red lake and a little lead-tin the upper surface, due combined with a lead white, yellow earth
yellow to the formation of lead copper green pigment and large black particles
• vermilion combined soaps and a few yellow earth
with lead white • at the right side of the particles
sample a thin layer of
red lake divides the two
layers of red lead
• red lead combined with
a few red lake particles.
This layer exhibits
significant lightening
due to the formation of
lead soaps
Deepest blue of Christ’s Dark orange-red of Martha’s Deep reddish-orange of Saint Deep yellow of collar of Bright green of cloak of
cloak. cloak. Peter’s cloak. apostle kneeling behind woman holding her cloak to
plate 10 plate 35 plate 30 Saint Peter, left-hand edge her face, behind and just to
• pure ultramarine Appendix plate 41 • realgar and pararealgar of picture. the right of Martha.
• pale pink layer of lead • red lake glaze (kermes (confirmed by Raman plate 26 plate 18
white combined with with a little madder) spectroscopy) • brilliant yellow layer of • copper green glaze
red lake with particles of • red earth combined lead white combined • copper green combined
powdered glass with lead-tin yellow with yellow earth and with lead white and
• vermilion combined and lead white lead-tin yellow a few carbon black
with red lake • pale mauve layer of particles
azurite combined with • brown underpaint
lead white, red and containing earth
yellow earth, vermilion pigments and some
and a particle of red large carbon black
lake pigment particles with lead
white
Patchy pale lilac-blue of Deep crimson shadow of Decayed orange of the cloak Bright yellow highlight of Brilliant lime-green (mid-
Saint Peter’s robe. Christ’s pink robe. of the apostle leaning over the yellow cloak held to cover tone) of sleeve of the apostle
plate 12 plate 36 Christ’s shoulder. the nose of the man on the below Saint John.
• lead white with a few Appendix plates 39 plate 31 far right of the painting, near plate 20
particles of ultramarine and 40 • red lake glaze edge. • copper green glaze layer
and a trace of very • red lake (kermes lake • red earth combined plate 27 • lead–tin yellow with
faded lake pigment with a little madder), with yellow earth and • lead–tin yellow inclusions which look
at the surface of the powdered glass and an arsenic sulphide • yellow earth like lead soaps
sample a few particles of pigment • dark green/black • copper green combined
• azurite combined with ultramarine • yellow earth copper-rich layer with lead white, lead-
lead white, a little red • red lake (kermes lake • yellow earth combined tin yellow and dark
lake and a few particles with a little madder) with lead white and a earthy particles
of carbon black and powdered glass few red particles • mixed dark green layer
combined with lead of black particles with
white yellow earth and copper
green

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 49

Jill Dunkerton and Helen Howard

Blue and Purple Red Orange Yellow and Brown Green

Pale blue of Martha’s sleeve. Light pink of Christ’s pink Orange-red of Martha’s Brown of the foliage of a Lightest green of Saint John’s
plate 13 robe. cloak. bush growing out of the lower left sleeve.
• disrupted surface layer plate 37 plate 32 part of the rocky outcrop. plate 19
of lead white combined • lead white combined • orange-red layer of red plate 17 • thin layer of ultramarine
with ultramarine and with red lake earth and vermilion • thin dark brown layer combined with lead
possibly a now faded (kermes lake with a with lead white of red earth and black white
red lake little madder), with • trace drawing in black particles • verdigris combined with
• ultramarine combined powdered glass and • lead white priming • thick dark brown layer lead white and a few
with red lake and lead a few particles of of black and red earth black particles
white ultramarine particles • verdigris combined with
• pale pink layer of lead • lead white combined • pale green of landscape lead white and a few
white with a few tiny with red lake (kermes background comprising black particles
red particles lake with a little a copper green, lead-tin • at the right side of the
madder) and powdered yellow, yellow earth and sample a black layer
glass lead white is visible beneath the
• thin dark brown/black copper green. A large
layer composed of black particle of red lake is
and red particles, with also present at this level
one large red earth
inclusion visible
Central area of sky, above Red of cloak which Mary Bright orange of Saint Peter’s Flesh of Lazarus’s left leg, Moss-green of cloak of the
cloud and below foliage: Magdalene kneels upon. cloak. light area. old apostle with hands raised
brilliant mid blue. (not illustrated Sample 11) (not illustrated Sample 4) plate 8 behind Christ’s left shoulder.
plate 15 • vermilion combined • arsenic sulphide • pale yellow flesh tone plate 21
• ultramarine combined with lead white and a pigment of lead white with • yellow earth with
with lead white and a little red lake • red earth combined yellow and brown earth copper green and a few
little red lake with lead-tin yellow pigments black particles
• azurite combined with and lead white • dark brown flesh tone of • copper green with lead
lead white, red lake brown, red and yellow white and yellow earth
and a few tiny particles earth pigments with particles
of other red mineral some lead white
pigments • two layers applied
wet-in-wet, the upper
part containing earth
pigments with lead
white, while the lower
contains large silica-
rich particles and earth
pigments
• dark layer containing
earth pigments

50 | NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30

 Sebastiano del Piombo’s Raising of Lazarus: A History of Change

Blue and Purple Red Orange Yellow and Brown Green

Central area of sky, just Red of shirt of man Deeper, more yellow orange Pale green tunic of man
above pale buildings: very supporting Lazarus at of cloak of the apostle in leaning over Lazarus (light
dark blue/black area. bottom right (mid-tone). profile at the left edge of the tone, resampling of S. 23:
plate 16 (not illustrated Sample 14) painting (green beneath?). pale blue/green)
Appendix FIG. 8 • red layer of vermilion (Sample not illustrated) plate 23
• azurite combined with with a little red earth • red lead combined with Appendix FIGS 6 and 7
a very few particles of in a matrix of calcium a little red lake. This • lead white with a
lead white in a very carbonate, silica and a layer exhibits significant translucent yellow-
darkened binding little lead white lightening, particularly green copper-
medium • pale brown layer with at the upper surface, containing pigment,
large silica-rich particles due to the formation of ultramarine and a few
in a matrix of yellow lead soaps shiny dark particles of
and brown earth • at the left side of the galena (lead sulphide)
pigments with a few sample a copper green • thick layer of azurite
black particles layer divides the two with lead white a few
layers of red lead. particles of yellow/
The copper green is brown earth pigments
combined with a little • pale mauve layer of
red lake lead white with azurite
• red lead combined with and a few particles of
a few red lake particles. red lake
This layer exhibits
significant lightening
due to the formation of
lead soaps
Deeper (mid) blue of Saint Bright ‘emerald’ green of
Peter’s robe. sleeve of man with a turban
(not illustrated Sample 7) at top right of the painting.
• ultramarine combined plate 24
with red lake and lead • very thin pale layer on
white surface at the right side
• single azurite particle at of the sample which
the base of the sample may contain lead-tin
yellow as well as lead
white
• verdigris combined with
lead white
• verdigris combined with
lead white
• purple layer of azurite
combined with red lake
and lead white
Deep mauve of Martha’s Green of foliage along bottom
sleeve. plate 11 edge of picture. (Sample 8)
• ultramarine combined • verdigris with a little
with red lake and lead lead-tin yellow
white • lead-tin yellow with
• red earth combined verdigris
with ultramarine and a • mixed brown paint
little red lake and lead layer of red, yellow and
white brown earth pigments
Purple shading of deep
terracotta pink of cloak of the
apostle below Saint John.
plate 34
• purple layer of red
lake combined with
vermilion, lead white,
azurite and a few black
particles

NATIONAL GALLERY TECHNICAL BULLETIN VOLUME 30 | 51