Historical Painting Techniques, Materials, and Studio Practice

Historical Painting Techniques, 

Materials, and Studio Practice

PUBLICATIONS COORDINATION: Dinah Berland 

EDITING & PRODUCTION COORDINATION: Corinne Lightweaver 

EDITORIAL CONSULTATION: Jo Hill 

COVER DESIGN: Jackie Gallagher-Lange 

PRODUCTION & PRINTING: Allen Press, Inc., Lawrence, Kansas 

SYMPOSIUM ORGANIZERS: 

Erma Hermens, Art History Institute of the University of Leiden 

Marja Peek, Central Research Laboratory for Objects of Art and Science, 

Amsterdam 

© 1995 by The J. Paul Getty Trust 

All rights reserved 

Printed in the United States of America 

ISBN 0-89236-322-3 

The Getty Conservation Institute is committed to the preservation of 

cultural heritage worldwide. The Institute seeks to advance scientiRc 

knowledge and professional practice and to raise public awareness of 

conservation. Through research, training, documentation, exchange of 

information, and ReId projects, the Institute addresses issues related to 

the conservation of museum objects and archival collections, 

archaeological monuments and sites, and historic bUildings and cities. 

The Institute is an operating program of the J. Paul Getty Trust. 

COVER ILLUSTRATION 

Gherardo Cibo, "Colchico," folio 17r of Herbarium, ca. 1570. Courtesy 

of the British Library. 

FRONTISPIECE 

Detail from Jan Baptiste Collaert, Color Olivi, 1566-1628. After 

Johannes Stradanus. Courtesy of the Rijksmuseum-Stichting, 

Amsterdam. 

Library of Congress Cataloguing-in-Publication Data 

Historical painting techniques, materials, and studio practice : 

preprints of a symposium [held at] University of Leiden, the 

Netherlands, 26-29 June 1995/ edited by Arie Wallert, Erma 

Hermens, and Marja Peek. 

p. cm. 

Includes bibliographical references. 

ISBN 0-89236-322-3 (pbk.) 

1. Painting-Techniques-Congresses. 2. Artists' materials- 

-Congresses. 3. Polychromy-Congresses. I. Wallert, Arie, 

1950- II. Hermens, Erma, 1958- . III. Peek, Marja, 1961- 

ND1500.H57 1995 

751' .09-dc20 95-9805 

CIP 

Second printing 1996 

iv

Contents 

vii 

viii 

Foreword 

Preface 

1 Leslie A. Carlyle, Beyond a Collection of Data: What We Can 

Learn from Documentary Sources on Artists' Materials and 

Techniques 

6 Sylvana Barrett, Dusan C. Stulik, An Integrated Approach for the 

Study of Painting Techniques 

12 Helen Glanville, Varnish, Grounds, Viewing Distance, and 

Lighting: Some Notes on Seventeenth-Century Italian Painting 

Technique 

20 Ann Massing, From Books of Secrets to Encylopedias: Painting 

Techniques in France between 1600 and 1800 

30 Sally A. Woodcock, The Roberson Archive: Content and 

Significance 

38 Arie Wallert, Libro Secondo de Diversi Colori e Sise da Mettere a 

Oro: A Fifteenth-Century Technical Treatise on ManUSCript 

Illumination 

48 Erma Hermens, A Seventeenth-Century Italian Treatise on 

Miniature Painting and Its Author(s) 

58 Beate Federspiel, Questions about Medieval Gesso Grounds 

65 Renate Woudhuysen-Keller, Aspects of Painting Technique in the 

Use of Verdigris and Copper Resinate 

70 Josephine A. Darrah, Connections and Coincidences: Three 

Pigments 

78 Kate I. Duffy, Jacki A. Elgar, An Investigation of Palette and Color 

Notations Used to Create a Set of Tibetan Thangkas 

85 Zuzana Skalova, New Evidence for the Medieval Production of 

Icons in the Nile Valley 

91 Helen C. Howard, Techniques of the Romanesque and Gothic Wall 

Paintings in the Holy Sepulchre Chapel, Winchester Cathedral 

105 Eddie Sinclair, The Polychromy of Exeter and Salisbury 

Cathedrals: A Preliminary Comparison 

111 Andrea Rothe, Andrea Mantegna's Adoration of the Magi 

117 Ulrich Birkmaier, Arie Wallert, Andrea Rothe, Technical 

Examinations of Titian's Venus and Adonis: A Note on Early 

Italian Oil Painting Technique 

127 Catherine A. Metzger, Barbara H. Berrie, Gerard David's St. 

Anne Altarpiece: Evidence for Workshop Participation 

v

135 Molly Faries, Christa Steinbuchel, and J. R. J. van Asperen de 

Boer, Maarten van Heemskerck and Jan van Scorel's Haarlem 

Workshop 

140 E. Melanie Gifford, Style and Technique in Dutch Landscape 

Painting in the 1620s 

148 J¢rgen Wadum, Johannes Vermeer (1632-1675) and His Use of 

Perspective 

155 Ilze Poriete, Dace Choldere, A Technical Study of the Materials 

and Methods Used by the Painters of the Latvian Churches in the 

Seventeenth Century 

158 Melissa R. Katz, William Holman Hunt and the «Pre-Raphaelite 

Technique" 

166 Jo Kirby, Ashok Roy, Paul Delaroche: A Case Study of Academic 

Painting 

176 Joyce H. Townsend, Painting Techniques and Materials of Turner 

and Other British Artists 1775-1875 

186 Stephen Hackney, Art for Art's Sake: The Materials and 

Techniques of James McNeill Whistler (1834-1903) 

191 John R. Gayer, Painting on a Photographic Substrate: Notes 

Regarding Materials and Techniques over the Past 100 Years 

196 Ernst van de Wetering, Reflections on the Relation between 

Technique and Style: The Use of the Palette by the Seventeenth 

Century Painter 

204 Index of Contributors 

vi

Foreword 

One of the first events organized by the Getty Conservation Institute as 

it began its activities about ten years ago was a symposium on paintings 

conservation. We felt at that time that our particular approach, based on 

multiple disciplines looking at a single problem, could contribute significantly 

to the field of conservation. Over the course of the past years we 

have continued to develop in that direction and with that belief, so it is 

particularly appropriate to see these important preprints come to light for 

the University of Leiden's symposium. 

The history of painting techniques is by nature a multidiSCiplinary area 

of study, combining research in science, conservation, and art history as 

well as specific expertise in paintings. Members of each one of these professions 

bring to the area their own detailed knowledge in artists' materials, 

techniques, or methods, whether it be information specific to pigments, 

binding media, signature style, or archival research. 

The field of history of painting techniques has been evolving very rapidly 

in recent years and opening enormous opportunities for further research 

and connOisseurship, as this important group of papers demonstrates. Not 

only scientific methods of examination but also bibliographical and archival 

research are making significant contributions. 

The authors of these preprints bring a wide array of expert knowledge 

as well as many fresh points of view that are certain to provoke serious 

questions and debate. The editors-Arie Wallert, Erma Hermens, and 

Marja Peek-have assembled this volume with a most inSightful approach 

and focus. So it is with a great deal of satisfaction and pleasure that we 

present these papers for the study, and enjoyment, of a fascinating and 

challenging field. 

Miguel Angel Corzo, Director 

The Getty Conservation Institute 

vii

Preface 

This volume of preprints, prepared for an international symposium on 

Historical Painting Techniques, Materials, and Studio Practice at the University 

of Leiden, the Netherlands, 26-29 June 1995, contains the results 

of work on historical painting techniques from all parts of the world. The 

suggestion to organize such a meeting was raised during the symposium 

on the Technology of Art Works from the Central European Region organized 

by the Archives of Art Technology in Prague in 1993. The Prague 

symposium emphasized Czech painters and their techniques. To broaden 

the scope of attention, Erma Hermens of the Art History Institute of the 

University of Leiden and Marja Peek of the Art Historical Department of 

the Central Research Laboratory for Objects of Art and Science in Amsterdam 

took on the task of organizing a second meeting in Leiden. 

The purpose of the symposium and this publication is to promote a 

greater understanding of the changing boundaries and interaction between 

art historians, conservators, and conservation scientists working in the fields 

of historical painting techniques-including wall paintings and polychrome 

sculpture-painting materials, and studio practice. 

In recent years, there has been an increasing interest in historical painting 

techniques. The study of the painting techniques and materials used 

throughout history and in various cultures is by nature an interdisciplinary 

exercise. In the past such studies were sometimes conducted with little 

interaction between art historians, conservators, materials scientists, and 

historians of science, because each discipline tends to present the results 

of its studies to different forums. This volume aims to present different 

approaches to the study of historical painting techniques in the hope that 

it will encourage cooperation among these various disciplines. 

Information about painting techniques can be gained in a variety of 

ways, including the chemical or physical analyses of the materials found 

in the paintings. Analyses of a large number of paintings attributed to 

certain regions, schools, workshops, or individual masters can contribute 

to a history of painting techniques. Analytical results can also help art 

historians assess attributions and can support or reject their hypotheses. 

The analysts, however, need the art historians to inform them about the 

stylistic idiosyncrasies and significance of those schools, workshops, or masters. 

It is often rewarding to see what artists have said about their own work 

and to study their written sources on painting techniques. This kind of 

information has come down to us in diaries, such as Neri di Bicci's Ricordanze; 

in painting handbooks such as Cennini's Libra dell'Arte, in anonymous 

recipe books; and even in model books, such as Stephan Schriber's 

Musterbuch. These recipe books tell scientists what substances to look for 

in analysis. Both scientific and art historical information help us understand 

the significance of sometimes rather obscure recipes and tell us whether 

the methods and materials described are common or exceptional for the 

particular period. Today it is not unusual to find information from historical 

sources incorporated in the examination of individual paintings. 

viii

The discussions in this volume present historical painting techniques 

from a variety of professional perspectives. With its wide range of topics 

and approaches to the study of historical painting techniques, this publication 

is presented in the hope that it will provide an impetus for further 

studies that involve material science, art history, conservation, archaeometry, 

and the history of science. We also hope it will be one in a series 

of such interdisciplinary and collaborative volumes. 

In addition to thanking all the colleagues at our institutions who helped 

us with advice and expertise, we also would like to acknowledge the 

invaluable work and support of several individuals in particular. Miguel 

Angel Corzo, director of the Getty Conservation Institute, immediately 

supported the idea to publish the presentations of the symposium and 

generously provided his time and support in the production of the present 

volume. Agnes Grafin Ballestrem, director of the Central Research Laboratory 

for Objects of Art and Science in Amsterdam, and A. W. A. Boschloo 

of the Art History Institute of the University of Leiden supported 

the entire project from its initial stages and advised us during the editorial 

process. Corinne Lightweaver and Jo Hill, independent editorial consultants, 

carried out invaluable work on the manuscripts, assisted by Joy Hartnett. 

In the preparation of these preprints for publication, we also wish to 

thank Dinah Berland, publications coordinator, the Getty Conservation 

Institute. 

Arie Wallert 


Erma Hermens 

Art History Institute of the University of Leiden 

Marya Peek 

Central Research Laboratory for Objects of Art and Science, Amsterdam 

ix

Plate la, left. Bartolomeo Vivarini, active 145149 1, 

Saint Peter, 1490. Tempera on partel. From Polyptych 

with Saint James the Greater, The Virgin and Child 

and Other Saints. The]. Paul Getty Museum, Malibu 

(71.PB.30). 

Plate 1 b, above. Detail from Vivarini's Saint Peter, showing 

clearly visible, individual brush strokes.

Plate 2a, above right. Attributed to Raphael, 1483- 1520, Portrait 

of a Young Man, 1505. Oil on panel. The J. Paul Getty 

Museum, Malibu (78. PB. 364}. 

Plate 2b, right. Detail from Portrait of a Young Man, illustrating 

the clear, luminous color of the transparent oil techniques.

Plate 3a, above. Peter Paul 

Rubens, 1577-1640, The 

Virgin as the Woman of 

the Apocalypse, 1623. 

Oil on wood. The]. Paul 

Getty Museum, Malibu 

(85.PB. 146). 

Plate 3b, right. Detail from 

The Virgin as the Woman 

of the Apocalypse, illustrating 

the use of impasto 

white under highlighted 

areas.

Plate 4a, above. Jacques 

Louis David, 1748-1825, 

The Sisters Zenalde and 

Charlotte Bonaparte, 

1821. Oil on canvas. The 

J. Paul Getty Museum, 

Malibu (86.PA. 740). 

Plate 4b, right. Detail Jrom 

The Sisters Zenalde and 

Charlotte Bonaparte, illustrating 

the imperceptible 

shift from dark to light tones 

resulting from physically 

blending pigments on the 

suiface

Plate 5, hft. Colorjading tests, 1837- 

1878. Fitzwilliam Museum, UnilJwil)' 

of Cambridge. 

Plate 6, above. Pigmelll {(llilailicrs: j)"illl 

bladder to collapsible Illbe. Fil':'lI'illi,1/1/ 

Museum, Universil), of Clll//Jri(!c.

Plate 7a, above. Valerio 

Mariana da Pesaro, Battle 

of San Fabiano, 1618- 

1620. Photograph by E. 

Buzzegoli, Laboratorio di 

Restauro. Courtesy of the 

Ujfizi Gallery, Florence. 

Plate 7b, right. Detail 

from Battle of San Fabiano, 

1618-1620, showing 

the landscape in the 

background built up with 

several transparent washes 

of color. Photograph by E. 

Buzzegoli, Laboratorio di 

Restauro. Courtesy of the 

Ujfizi Gallery, Florence.

Plate 8. Gherardo Cibo, "Colchico, " from Herbarium (MS ADD 22333), jo lio 17r, ca. 

1570. Courtesy oj the British Library.

Plate 9. Marco Palmezzano, Mystic Marriage of Saint Catherine, signed and dated 

1537. Photograph by Christopher Hurst. Private collection.

Plate 10. Detail from Marco 

Palmezzano's Mystic Marriage 

of Saint Catherine, 

showing green glaze dabbed on 

St. Catherine's garment. Area 

shown measures approx. 3 X 

4 cm. Photograph by Christopher 

Hurst. Private collection. 

Plate 11. Cross section from PalmezzarLO's Mystic Marriage 

of Saint Catherine, showing of the green baldachin 

of the throne in normal light (X200). Photograph by Christopher 

Hurst. Private collection. 

Plate 12. Cross section from Palmezzano's Mystic Marriage 

of Saint Catherine, showing the oiling-out layer between 

the underpainting and the final glaze in UV light 

(X220). Photograph by Christopher Hurst. Private collection.

Plate 13, right. Surya 

in his Chariot, ca. 

1770, Bundi miniature. 

Smalt in the blue sky. 

Photograph by Paul 

Robins (Photo Studio), 

courtesy oj the Victoria & 

Albert Museum (D.379- 

1889). 

Plate 14, below. Krishna 

and Girls, 1730- 

1735, Basohli mInIature. 

Smalt with high arsenic 

content in the sky 

and Krishna's skin. Photograph 

by Paul Robins 

(Photo Studio), courtesy 

if the Victoria & Albert 

Museum (I.M. 87-1930).

Plate 15. The Buddha Shakyamuni 

Preaching at Dhanyakataka. 

Glue bound paint on cotton, 80.4 X 

3 cm. Denman Waldo Ross Collection, 

Museum of Fine Arts, Boston 

(06.333).

Plate 16. The Fifth King of 

Shambhala. Clue bound paint on 

cotton, 80.9 X 43.3 em. Denman 

Waldo Ross Collection, Museum of 

Fine Arts, Boston (06.324).

Plate 17, top. Icon of Virgin with Child Enthroned 

between Archangels, Nine Church 

Fathers and Nine Coptic Monks in the church 

of St. Mercurius Abu's-Saifain. This photograph 

documents its state of preservation in 1990. Photograph 

by Ayman el-Kranat, courtesy if the Archive 

of the church of St. Mercurius Abu's-Saifain. 

Plate 18a, b, above lift and right. Icon of Six 

Equestrian Saints in the church of St. Mercurius 

Abu's-Saifa in, lift and right sides, showing the 

beam's state of preservation in 1990. Photograph 

by Ayman el-Kranat, courtesy of the Archive of 

the church of St. Mercurius Abu's-Saifain. 

Plate 19, right. Detail of the head of Joseph of 

Aramathea from Deposition on the upper tier of 

the east wall. Photograph courtesy of the Conservation 

of Wall Painting Department, Courtauld 

Institute of Art, London.

Plate 20. Cross section oj Sample 3, taken 

Jrom the olive-green background near the leg 

oj Nicodemus in the Deposition on the 

east wall. The charcoal black inclusions 

within the plaster substrate are clearly visible; 

over this, a layer oj natural ultramarine 

is present. The present green appearance 

is due to the layer oj vivianite on the 

suiface oj the sample. Originally a blue color, 

some oj the particles have altered to a 

yellow color, and in the center oj some oj 

the yellow particles, the original blue color 

is just discernible. Photograph courtesy oj 

the Conservation oj Wall Painting Department, 

Courtauld Institute oj Art, London. 

Plate 21. Photomicrograph oj sample from 

central porch tympanum, Salisbury. Cross 

section shows complex structure, representing 

several repaintings and including vermilion, 

red and white lead, iron-oxide red 

and yellow, black, blue, and possibly gold. 

Plate 22. Photomicrograph oj sample Jrom 

Exeter west Jront. Cross section shows pink 

primer, white lead, and verdigris top layer.

Plate 23, top. Andrea Mantegna, The 

Adoration of the Magi. Distemper on 

fine linen, 54.6 X 69 .2 cm. The). Paul 

Getty Museum, Malibu (85.PA.417). 

Plate 24, lift. Detail from The Adoration 

of the Magi, upper lift-hand comer, 

showing feathered-out colors on the edges 

of the painting, originally under theframe. 

The ). Paul Getty Museum, Malibu 

(85.PA.417). 

Plate 25, above. Detail from The Adoration 

of the Magi, lower edge, showing 

feathered-out colors on the edges of the 

painting, as in Plate 24.

Plate 26, top. Titian, Venus and Adonis. 160 X 196. 5 em. The] . Paul Getty Museum, 

Malibu (92.PA.42). 

Plate 27, bottom. Titian, Venus and Adonis. 107 X 136 em. Courtesy of the National 

Gallery of Art, Washington, D.C., Widener Collection (1942.9.84 [PAJ).

Plate 28. Gerard David and Workshop, St. Anne Altarpiece, ca. 1506: lift, St. Nicholas, 236 X 76 cm; center, St. Anne with the 

Virgin and Child, 236 X 96 cm; right, St. Anthony of Padua, 235 X 76 cm. Courtesy of the National Gallery of Art, Washington, 

D. C. Widener Collection.

Plate 29. Cross sections of the St. Anne Altarpiece, shown in Plate 28. Cross section (a) 

through the underdrawing line below the cherub's foot (photographed at X100) shows fo ur 

layers: ground, underdrawing, pale gray, and blue-gray. Cross section (b) through a line in 

St. Anne's red drapery (photographed at Xl 00) shows multiple layers: ground, type III 

underdrawing, two layers of orange red, and multiple layers of red glaze (the layering is not 

observable in normal light). Cross section (c) through the "correction" layer in the Virgin's 

sleeve (photographed at X 100) shows five layers: ground, underdrawing, pale underpaint, 

"correction" layer, and dark violet glaze. Sampling by B. Miller. Cross section (d) through 

the Virgin's blue drapery (photographed at Xl 00) shows fo ur layers: ground, underdrawing, 

pale blue, and dark blue. Cross section (e) through the dark green part of the cloth of honor 

shows seven layers: ground, scattered black particles, blue-green, dark green, dark-green and 

green glazes, and discolored varnish. Cross section (f) through St. Anthony's cassock (photographed 

at Xl 00) shows three layers: ground, underdrawing, and violet-gray. Sampling by 

B. Miller.

Plate 30. Cross section from Baptism, 

near an edge: layer 1 (at the bottom), intermediate 

white layer, here thicker; layer 

2, pink layer with madderlike particles; 

layer 3 (on top), white with natural ultramarine. 

Photograph by]. R. J. van Asperen 

de Boer. 

Plate 31. Paint cross section, light green 

Jo liage, upper part oj central trees. Workshop 

oj Jan Bruegel, Noah's Ark. Layers 

Jrom bottom: (1) chalk ground; (2) imprimatura: 

red and black in a translucent medium; 

(3) sky: white lead, pale smalt, 

black; (4) Joliage underpaint: azurite, 

earth, white lead, black. Magnifrcation on 

35 mmfilm: X64. 

Plate 32. Paint cross section (dark sky at 

upper edge) from Esaias van de Velde's 

Winter Landscape. Layers Jrom bottom: 

(1) ground: chalk with white lead, earth, 

and black; (2) sky: white lead, smalt, and 

earth black; (3) darker sky: smalt and 

white lead; (4) overpaint. Magnifrcation 

on 35 mm film: X70.

Plate 33. Paul Delaroehe, The Execution of Lady Jane Grey, 1833. Canvas, 246 X 29 7 em. London, National Callery (1909).

Plate 34a, b, c, d. Paint sampLes Jrom 

The Execution of Lady Jane Grey: (a) 

cross section, pinkish red oj executioner's 

tights, original magnification X540; (b) 

cross section, grayish shadow on Lady Jane's 

dress, original magnification X750; 

(c) cross section, red and brown brocade 

draped over Lap if seated attendant Left, 

originaL magnification X750; (d) unmounted 

fragment, bLack oj Sir John Brydges's 

gown, photographed from the reverse 

at X275.

Plate 35, right. Detail Jrom the lower lifthand 

comer oj J. 1'1'1.. W. Turner's The 

Dawn of Christianity shows where Turner 

used the canvas to try out colors. 

Plate 36, below. Detail oj sheep Jrom 

William Holman Hunt's Strayed Sheep 

(Our English Coasts), 1852. 432 X 

584 mm. Courtesy oj the Tate Gallery, 

London (N05665).

Plate 37, lift. James Whistler, Arrangement in 

Flesh Color and Black: Portrait of Theodore 

Duret, 1883-1884, oil on canvas. Wolfe Fund, 

Catharine Lorillard Wolfe Collection, 1913 

(13.20). © 1984 by The Metropolitan Museum 

if Art. 

Plate 38, above. Detail of a reflection from Whistler's 

Nocturne: Blue and Silver, showing that 

the paint has been dragged horizontally when both 

the white overpaint and blue underlayer were still 

wet, to produce the iffect if light on the water surface.

Abstract 

This paper argues for an interdisciplinary 

approach to the study of artists' 

painting materials and painting 

practices. Recent research into British 

documentary sources on nineteenth-century 

oil painting reveals 

information useful in technical examinations 

of paintings. Examples 

illustrate how important a full understanding 

of the artist's physical 

environment and contemporary beliefs 

can be in accurately interpreting 

evidence from a painting. 

Beyond a Collection of Data: What We Can Learn 

from Documentary Sources on Artists' Materials 

and Techniques 

Leslie A. Carlyle 

Canadian Conservation Institute 

1030 Innes Road 

Ottawa, Ontario KiA OC8 

Canada 

Introduction 

In the preface to the catalogue for the National Gallery of London's exhibition, 

Art in the Making: Impressionism, the sponsor remarks, "We see not 

just painted surfaces, but are given a multi-disciplined information which 

brings the paintings themselves to life" (1). The desire to know more about 

the whole object, to go beyond the image, the preface continues, to get 

"behind the pictures, and even through them," is very much a feature of our 

era, just as the delight in finding the "real" information hiding behind our 

conventional views fu els so many contemporary investigative endeavors in all 

disciplines. Our enthusiasm for complete knowledge is something we have 

in common with our predecessors of the late eighteenth century. 

Fresh from the age of enlightenment, one author of a late eighteenth-century 

technical manual on oil painting writes in his preface (2): 

Facts judiciously arranged, and published from time to time as they accumulate, 

are productive of infinite advantage . ... Every branch of science 

is much fa cilitated and advanced by public communication, which distinctly 

points out the present, and opens a free channel to future discoveries . ... 

Records oj this kind act, therefore, as stimulants to general improvement: 

what is already known need not be retraced, and what is discovered in 

future [sic] may be occasionally added: thus, the needy and diffident will 

be taught with oeconomy [sic] and ease, and mystery will be unfolded and 

converted into truth. 

Few today would argue for such a completely linear view of knowledge, but 

the desire to "know all" has not left us. 

Nineteenth-century technical literature on oil painting materials and practices 

shows that the search for and collection of the "facts" was underway in earnest 

at that time. By the end of the century, however, we find the optimism 

somewhat chastened: all the new chemical knowledge, all the new facts, still 

could not provide oil painters with any guarantees, once and fo r all, for the 

durability of their work. 

A similar sort of optimism existed for those in the twentieth century who 

examined the role of science in unraveling the mysteries of oil paintings. The 

belief was that with enough instruments and enough analysis, we could know 

the secrets of the old masters, we could know of what a painting is made. 

However, as one of our discipline's critics Michael Daly points out, "Such 

technical analysis can only ever say what a material is, never what its purpose 

was" (3). 

Although scientific instrumental analysis is a highly sophisticated branch of 

inquiry in itself, results from it alone are not sufficient. It is only in partnership 

with other forms of investigation that we can hope to unravel the meaning 

behind what we find through analysis. Now, nearing the end of the twentieth 

century, having penetrated much of the "mystery" and converted it into truth, 

we find that the gifts of science are not enough. It is to "multidisciplinary 

information" that we turn in order to understand the purpose of the materials 

we find. 

Carlyle 1

Fortunately, the spirit of the eighteenth century in Britain fired a great enthusiasm 

for publications on technical matters, including the materials and 

techniques of oil painting. Thus, we find a rich source of information in the 

various treatises, manuals, and handbooks that continued to be published into 

the nineteenth century. The following will be a discussion of the kind of 

information these documentary sources can provide and how this information 

can influence our interpretation of cross sections and analytical results as well 

as fu rther our understanding of painters' techniques. 

The painter's environment 

The anonymous eighteenth-century author quoted previously placed great 

faith in "facts judiciously arranged, and published from time to time as they 

accumulate." Yet, however important a discrete piece of information, such as 

the date of introduction of a new pigment, may be in the study of painters' 

instruction books, it is not always this information that provides insight into 

the painter's choice of materials. Sometimes it is the tangential information 

about the experience of living at a given time that provides a context fo r 

what we observe now. 

In cross sections of paint, the build-up of dirt between layers of paint must 

be interpreted in relation to past conditions for lighting and heating. We 

cannot, based on our present-day experience, extrapolate from the thickness 

of a dirt layer the length of time between episodes of painting. Here, those 

who study the history of technology and of domestic life are of great help. 

In one source on the history of domestic environments, we find that even as 

early as 1700 the use of coal fo r heating in London resulted in a "Tartanous 

Smoak" that sullied the environment both indoors and out: "All sorts of 

Hangings, especially the Tapestry, are in a few Years totally defil'd by it ..."(4). 

Because painters who fo llowed the technique of "painting in stages" were 

obliged to wait between applications of paint for the underlayers to dry, a 

fairly rapid build-up of dirt could be expected under the conditions described, 

far more than our late twentieth-century environments would convey. 

The level of air pollution in the days of coal heating also caused great concern 

among artists and their chemist advisors with regard to the role of lead in 

paintings. It was thought that the high levels of sulfur in the air caused reactions 

with lead-white pigment and with lead-treated oil, resulting in an 

overall darkening of these materials due to the reaction product, lead sulfide. 

Various solutions to this problem, including the application of nonreactive 

zinc white over lead-white underlayers, were recommended. This advice to 

apply zinc white over lead white was given not only for paint layers, but fo r 

grounds as well; it was believed that a lead-white ground preparation could 

also be subject to darkening. Cross sections taken from a painting in which 

this advice had been followed show layers of two different white paints, the 

presence of which would not be immediately obvious without the knowledge 

of the remedial steps taken to obviate the so-called lead-sulfide darkening (5). 

Aside from the dirt and soot from coal heat and tallow candles, interior 

environments were also substantially colder in the winter months. In the 

absence of central heating, painters found that their colors dried significantly 

more slowly during the winter months, hence the advice to add materials 

that hasten drying at this time of year (see below) . 

Beliefs influencing artists' practices 

Artists' practices were also influenced by views and beliefs that are foreign to 

our era. Conservators have discovered empirically that it was not uncommon 

for nineteenth-century painters to use similar varnishes in the paint medium 

to those used as a final varnish. In the literature, painters were quite explicit; 

they believed that using the same resin in the medium as in the final varnish 

would, by ensuring homogeneity of materials, reduce the likelihood of cracking 

(6). In a pharmaceutical dictionary and recipe book published in 1764, 

2 

Historical Painting Techniques, Materials, and Studio Practice

we find a description of the affinity of like materials that could well be the 

antecedent of this nineteenth-century practice (7) . 

The widespread addition of varnish to the oil painting medium was itself 

fostered by the belief that this was the method used by the old masters to 

achieve the particular translucent quality observed in their paint (8). 

Seeing the past through present-day lenses 

David Lowenthal, in his book The Past Is a Foreign Country, wrote (9): 

However fa ithfully we preserve, however authentically we restore, however 

deeply we immerse ourselves in bygone times, life back then was based on 

ways oj being and believing incommensurable with our own. We cannot 

help but view and celebrate it through present day lenses. 

Just as we can enhance our understanding of artists' practices by learning 

more about the implicit assumptions they made based on the beliefs common 

to their era, we must be especially careful not to impose our own assumptions 

on the past. For example, the identification of fu gitive colors in a nineteenthcentury 

painting could lead to the conclusion that the artist knowingly opted 

for fugitive colors. Because information regarding which pigments were unstable 

was widely available at this time, the presence of these colors in a 

painter's work must mean that the painter did not "care" if the colors would 

fade. But this is not necessarily true; in fact, the painter may well have conscientiously 

purchased stable colors, but may have been unknowingly supplied 

with substituted materials by the colormen. 

A thorough study of the literature, combined with scientific analyses of paint 

samples, revealed that in the nineteenth century the name of a color did not 

always offer a reliable indication of composition. Naples yellow is a good case 

in point. Traditionally a lead-antimony compound, by the late nineteenth 

century Naples yellow was reported to have been substituted with more reliable 

coloring agents: lead white and cadmium yellow. But the sample labeled 

Naples yellow in a Winsor & Newton oil-sample book actually consisted of 

lead white, red lake, and yellow lake. Since nineteenth-century lake colors 

were not particularly stable to light, this particular Naples yellow would be 

unlikely to retain its hue indefinitely (10). 

Documentary research also indicates that media analysis sampling that is restricted 

to one color area cannot be assumed to apply to the whole painting. 

Early instruction books, in which the artists were still instructed in grinding 

their own paint, indicate that the paint medium would be changed according 

to the character of the pigment used. Because of its initial yellow color and 

because it was believed to after-yellow the most, linseed oil was generally 

recommended for dark colors, and the less-colored poppy and nut oils were 

reserved for light colors. There were exceptions, however. Lake colors dried 

slowly; therefore, linseed oil-the fastest of the three to dry-was recommended 

for use with these "light" colors (11). Drying oil in combination 

with copal varnish was recommended as the medium for Kings yellow or 

orpiment (12). Many other pigments received individual treatment and admixtures 

with varnish (13). 

As we have seen, the choice of oil depended not only on the pigment used, 

but also on the season. Oils treated either by boiling alone or in conjunction 

with metallic compounds (driers) to hasten their drying time were sometimes 

recommended for the winter months only, when damp, cold weather lengthened 

drying time. Conversely, painters were warned that such treated oils 

would "in summer ... dry so soon as to be troublesome" (14). Since the 

essential nature of oil paint had not changed by the early nineteenth century, 

it is not surprising to find this kind of advice appearing in print much earlier. 

In 1693 Marshall Smith recommended, "If in the hottest weather your greatest 

Dryers dry too fast, as White, Umber, &c and so grow too stiff to work 

with, you may prevent it by mixing a little Sallat Oyle with Colours" (15). 

Carlyle 3

Just as we cannot extrapolate to the whole painting from media analysis in 

one color area alone, documentary sources indicate that the medium could 

change not only from color to color, but also from paint layer to paint layer. 

Once again, this was not confined to the nineteenth century. Marshall Smith 

instructed that lead white be mixed with nut oil, but noted that linseed oil 

could be used in dead-coloring (16). Advice to vary the medium according 

to the layer continued to appear in the literature, the faster-drying linseed oil 

again being recommended for underlayers such as dead coloring, with poppy 

or nut oil in the finishing layers (17). 

Interestingly, the medium for the first lay or dead coloring need not have 

been oil at all. There were references to the use of watercolor, egg tempera, 

and a combination of two-thirds starch to one-third oil (18). 

Contemporary experience with oil painting materials may also lead to assumptions 

that require examination. Today, if we wish to prepare a "traditional" 

lead-white ground, the first step would be to size the canvas using a 

hide glue such as rabbit-skin or parchment size. Although there are indications 

that the use of glue size, including isinglass, was common in the past, this was 

not the only material used. Starch was also employed as a size layer and 

appears in recipes throughout the nineteenth century. There were also indications 

that the addition of a plasticizer or humectant such as honey, sugar, 

or glycerine would not have been unusual. Near the end of the century, we 

find a reference to the use of collodion (cellulose nitrate) as a replacement 

for the size layer (19). 

Our present-day lenses can also result in our underestimating the importance 

of materials that in our own day are no longer in use or have become precious 

and rare. Isinglass, a glue prepared from the swim bladder of the Russian 

sturgeon, is not widely available today. In eighteenth- and nineteenth-century 

England, isinglass was commonly used for a variety of purposes: to clarifY 

beer, wine, and soup, and as a sizing agent for fabric, ribbons, and paper. 

Therefore, the presence of what is now quite rare, but was then a relatively 

commonplace glue in the size layer for an oil painting, is not surprising. 

Another such material that has dropped out of use entirely is sugar of lead 

(lead acetate). A white crystalline powder widely used as a drier fo r oil paint, 

it was added directly to the pigment-oil mixture and was also present in 

medium recipes. Lead acetate could be purchased easily from apothecaries 

and appears to have been in wide use by painters in the late eighteenth and 

nineteenth centuries. By the twentieth century, however, it was never mentioned 

in sources on oil painting materials and techniques, although other 

traditional lead driers, such as litharge or metallic lead, do receive notice. As 

a result, the important role that this material played has never been acknowledged 

or studied in the twentieth century. 

Conclusion 

Although we must accept Lowenthal's observation that life in the past was 

"based on ways of being and believing incommensurable with our own," we 

should not see the exercise of studying past practices and materials as fu tile 

(20). Rather, we should equip ourselves with the knowledge that we are 

handicapped by our late twentieth-century standpoint. By making use of a 

variety of disciplines, by not concentrating our energies too much on only 

one avenue of inquiry, we can continue the search to "know all." As much 

as possible, we should look outside of our immediate disciplines for researchers 

who are also mining the past, as it is this multidisciplinary approach that 

will enrich our understanding and interpretation of the "facts." 

Notes 

1. Forester, Sir Archibald. 1990. Sponsor's preface. In Art in the Making: Impressionism, 

D. Bomford,]. Kirby, ]. Leighton, and A. Roy. London: The National Gallery 

and Yale University Press. 

4 


2. A Practical Treatise on Painting in Oil Colours. 1795. London, v. 

3. Daly, M. 1993. The varnished truth. Art Review. Vol. XLV (November): 57. 

4. Davidson, C. 1986. A Woman's Work 15 Never Done. London: Chatto & Windus, 

132. 

5. Carlyle, L., and J. Townsend. 1990. An investigation oflead sulphide darkening 

of nineteenth-century painting materials. In Dirt and Pictures Separated, London: 

United Kingdom Institute fo r Conservation, 40-43. 

6. Carlyle, L. 1990. British nineteenth-century oil painting instruction books: a 

survey of their recommendations fo r vehicles, varnishes, and methods of paint 

application. In Cleaning, Retouching, and Coatings: Technology of Practice fo r Easel 

Paintings and Polychrome Sculpture. London: International Institute for Conservation, 

79. 

7. "Experiments concur with daily Observation to prove, that different Bodies, 

whether Principles or Compounds, have such a mutual Conformity, Relation, 

Affinity, or Attraction, as disposes some of them to join and unite together, 

while they are incapable of contracting any Union with others. Substances having 

an affinity together, will unite and form one Compound. It may be laid 

down as a general Rule, that similar Substances have an Affinity with each other, 

as Water with Water, Earths with Earths, &c." James, R. 1764. Pharmacopoeia 

Universalis: Or, A New Universal English Dispensatory. London, 11. 

8. Carlyle, L. 1991. A critical analysis of artists' handbooks, manuals and treatises 

on oil painting published in Britain between 1800-1900: with reference to selected 

eighteenth-century sources (Ph.D. diss., Courtauld Institute of Art, University 

of London), Vol. I, 27-28. 

9. Lowenthal, D. 1985. The Past 15 a Foreign Country. Cambridge: Cambridge University 

Press, xvi. 

10. Townsend, J., L. Carlyle, S. Woodcock, and N. Khandekar. n.d. Later nineteenth-century 

pigments: analysis of reference samples, and evidence for adulterants 

and substitutions. The Conservator. United Kingdom of Conservation. 

Forthcoming. 

11. Templeton, J. S. 1846. The Guide to Oil Painting. London, 28. 

12. Ibid., 58. 

13. Carlyle 1990, op. cit., 76. 

14. Williams, W. 1787. An Essay on the Mechanic of Oil Colours . ... Bath, England, 

47. 

15. Smith, M. 1693. The Art of Painting . ... London, 73. 

16. Ibid., 71. 

17. For example, "The oils are the mixture of oil and turpentine; and as the portrait 

advances towards the finishing sitting, nut or poppy oil may be substituted in 

the mixture fo r boiled oil." Cawse, J. 1822. Introduction to the Art of Painting in 

Oil Colours. London, 15. 

18. Carlyle 1991, op. cit., 288. 

19. Ibid., 247. 

20. Lowenthal, op. cit. 

Carlyle 5

Abstract 

Methods of art historical research, 

painting expertise, and scientific research 

can contribute to the detailed 

study of painting techniques. Conversely, 

the knowledge of painting 

techniques could support art historical 

research. The need for collaboration 

between individual disciplines 

during all stages of research is 

strongly stressed, from the initial formulation 

of working questions and 

hypotheses to final conclusions about 

the techniques of an artist. 

An Integrated Approach for the Study 

of Painting Techniques 

Sylvana Barrett and Dusan C. Stulik* 


4503 Glencoe Avenue 

Marina del Rey, California 90292 

USA 

Introduction 

The study of artists' techniques in general and the study of an individual 

artist's techniques in particular are important for several reasons: 

(a) Art historians can use detailed knowledge of an artist's technique and its 

developmental evolution throughout the artist's career in the authentication 

process. This information can also aid in the establishment of a proper chronology 

for the known works of a given artist. 

(b) Artists of various historical periods were able to achieve specific visual 

effects by the use of special artists' materials or by methodical application of 

proven painting techniques. Because detailed documentation is seldom available 

today to help artists learn the steps needed to re-create a given visual 

effect, artists must rely on the results of systematic art research to learn old 

master techniques. 

(c) Museum conservators, in order to ensure a safe working strategy when 

planning a conservation or restoration treatment, rely on specific information 

about pigments, binding media, and materials, including those of earlier restorations, 

as well as detailed knowledge of the structural arrangements of these 

materials. 

The art historian and art research 

The role of the art historian in the realm of art research is critical. It is the 

art historian who must set the foundation, into which the information gathered 

by individual researchers of a painting-technique research team is organized, 

for final interpretation. In the study of painting techniques, two tools 

used by art historians are very important: connoisseurship and archival research. 

Connoisseurship. Beyond establishment of individual and historical chronologies, 

connoisseurship itself-the expert knowledge of style and technique that 

the art historian develops through the course of a career-is of immense help 

in the research of artists' techniques. This keen sense of discrimination can 

be used to identify idiosyncrasies particular to a given artist's work. When 

drawn from works of undisputed provenance, the art historian can use the 

idiosyncrasies to establish a signature of style, materials, and techniques for 

any given artist. This "signature style" is critical for researchers in all the 

related disciplines. The signature style for a particular artist establishes the 

standard of measure against which all data can be judged. 

Archival research. The study of primary documents-such as municipal, guild, 

or financial records and chronicles in which artists' names can be directly 

located-provides crucial information about artists' lives, training, professional 

and social standing, and other socio-economic factors that influenced the 

development of their working methods and personal painting styles. When 

an artist's notes, letters, books, diaries, and travel journals are available, the 

study of these materials often provides important information leading to 

* Author to whom correspondence should be addressed. 

6 


knowledge and understanding of the artist's intent, as well as the rationale 

behind the use of a particular material and painting technique. 

The study of primary sources-such as books and treatises on painting, old 

recipes for the preparation of artist materials, and guild practices and procedures-can 

contribute to our understanding of the artists' working environment. 

The study of such documents can be invaluable, but extreme caution 

should be applied when interpreting old manuscripts. A number of misunderstandings 

have arisen because of the often confusing nomenclature used 

in old texts and recipes and the distorted facts found in old biographies of 

artists. Even so, the study of secondary documents, such as biographies of 

painters, provides important information of the creative, social, and economical 

environment in which the artist worked. 

Artists and art research 

A painting is composed of elements that can be separated for study. As there 

are schools of artistic thought, there are schools of painting technique. Each 

school of painting technique has a specific procedural approach to painting 

construction. Although there are many elements common to all painting 

techniques, there are also specific elements unique to each. It is possible to 

isolate and define these unique qualities for each technical approach and to 

establish markers for a detailed study. These markers, when encountered during 

an examination, provide keys to the likely construction of visual effects 

within the piece. 

An initial, standardized visual examination of a painting by an artist specializing 

in painting techniques can id in any art historical or subsequent scientific 

investigation, not only helping to orient the researcher but also assisting 

in the interpretation of the results. If, for example, it was determined through 

visual examination that a particular piece was a multilayered, glazed construction 

on panel, it might then be assumed, based on knowledge about the 

particular technique, that the piece would have an oily imprimatura on a 

gesso ground. Knowing this beforehand, a researcher who discovered an oily 

component in the ground layer might investigate the possibility of its having 

been absorbed from the imprimatura by the lean ground rather than assuming 

the discovery of a novel gesso recipe. 

Technical approaches used for creating the illusion of volume 

A basic goal of all representational artists-to present an illusion of volumeis 

accomplished in painting through the juxtaposition of dark and light values, 

and of highlights and shadows. This illusion is accomplished traditionally 

through one of four basic techniques. A brief analysis of these techniques will 

illustrate the possibilities of standardizing the visual examination of paintings 

and the usefulness of the visual markers that can be established as a result of 

this approach. 

Basic technique. In a direct approach, dark and light values are placed by single, 

individual brush marks onto the surface of the painting. There is no blending 

of the pigments themselves. Dark values are used to indicate shadow and light 

values to indicate highlights, effectively indicating volume. Despite the obvious 

simplicity of this approach, it can be highly successful. The tempera 

paintings of the Italian Trecento illustrate the effectiveness of this technique. 

Visual markers for this technique include a uniform surface of clearly defined, 

individual brush strokes that retain their original distinct color and do not 

physically blend into surrounding pigments (Plate la, b). 

Transparent oil technique. This more complex, systematic approach, which was 

developed with the advent of transparent oil media, is exemplified by the 

Flemish and early Netherlandish masters. These artists conceived of the painting 

from its inception as a multilayered object with a structural separation of 

color and fo rm. Volume, developed through highlights and shadows in a monochromatic 

underpainting, was followed by color embellishments. 

Barrett and Stulik 7

Well aware of the optical properties of both light and color, the artist worked 

on a highly reflective, white ground layer. The underpainting could be a 

complete gray-toned version of the finished image, painted in a manner such 

as that described in the preceding basic technique. It could also be constructed 

through a more sophisticated technique, as seen in the unfinished panel of 

Santa Barbara by Jan van Eyck. An underdrawing, which establishes contours 

and darks on the white ground layer, is covered with an imprimatura, a thin, 

transparent layer of paint that allows the drawing to show through the ground 

while also establishing a middle tone throughout the painting. Highlights 

could then be added in white paint where appropriate, thus, with less work, 

completing the values and creating a finished monochromatic underpainting. 

Regardless of the approach taken toward the underpainting, its creation was 

essential to the technique itself. Color applied as thin transparent glazes allowed 

the fully developed underpainting to define the fo rms while the color 

itself remained clean, pure, and unadulterated. Highlighted areas could be 

achieved with the thinnest possible application of local color, as the white of 

the underpainting had merely to be tinted appropriately. Dark tones, however, 

posed some problems with the clear transparent pigments: many layers were 

required to cover the underdrawing and establish the proper local color. 

By fo cusing on these highlights and shadows, visual identification of the technique 

is quite simple. Highlighted areas are very thin and fine. Color applied 

in thin glazes tends to be clear, luminous, and devoid of brush marks. Shadows 

and dark colors, however, appear as thickly built-up surfaces, creating ridges 

clearly visible in raking light where they come into contact with the delicate 

light areas (Plate 2a, b). 

Highlighting with impasto white. Allowing a freer painting style and facilitating 

larger formats, this more flexible technique is typical of the Baroque masters. 

Any support suffices; the underdrawing is optional. The artist tones the surface 

with a middle or darker value, then creates the image with an underpainting 

of washes that may be controlled or completely free and spontaneous. The 

areas of the painting to be highlighted are now created with a heavy impasto 

white paint. 

This simple procedure accomplishes the same optical effects as the complete 

monochromatic underpainting of the previous transparent oil technique, yet 

it allows the image to evolve as it is constructed. The continued separation 

of value from color still allows for beautiful luminous color. Because the image 

originates in the loose, dark washes, contours need not be highly defined and 

extreme chiaroscuro is possible. The resulting work is often quite dramatic in 

nature. 

This technique provides very specific optical markers. The darks are thin and 

transparent, often revealing the preliminary wash or imprimatura. The highlights 

that define the volume appear thick and visibly raised from the painted 

surface (Plate 3a, b). 

Direct suiface blending. In the controlled technique of surface blending, individual 

colors and values are mixed and applied to appropriate locations of 

the surface to indicate highlight and shadow. Each new application of color 

is carefully blended into the surrounding paint, resulting in a smooth, continuous 

flow. The underdrawing and underpainting serve only as a guide for 

the surface painting; they do not actively affect the surface itself. 

All traces of brushwork can be blended out if desired; consequently, the technique 

lends itself well to smooth, detailed, controlled styles such as found in 

the work of the Neoclassicists. 

Visual markers fo r the technique include a smooth, continuous surface with 

gradual, imperceptible shifts from highlight to shadow. Direct blending of the 

pigments creates an opaque quality in contrast to the luminous character of 

colors in the multilayered approaches (Plate 4a, b). 

8 


Table 1. Scientific methods used Jo r identification oj critical, unaltered parts oj a painting. 

M ethod In ormation LimitalJon 

observatIOn under UV light I dark areas of repamlJng III - does not worK lor all 

comparison wilh light areas 

binding media 

of heavily oxidized oil paints - possible interference with 

(autofluorescence) 

some varnishes 

UV photography documentation of repainted as above 

areas 

false color infrared 

identification of repainting if only for orientation, should 

photography done using pigments of the be confirmed using other 

same color but different methods of chemical 

chemical composition analysis 

X -ray radiography localization of support high contrast only for heavy 

alterations, ground and paint chemical elements (Pb, Hg, 

layer repairs and alterations Au) 

X-ray f1uorescence - identification of repainted - possible difficulties with 

spectrometry (XRF) areas if done with inorganic mixed pigments and 

pigments of different complex multilayer paint 

composi tion than original structure analysis 

(e.g., zinc or titanium white - no analytical information 

instead of lead white) for low atomic weight 

- a non contact analysis elements (Z :$ Na) 

microchemIcal methods Identification of repainted - microsampling 01 paint 

(polarized light microscopy, areas based on pigment material needed 

chemical microscopy, identification - uncertain identification of 

electron microprobe, X-ray 

some organic pigments and 

diffraction) 

dyes 

The visual traces left by the physical manipulation of painting materials on 

paintings can be identified fo r study. Idiosyncrasies of color, brushwork, paint 

consistency, form, and so on can each be evaluated to reveal specific unique 

qualities within given predefined techniques. This information can then be 

used for the general analysis of historical paintings. The works of individual 

artists can be analyzed in a similar manner, identifYing a signature style and 

painting procedure through specifiC visual clues. 

The role of scientific and technical examinations 

The process of applying scientific methodology to the study of painting techniques 

can be divided into the following three major steps: (1) identification 

of the critical (unaltered) parts of the painting, (2) authentication of the painting, 

and (3) study of detailed physical and chemical structure of paintings. 

Identification oj the critical (unaltered) parts of the painting. The majority of paintings 

in museums or private collections have a long history of cleanings, restorations, 

and alterations. Before embarking on a study of painting techniques, 

it is essential to identifY areas of the painting in which the painting technique 

of the original artist has not been altered by later treatments. Scientific methods 

for such a study are described briefly in Table 1. 

Authentication oj the painting. When a series of paintings by a particular artist 

is examined with the goal of studying painting techniques, it is crucial that 

the authorship of the pieces is established "beyond a reasonable doubt." Scientific 

research cannot establish the relationship between the artwork and the 

artist. What scientific research can do very successfully is to effectively eliminate 

paintings that, based on clearly defined scientific facts, could not have 

been created during the active life of the artist in question. Several powerful 

scientifiC methodologies that can be used to help authenticate paintings are 

described in Table 2. 

Study of detailed physical and chemical structure of paintings. Usually, only the top 

paint layer, with its corresponding brushwork and surface treatment, is accessible 

to visual observation. This is a very severe limitation when studying 

painting techniques because before the alia prima technique became widely 

used during the second half of the nineteenth century, the majority of paintings 

were created in complex multiple-step and multilayer processes. Table 3 

shows several scientific methodologies that can be used in probing and analyzing 

such painting structures. 


Table 2. Scientific techniques Jor the authentication oj paintings. 

VISI ble observation and X- visible and hidden tool - knowledge 01 ancient 

I Method Inlonnation Limitation 

ray radiography marks; materials (machine technologies required 

made canvas, nails, etc.) - new repairs may pose a 

problem 

pigment chronology identification of the - detailed knowledge of 

inconsistency studies systematic use of pigments pigment chemistry and 

(microchemical methods) in the painting which were technology required 

not available in times when - problem of inteIpretation 

the painting was presumed when only repainted areas 

to be painted 

analyzed 

- the negati ve resul ts cannot 

be considered a certain proof 

of authenticity 

radiocarbon dating absolute dating method of - does not provide good 

the actual age of natural results for materials younger 

product organic materials in than 300 years 

paintings (wood, canvas, - prOblems of impurities 

binding media) 

dendrochronology or tree absolute dating method quality sample containing a 

ring dating actual age of wood sam pIes number of easily measured 

tree rings and corresponding 

calibration data is needed 

INTEGRATED APPROACH 

FOR THE STUDY OF 

PAINTING TECHNIQUES 

Figure 1. Diagram of integrated approach for 

the study of painting techniques. 

Conclusion 

Art historians, conservation scientists, and artists use different means to study 

artists' techniques. Each of their approaches can contribute valuable information 

about the painting techniques of an individual artist, a school, or 

an art historical period or movement. But each approach leaves something 

unexplained, something missing from the whole picture that encompasses 

everything from the artist's brushwork to preference for certain materials and 

formulas. To provide a real understanding of artists' techniques, it is necessary 

to establish a close collaboration between all the above-mentioned disciplines, 

not only to secure a more complete set of data but, more importantly, to 

stimulate interdisciplinary formulation of more holistic answers about artists' 

techniques (Fig. 1). A painting should not be studied by individual specialists 

from each discipline, but rather by representatives of all disciplines who view 

Table 3. Scientific methods used Jo r the analysis oj painting structures. 

Method In onnation Llmltallon 

mfrared refiectography Identillcation and study of I difficult to Idenllly 

underdrawing 

underdrawing beneath thick 

layer of IR-opaque paint 

layer 

X-ray radiography study of support, lead white - superimposition of several 

underpainting, brushwork, layers of painting in one 

and changes in composition X-ray radiograph 

- difficult inteIpretation for 

features done in organic 

materials or pigments of low 

atomic number 

X-ray tomography detailed study of individual - methodology under 

paint layers 

development 

- high cost 

- problem of interpretation 

when individual paint layers 

are of uneven thickness 

cross section analysis detailed material - cross section sample 

(microchemical methods, IR identification (pigments, needed 

microscopy) binding media) and - high cost of analysis 

sequence of individual paint - positive identification of 

and material layers of the some organic pigments 

painting 

might need additional 

sampling, followed by 

organic microchemical 

analysis 

- additional sampling might 

also be needed for detailed 

analysis of binding media 

using gas chromatography - 

mass spectrometry 

10 


the painting together, share background information, and actively collaborate 

in formulating a working hypothesis, work strategy, and research goals. Successful 

research calls fo r broad international and multidisciplinary collaboration 

when art technique findings are used for authentication purposes. 


Abstract 

This paper provides an overview of 

technical aspects of the search for 

verisimilitude in seventeenth-century 

Italian painting. In particular the role 

of varnishing will be examined in 

relation to technical problems caused 

by absorbing grounds. In addition, 

some theories on viewing distance 

and lighting will be discussed. 

Varnish, Grounds, Viewing Distance, and Lighting: 

Some Notes on Seventeenth-Century Italian 

Painting Technique 

Helen Glanville 

au Fourquet 471 20 

Pardaillan 

France 

Introduction 

In the sixteenth century, the dichotomy between Disegno and Colore, between 

Titian and Raphael, was seen as one between those artists who chose to 

imitate nature and those who chose the Antique as their model. In the seventeenth 

century, this dichotomy could be reduced, simplistically speaking, to 

two groups of artists: those who were more strongly influenced by Raphael 

and the Antique and who painted with a smoother, apparently more finished 

technique; and those who were more influenced by Titian and the Venetian 

school of painting and the more open texture that accompanied this type of 

representation of naturalistic effect-what Poussin's friend Du Fresnoy termed 

"the great Lights and Shadows, the Effect of the whole together" (1). These 

two tendencies have been seen to coexist through the end of the nineteenth 

century. 

Although these two schools differed in their approaches to handling paint, 

they both subscribed to the idea that painting should be the representation 

of natural appearances on a flat surface; and most importantly, that through 

this representation the public should be able to grasp a higher and greater 

truth. Ideas as to what form this imitation of nature should take varied, but 

the essential concept can be found in the writings of theorists as divergent 

in other respects as the arch-Venetian Boschini and Bellori, the epitome of 

Roman Classicism (2, 3). 

The first Academy of Painting, founded in 1586 by the Carracci in Bologna, 

was crucial to the development of painting in seventeenth-century Italy. Painters 

such as Domenichino, Reni, Albani, and Guercino (as well as Annibale 

Carracci) who had come to Rome after training at the Carracci Academy 

profoundly influenced their contemporaries in Rome. They brought not only 

the teachings of their masters (i.e., that the painter had to emulate nature 

accurately on a flat surface, while also illustrating the essence, the "Truth," of 

what was depicted, that which was beyond simple appearances). This concept, 

essential to painting in its newly reacquired status as a liberal art, was also in 

complete accordance with the tenets laid down by the Counter-Reformation. 

Two camps emerged concurrently: those who described the thing itself, and 

those who described the impression on the beholder. 

St. Philip Neri and Paleotti both required that artists, through verisimilitude 

or realistic representation, appeal to the hearts and minds of the people. 

This paper provides an overview of technical aspects of the search fo r verisimilitude 

in seventeenth-century Italian painting. In particular the role of 

varnishing will be examined in relation to technical problems caused by absorbing 

grounds. In addition, some theories on viewing distance and lighting 

will be discussed. 

The use of varnish 

Andre Felibien, who moved in the artistic circles of Rome in the 1640s and 

who was a friend of Poussin, Guercino, and Cigoli, and probably knew Galileo, 

wrote, "When a work is painted to the last degree of perfection, it can 

be considered from close to [sic]: it has the advantage of appearing stronger 

and three-dimensional" (4). This same effect, wrote Felibien, can be created 

12 


y distance with "the aid of the air interposed between the eye and the object, 

using different distances," or by the application of a varnish "which is why 

we cover paintings with a varnish that blunts that brightness and sharp edge 

[qui emousse cette pointe brillante et cette vivacite] which at times appears too 

strongly or unevenly in freshly painted works; and this varnish gives them 

more strength and softness [douceur]. ... We use all these different methods 

to give painted objects that relief, that roundness that they require in order 

to resemble what one is imitating" (5). 

Marco Boschini, the Venetian art theorist who in 1660 published his poem 

on the art of painting, "La Carta del Navigar Pitoresco," had a low opinion 

of the glossy varnishes which he says "foreigners" used: "They make such a 

commotion about it, that it would seem that gloss is the only beauty, and 

varnish the apogee of art" (6). Later in the poem, he makes the distinction 

that is central to artistic theory and to seventeenth-century Italian artists in 

Venice as well as Rome. Comparing the aforementioned "foreigners," Boschini 

praises the Venetian painter fo r imitating the effects of gloss (in armour 

and mirrors) in paint, rather than resorting to the use of varnish to produce 

this effect physically on the painting's surface, writing, "Ii ha fatti straluser co'i 

colori" (he has made them gleam with his paints). 

As Boschini wrote, "diligent painting can be copied," but not what he termed 

the colpi di dottrina (which rather defies translation). It is the element of skill 

that is paramount, rather than the materials. 

A high-gloss varnish may have been a requirement of northern seventeenthcentury 

artists, at least according to Boschini, but this paper investigates the 

role played by varnish in Italian paintings of the same period, both those 

requiring "close scrutiny" and those "made to be seen from afar," as Horace 

discussed, distinguishing these two kinds of fundamentally different paintings, 

when pleading for flexibility in the judgment of poetry, saying that it should 

be judged like painting which exhibits not only a detailed style that requires 

close scrutiny, but also a broad impressionistic style that will not please unless 

viewed from a distance (7). 

Interesting information on the use of varnish in seventeenth-century Italian 

painting can be gained from the highly illuminating lecture entitled "II Lustrato" 

delivered on 29 December 1691 by Filippo Baldinucci to his colleagues 

at the Accademia della Crusca in Florence. He dealt with the subject of 

varnish in contemporary painting as opposed to its use in ancient times and 

in the fourteenth century (8): 

VVhat we are told oj the practices oj painters in antiquity leads us to believe 

that they did not use oil as their medium; what we are told oj the practice 

oj Apelles always remains with me, that is, that he fo und a certain dark 

color [9], or maybe a varnish, which no one was able to imitate. This 

varnish he applied to his works after he had finished them, and with such 

a skill that the bright colors did not offe nd the eye and appeared from afa r 

as through a glass (and please take note of this detail). Harsh colors acquired 

through it an element of austerity, oj saturation. This is precisely what our 

painters in the fo urteenth century did, bifore the discovery of oil as a 

medium; that is, they applied a varnish to their panels, which was of a 

composition that gave their languished paintings a certain depth and added 

strength, and by dimming the overpowering highlights, brought the whole 

closer to natural appearances. And then if you hear it told that modern 

painters sometimes also use a varnish on their oil paintings, I reply that 

such practice (which is con f ined to a few painters) is not to make up a 

deficiency in oil painting-that is, to bring depth to the dark colors, and 

render the highlights softer and less garish-things of which oil painting 

has no need, but rather to remedy an accidental mishap. This can occur 

because oj the imprimatura, the paint mixture which one applies to the 

canvases or panels, or because oj a difect of the canvases or panels themselves. 

They attract the oil liquid so strongly, almost stealing it from the 

Glanville 13

paint, that the paint remains drained [prosciugato], so that it can no 

longer be seen evenly on the suiface, as would have been the case when 

such an accident would have been prevented. By means oj another unctuous 

substance, which is the varnish, applied to those areas where the oil is 

missing on the suiface (and this is the crux of the matter), the dark colors 

are made to reappear. These are the dark pigments which are really present 

in the oil painting, not those darks that just appear dark but are not 

physically present, as was in fa ct the iff ect created by Apelles's varnish in 

some very small areas on his paintings. 

Many of Baldinucci's points require and deserve investigation; I will confine 

myself to examining his assertion that varnish in his time was applied as a 

retouching remedy locally rather than over the whole work (as the oil medium 

itself provided the saturation required to "bring the whole closer to 

natural appearances"). I shall also fo llow up his point linking this localized 

"sinking-in" with absorbency of the ground layer. 

We are indebted to the Englishman Richard Symonds, an amateur painter 

who spent 1646-1647 in Rome, fo r the very detailed and painstaking notes 

he took while watching Gian Angelo Canini painting in his studio. Canini 

was a fr iend of Poussin's, and like the latter, trained in Domenichino's studio 

in Rome. 

Symonds describes on several occasions that portions of Canini's paintings 

were "sinking in" (prosciugated), and he also described the remedies that 

Canini applied. Giving an account of the painting of a portrait of Sir Thomas 

Killigrew, he describes how: "the face and field were prosciugated ... nitwithstanding 

oyle was putt upon the back side of the cloth." This he blamed 

on the fact that the "cloth" (and he marked "imprimatura" in the margin) 

"was not as perfectly dry as it ought. I askt him how he would fetch the 

colour of the face, he sayd he would give it a semplice chiara d'uovo [eggwhite] 

beaten together . .. or olio di sasso, another varnish" (10). 

Elsewhere, Symonds describes Robert Spenser's portrait, painted in one sitting 

by Canini (1 1): 

The scaife which was crimson he did with lacca / biacca Qake and lead 

white]. The whole scaife being done, and afo re it was dry he putt on gold 

colour for the fringe, all which kept his fresh colour and needed no varnish. 

Not three days after, when one would thinke it was scarce dry he with a 

pencill of setola [hog] putt on his varnish over the first field & face & 

Armour & hands, but not the scaife or benda, / this kind of varnish he 

esteemed above that of Olio di Sasso. 

For the preferred varnish, he gives the following recipe on f.20: "2 oz. of 

seven times distilled aquavita (spirits of wine) 112 oz. of ground sandarac 

112 oz. of olio d'Abezzo [Strasbourg Turpentine]." This "final" varnish was 

only applied to those areas which he felt required it. He was also quite clear 

that certain pigments should not be varnished. 

Symonds questioned Canini about the sky in a monumental painting that he 

executed on the subject of Anthony and Cleopatra: "2 days after the Azzurro 

ayre was dry, and I askt him what if it should prosciugare [sink in], so much 

the better said he, because you never put vernish over azzurro" (12). 

The reason for this answer may be in the yellowing characteristics of the 

varnish or in a deliberate choice fo r the matte quality of the blue with a 

consequent increase in scattered reflection, and of the white component in 

light. This effect, to be avoided in darks, but desirable for achieving the effects 

of aerial perspective, was advised by Leonardo in his treatise on painting. One 

should not forget that all the artists in Poussin's circle had a strong interest in 

optical matters, and studied mathematics and optics with the best mathematicians; 

in addition, Galileo moved in this same circle. 

Canini was also aware that certain pigments, such as blacks, caused problems 

by drying matte: "Black, even varnished, sinks in," meaning that unless cor- 

14 


ected, shadows would appear matte and therefore cool and come forward 

rather than giving the intended effects of depth (13). 

Armenini is also aware of this problem. In his De' veri precetti della pittura 

(1587), speaking of shadows and contours he says: "They never remain sufficiently 

finished ... but will always sink in, or the darks become cruder so 

that it is necessary to come back to them many times in order to make them 

fresh, blended, soft, and pleasing" (14). 

Absorbency of the grounds 

The problem of overabsorbent grounds, which is identified by Baldinucci as 

the cause for localized applications of varnish, is also a problem that reoccurs 

in Symonds's notes and obviously concerned Canini and presumably the other 

artists in this circle. 

Almost all of the samples taken from paintings produced in central Italy circa 

1600-1800 showed the presence of chalk in varying but significant proportions. 

Sometimes chalk was found as a single ground layer, as in the case of 

Domenichino's Vision of St. Jerome of 1603 (National Gallery, London); in 

Annibale Carracci's Boy Drinking (Christ Church Gallery, Oxford) which 

probably dates from the end of the sixteenth century; and in the much later 

Cavallino, Christ Driving the Traders from the Temple (National Gallery, London). 

This type of ground derives from Venetian painting and is found, for instance, 

in Titian's Madonna and Child (ca. 1525), in which what appears to be the 

dark ground is in fact the saturated color of the canvas. The refractive index 

of chalk is so low that it is almost completely transparent when associated 

with oil. 

Not enough analyses have been carried out to ascertain whether the chalk 

was applied in an aqueous or oil medium. More often the chalk is found 

mixed with earth pigments and small inclusions of other pigments such as 

charcoal black, or with driers such as lead white or an umber, as is the case 

with Caravaggio's works in the National Gallery in London (15). In a thin 

section taken from Poussin's Finding of Moses (National Gallery, London), one 

can see that the absorbency of the ground was recognized and an isolating 

oil layer applied. The sample was taken from a blanched area, in which both 

the ground and the paint film were very lean. Interestingly enough, in other 

areas containing lead white, the paint film was less lean and less absorbent. 

Symonds wrote, "It is good to give a mano [hand] of gesso first, and then oil 

on top." Earth from which bricks are made is ground and used for the imprimatura, 

Symonds wrote, recording the recipe for the ground favored by 

Canini: After sizing the canvas, he applied a mixture of "red earth, a little 

white lead, a little charcoal black, and chalk" (16). Until now the latter seemed 

always to have been interpreted as white clay, although brick clay seems perfectly 

adequate. This mixture was applied to the canvas in an oil medium. 

The absorbency of the ground seems to have been both feared (because of 

the problem of sinking in) and desired (because it kept the colors fresh) . As 

Symonds wrote, "These cloths that have gesso in their imprimatura-the 

gesso makes the colour keep fresher and does drink up the evil of the oil, 

but they crack sooner, and that is the worse of the gesso" (17). The sinkingin 

of colors, Symonds's "prosciugare," seems to be linked in their minds to 

the poor drying of the ground, either because the commercial primer left 

out the lead white for economy's sake (18), which seems to reduce the porosity 

of the paint film, or because it had not dried sufficiently before the 

painting was executed. 

Then, as today, the poor drying qualities of earth pigments were recognized 

but not fully understood. From analyses we know that when the pigment in 

the oil film is an earth color, there is a surprisingly high percentage of scission 

Glanville 15

products in the dried oil film (19). These short-chain scission products, by 

giving the structure a swollen matrix, confer flexibility to the oil film. Earth 

pigments are then particularly suitable as grounds on flexible supports. To 

speed the drying process of these grounds, artists advocated the use of prepolymerized 

oil (20). 

Most of the inorganic components of these colored grounds are hygroscopic 

and are not easily wetted by organic media. Free fa tty acids in the oil would 

act as surface active agents, and therefore an unpolymerized oil would be 

preferable to ensure wetting of the particles and pigment aggregates. 

Groen and Burnstock have fo und air pockets with earth pigments bound in 

unpolymerized linseed oil (21, 22). This problem can only be aggravated by 

the use of an oil with a reduced proportion of free fatty acids (23). The 

presence of air pockets makes the ground porous, causing the paint layers 

above to "sink in" as the paint medium is drawn down into the absorbent 

ground. 

Speedy drying achieved by applying the ground as an emulsion, as advocated 

in Pierre Lebrun's Treatise, would also result in a porous film because of the 

voids formed during evaporation of the aqueous phase (24). The absorbency 

of the ground would be particularly problematic with artists such as Guercino 

who, in his first phase, painted aUa prima, using the dark ground fo r his 

shadows and often applying only one layer of paint. Malvasia, describing 

Guercino's technique before the artist left fo r Rome, says that Guercino had 

"an extraordinary speed of execution, in one go laying-in (bozzando) and 

finishing" (25). In the case of Guercino, the complaints came from his clients 

who felt that since they paid Guercino by the figure, they wanted to see the 

whole figure, not one in which more than half was drowned in shadow. 

Lanfranco and later Giordano were also known for their speed of execution. 

This speed was castigated by most art theorists, because it was seen as pratica, 

simple manual facility rather than the fruit of matured intellectual thought 

and skill. 

Exposed or thinly covered porous grounds would pose a problem when it 

came to varnishing (for the restorer as well as the artist), but grounds suffering 

from what the French call lithargeage present a similar problem in appearance. 

The coarsely ground lead white would present a less absorbent ground, but 

the rough granular surface scatters light, obscuring detail in the worst cases 

and flattening the composition. Many seventeenth-century paintings suffer 

from this problem (26). 

Viewing distance and lighting 

It is clear that in a painting such as Guido Reni's Abduction of Helen (Louvre), 

which does not depend on the warmth of its shadows for the depiction of 

depth and its illusion of space but rather on what Leonardo termed aerial 

perspective and the diminution of colors, the impact oflithargeage is minimal, 

and saturation of the picture surface not essential. 

This is the case with paintings of what I have termed the "high finish" school; 

that is, those by painters who did not paint alIa prima, nor availed themselves 

in the Venetian manner of the artifice of large areas of shadow and light, but 

rather achieved the illusion of space through "the diminution of the hues, as 

much through their quality as through their strength" (27). 

The relative importance of saturation in paintings that rely on their dark 

ground for illusion of space was especially visible in a recent exhibition of 

Neapolitan painting in Bordeaux. Two paintings by Caravaggio were hung 

next to each other. One was a privately owned Doubting Thomas, with a high 

gloss varnish, and the other painting was a Salome with the Head of St. John 

the Baptist (National Gallery, London), which had a much more matte appearance 

(28). 

16 


The differences in saturation, clearness of detail, and illusion of space were 

very clear at a normal viewing distance. If one stepped back far enough to 

be able to see both paintings, the detail and saturation of the two works 

appeared to be the same, and both surfaces had a matte appearance. The very 

effect of the air interposed between the eye and the object was alluded to by 

Felibien. According to this phenomenon, first objectively described and explained 

by Leonardo, objects in the distance will appear more blue because 

of the interposition of the tiny particles contained in the air. 

The warming, darkening effect of varnish, that of light transmitted through 

the varnish, will be counteracted by the blue quality of the reflected light. 

This means that with paintings executed to be seen at a distance, a glossy 

saturated varnish would be unnecessary; at the correct viewing distance, the 

painting will appear somewhat matte. 

Boschini, characteristically, writes, "Do you think painting is enjoyed more 

directly under the eye or from an appropriate distance? Certainly at a distance; 

because once you have hung the pictures and placed them at the wanted 

height, there's no reason for you to go clambering up there" (29). 

Another variable that is too infrequently considered when exhibiting paintings 

is the lighting. Paintings from that period were never painted in uniform 

daylight, nor were they to be seen in such conditions. Daylight would have 

been directional, and artificial light would have been warm candlelight. Restorers 

sometimes have the opportunity oflighting the painting with tungsten 

lights from the direction in which the painting was executed, and the effect 

in increased naturalism and illusion of space can be quite dramatic (30). 

This is not a new observation. In 1628, in a text on the artist Cigoli, the 

biographer Commodi wrote about Cigoli's Martyrdom of St. Lawrence: "It 

appeared much more admirable in the room in which it was painted which, 

receiving little light through a small window, was indeed in proportion both 

with the handling and with what was represented in the scene. Having seen 

the painting in this room several times, and seeing it again in the open air, 

he said to himself 'This just is not the same, even though everyone praises it 

also out here' " (31). 

Viewing distance is also crucial for works executed with a painting method 

that defies legibility at close scrutiny. This is what we now may revere as the 

"magic" of painting, but to Poussin and his circle it came close to pratica, 

manual skill not guided by the intellect and therefore "inferior" and not 

worthy of painting as a liberal art. 

Rather surprisingly, Bellori appreciated the skill involved in producing effects 

to be appreciated from a distance. In his biography ofLuca Giordano, speaking 

of a sotto-in-su figure of Christ, he describes how Giordano would constantly 

descend from the scaffolding to check on the effect of his painting and the 

"huge brush strokes" and "coarse and garish highlights in the blue drapery," 

which the distance of the eye blends and harmonizes together so beautifully 

and makes one understand how great his mastery is" (32). 

Boschini's poem, a eulogy of Venetian painting and technique, extols the 

macchia of which the "high finishers" are so scornful. He carefully distinguishes 

this from the rapid, sketchy alla prima technique he terms prontezza (33). He, 

too, emphasizes the skill involved, rather than the material aspect. The Venetian 

macchia "is what unites the wet paint with the dry paint beneath, 

which is so important to the beauty of this way of painting" (34). 

Like Armenini and Felibien, he emphasizes the handling of the paint and the 

color, rather than its nature and strongly argues that painting should appeal 

to the mind and not the eye-that the spectator must see the hand inside 

the glove, while the eye only sees the glove (35, 36). 

This brings us back to the essential message of the Counter-Reformation 

and of the new stature of painting as a liberal art: to convey essential truths 

Glanville 17

epresentationally so as to appeal to the heart and mind as well as to the 

senses. As Horace said, "Painting should instruct as well as delight by the two 

ways open to it: close scrutiny or distant viewing." 

Notes 

1. Du Fresnoy, C. 1750. Observations on the Art of Painting. Dryden translation, first 

published in 1667. 

2. Boschini, M. 1660. La Carta del Navegar Pittoresco. Venice. 

3. Bellori, G. P. 1672. Le vite de' Pittori, Scultori e Architetti Moderni, first published 

in 1672. Pis a 1821 edition. 

4. Felibien, A. 1685. Entretiens sur les vies et les ouvrages des plus excellents peintres 

anciens et modernes, 2nd ed. Paris. Vo!' 2, Entretien VII:240. 

5. Ibid., 242. 

6. This passage in a different translation is quoted by Robert Ruuhrs in his article 

"Matte or glossy? Varnish for oil paintings in the seventeenth century," which 

appeared in an English translation in the ICA Newsletter, Oberlin, Ohio: November 

1985. The article was originally published in Maltechnik-Restauro in July 1983: 

"E i forestieri certo in quela parte / Fa tanto cavedal, che la lustrezza / A lori 

ghe par l'urUca belezza. / Ghe par che la sigila tuta l'arte." 

7. Horace. Ars Poetica. 361. See Lee, R. W 1940. Ut pictura poesis: the humanistic 

theory of painting. The Art Bulletin (XXII). 

8. Baldinucci, F. 1692. II Lustrato. Paper presented at the Accademia della Crusca 

29 December 1691 and 5 January 1692. Florence. On page 25: "Trovansi bene 

notizie di Pittori antichi, che fanno credere, che tale uso non vi fosse; fra Ie quali 

potra sempre appresso di me cio che fu scritto d' Apelle, cioe, ch' egli fu ritrovatore 

d'un certo color bruno, 0 vernice che si fosse, la quale rUuno seppe imitare, e 

davala all'opere dopo averle finite; e che servivane con tanto giudizio, che i colori 

accesi la vista non offendevano, facendoli vedere da lungi come per un vetro (e 

notate questa particolarita) e che Ie tinte lascive, mediante quella acquistavano 

un certo che d'austero, 0 di scuro, che e tutto quello appunto, che facevano i 

nostri pittori del 1300, avanti il ritrovamento della tempera coll'olio, cioe, che 

davano sopra Ie tavole una vernice, che era una certa mestura, che alia loro 

dilavata pittura un certo che di pili profondo, e di forza maggiore aggiungeva, 

ed il soverchio chiaro alquanto smorzando, riduceva a maggior somiglianza del 

naturale ...." On page 31: "Se poi sara detto, che i moderni Pittori usano anch'essi 

talvolta vernice sopra Ie lor pitture a olio, io rispondo, che tale usanzach'e 

di pochi-non e per supplire al mancamento della pittura a olio, cioe, per 

render pili profondi gli scuri, e i chiari pili mortificati, e pili carnosi, cose tutte 

delle quali la pittura a olio non ha bisohno rna bensi per rimediare ad un'accidental 

disgrazia, che occorre taJora a cagione dell'imprimatura, mestica, 0 altro 

che dassi sopra Ie tele, 0 tavole, 0 pure proviene dalle medesime tele, 0 tavole, 

cioe d'atrarre cosi forte il liquido dell'olio, quasi rubandolo al colore, ch'e' venga 

in qualche luogo prosciugato per modo, ch'e' non possa farsi vedere in superficie 

per tutto ugualmente, com'egli avrebbe fa tto col cessare di tale accidente; con 

che per mezzo d'un'altra cosa untuosa, che e la vernice data dove I'olio in 

superficie manco, fassi apparire (e questo e il punto stretto e forte) con che fassi 

apparire 10 scuro, che gia nella pittura fatta a olio veramente e, non quello che 

non v'e, ch'era appunto l'effetto, che in qualche piccolissima parte faceva aile 

sue pitture la Vernice d'Apelle." 

9. In Italian, the word used is bruno. It should be noted that imbrunire is the word 

used for burnishing gold, that is, making it darker and increasing its sheen. 

10. Beal, M. 1984. A Study of Richard Symonds: His Italian Notebooks and Their Relevance 

to Seventeenth-Century Painting Techniques. New York: Garland Publishing, 

fo!' 46. 

11. Ibid., fo!' 143. 

12. Ibid., fo!' 75. 

13. Ibid., fo!' 25. 

14. Armenini, G. B. 1587. De veri precetti de la pittura. Ravenna. 

15. A good example can be found in Caravaggio's Boy with Lizard. National Gallery, 

London. 

16. Symonds in Beal, op. cit., fo!' 10. "Terra Rossa, biacca da corpo a little. Creta 

un tantino, negro Carbone. He complayned the ordinary imprimers putt in no 

biacca." 

17. Ibid. 

18. Ibid. 

18 


19. White, R. Personal communication. National Gallery, London. I am indebted to 

Raymond White for much enlightening discussion around the subject. 

20. Symonds, op.cit., fol. 10. "To make imprimatura wich shall stand abroad in ye 

ayre, tis best to use olio Cotto. Also tis good dare una mano di gesso prima / 

poi oglio suopra." 

21. Groen, K. 1988. Scanning electron microscopy as an aid in the study of blanching. 

Hamilton Kerr Institute Bulletin 1 :48-65. 

22. Burnstock, A. An examination of blanching in the green paints from a selection 

of seventeenth-century French School paintings. Ph.D. diss. Courtauld Institute 

of Art, London. 

23. Groen, op. cit. 

24. Lebrun, P. 1635. Recueuil des essaies des merveilles de la peinture. In Merrifield, 

M. p. 1849. Original Treatises on the Arts of Painting. Vol. II. London:John Murray, 

759-841. On page 821: "To prime a canvas quickly, so that a person may paint 

on it the same day that it has been primed, you must grind together some 

parchment glue and oil priming, and immediately prime the canvas with this; it 

will harden directly, but the priming is apt to scale off when the canvas is rolled 

up." 

25. Malvasia. 1646. Vite di Pittori Bolognesi. Unpublished addenda. In Sce/te di curiosita 

letterarie, inedite 0 rare dal XIII al XIX, ed. Adriana Arfelli. Speaking of 

Guercino's Circumcision in the Chiesa delle Monache di Gesu Maria in Bologna 

which was executed overnight: "La sua formidabil velocita nell'operare, e alla 

prima bozzando, e fmendo nello stesso tempo." 

26. For instance, Caravaggio's Salome with the Head of St. John the Baptist (National 

Gallery London) and Mattia Pretis's The Beheading of St. John the Baptist (National 

Gallery of Ireland, Dublin). 

27. Felibien, op. cit., Entretien VIII, 401. "[L]a diminution des teintes ... autant par 

la qualite que par la force des couleurs." 

28. Either an optical effect due to the pronounced lithargeage of the ground or the 

inherent matte appearance of the varnish itself. 

29. Boschini, op. cit., 295. "Credela che se goda la Pitura / Piu soto l'ochio, 0 in 

debita distanza? / Certo in distanza debita: perche / Quando have messi i quadri, 

e situdi / In l'altezza de iluoghi destinadi, / Non e rason, che la ve rampeghe." 

30. This was the case during the examination of Gainsborough's The Linley Sisters 

in the Dulwich Picture Gallery. Gainsborough had painted the portrait with the 

light coming from the left through holes he had drilled in his shutters! See 

Glanville, H. 1988. Gainsborough: Artist or Artisan? (original unedited version). 

In A Nest of Nightingales, s.l. exhibition catalogue. 

31. Biography of Ludovico Cardi, in Cardi, G. B. 1628. II Cigoli. Florence, 19. 

"[M]olto piu ammirabil si rendeva in quella stanza dove fu lavorata, la qual per 

una picciol finestra ricevendo poco lume era appunto proporzionato et alla maniera 

et a quello che nella storia si rappresentava; et avendola ivi veduta piu volte, 

e, rivedendola poi all' aria aperta, diceva fra se: questa non e quella, sebbene quivi 

ancora lodata da tutti." 

32. Bellori, op. cit. Vita di Luca Giordano, 51. "[E]ssendo il tutto condotto con grossi 

trattizzi ne' contorni, e cosi anche il Cristo, benche sia alquanto piu terminato, 

come figura principale del soggetto, ne! cui panno azzurro si veggono chiari 

terribili di biacca imbrattata di carnatura, con tratti di pennello piu grossi, il quale 

unisce ed accorda tanto bene con la distanza dell'occhio, che fa comprendere 

quanto sia grande la maestria de! suo pennello." 

33. Boschini, op. cit., 339. "[L]a prontezza ... impastar figure in copia, e senza natural. 

..." 

34. Boschini, op. cit., 339. ''L'e que! unir quel fresco con quel seco / Che in sto 

operar si belo importa tanto." 

35. Boschini, op. cit., 294. "Perche se se vardasse al color solo, / EI piu vago sarave 

anche il piu belo / Ma varvandose el trato del pene!o / Casca sta consequentia 

a rompicolo." 

36. Boschini, op. cit., 290. "[L]'ochio no'vede altro, che'l guanto." 

Glanville 19

Abstract 

French publications on painting materials 

and techniques before 1800 

include books of "secrets," treatises 

written by artists themselves or 

about their practice, articles from 

learned journals, dictionaries, and 

encyclopedias. The traditional methods 

of painting listed in these publications 

include dhrempe (a reference 

to a water-based media, which can 

be glue, gum, or egg tempera), fresco, 

miniature, and oil painting, 

enamel and glass painting, as well as 

painting on porcelain. Some of the 

more unusual techniques are outlined: 

encaustic, eludoric, and glass 

painting, including glass transfer 

techniques, and finally a satirical 

contribution, "peinture en fromage 

ou en ramequm." 

From. Books of Secrets to Encyclopedias: Painting 

Te chniques in France between 1600 and 1800 

Ann Massing 

Hamilton Kerr Institute 

Whittlesford 

Cambridge CB2 4NE 

United Kingdom 

Introduction 

Until the mid-seventeenth century, the business of painting in France was 

strictly regulated by guilds. The skills of the painting craft, handed down from 

master to apprentice, were treated as valuable personal possessions and protected 

by secrecy (1). Few records of the painting techniques themselves were 

published before the mid-eighteenth century. The first information contributing 

to the history of painting materials and techniques in the French language 

is found in books of "secrets," collections of recipes on various topics, 

often compiled from diverse authors. The "secrets" of the earliest such books 

were not contemporary recipes, however, but translations from Italian or Latin, 

often with reference to ancient Greek or Roman authors. 

Books of secrets 

The first of the books of "secrets," Les secrets de reverend Alexis Piemontois (Paris, 

1557) was a compilation, translated from the Italian, that included remedies 

for illness, wounds, and accidents, as well as instructions on how to make 

perfumes,jams, dyes, and pigments. The book was reprinted in 1573 and was 

expanded to include recipes from Dioscorides, Galen, and others, as well as 

some recipes relating to the history of painting techniques such as gilding, 

pigments, inks, and the making of varnishes. It was republished several times 

during the sixteenth and seventeenth centuries, as were other secret books, 

such as Les secrets et merveilles de nature (Lyon, 1586) by Jean Jacques Wecker, 

a doctor from Colmar. Wecker's recipes fo r magic and sorcery, health, cooking, 

beauty, and the manufacture of pigments, "des secrets des falseurs et vendeurs 

de couleurs et peintures," were also taken from Latin sources; he credits seventy-five 

authors from Aristotle onward. Fortunately, the book includes a 

table of contents, fo r the recipes are not arranged alphabetically but in a 

systematic order that reflects Wecker's vision of the universe and places painter's 

secrets on the same level as recipes fo r cooks, druggists, carpenters, and 

Jomers. 

Several books of secrets were published in the seventeenth century, but perhaps 

the most popular and the most relevant to our topic was Le Sieur 

D'Emery's Recueil des curiositez rares et nouvelles des plus admirables iff ets de la 

nature ret de l'art] (Paris, 1674) (2). Included were many recipes related to the 

history of painting techniques, recipes on how to copy drawings, make pastels, 

imitate marble, stain wood, gild, and make engravings look like old master 

paintings. 

Books of secrets continued to be written and reprinted throughout the eighteenth 

century and beyond. In the eighteenth century, the two compilations 

most frequently reissued were Secrets des arts et metiers (Brussels, 1755) and 

L' Albert moderne ou nouveau secrets eprouves et licites (Paris, 1768). This "modern" 

Albert was a revision of "old" Albertus Magnus, with the superstitions and 

enchantments deleted, keeping only "useful" advice, such as how to cure a 

toothache with two live moles. (The method begins by holding a mole in 

each hand and squeezing gently, without letting go, until they die; this process 

should take about five hours.) Large sections are devoted to medical and 

agricultural recipes; the third section includes "divers moyens de se faire une 

20 


occupation agreable pour charmer l'ennui," with recipes about liqueurs, flowers, 

painting secrets, and so on. 

Technical treatises 

The most widely read of the early painting manuals, judging by the number 

of editions, was Claude Boutet's treatise on miniature painting, published in 

1672, called by the author an "ABC de La mignature" (3). Following the conventions 

of his time, Boutet dedicated his treatise not to the specialist but to 

an amateur, a Mlle. Fouquet, who wished to perfect her abilities in the craft. 

In the introduction, Boutet claims his "secrets" came from the Italians, and 

states that although he could have profited by keeping them to himself, he is 

giving away his knowledge fo r the benefit of others. He provides an extremely 

useful summary of French painting practice in the second half of the seventeenth 

century. The treatise discusses the various pigments and how to use 

them, based on a three-step sequence (as in oil painting of the period) : ebaucher, 

pointiller, and finally finir or rehauser. 

A seventeenth-century treatise of great importance for the history of technique 

is De La Fontaine's Academie de La peinture (Paris 1679), dedicated to 

the tutor of the future king of France (4) . He summarizes much practical 

detail, especially on pigment mixtures, mentioning dhrempc, fresco, pastel, grisaille, 

and miniature as well as oil painting. He refers to the art of perspective, 

which he believes a painter should know both theoretically and practically. 

His dependence on Italian painting is evident, and he lists the names of the 

Antique painters who have contributed to this knowledge (5). He also discusses 

the origin of painting; although several "philosophers" attributed the 

origin of painting to a shepherd who began to trace his shadow and saw the 

resemblance to the human fo rm, De La Fontaine allows that the more common 

opinion is that the discovery of painting was made by the Hebrews and 

that they transmitted it to the Greeks and to the Romans. Until the late 

eighteenth century, it was believed that this knowledge was then "lost," to be 

"rediscovered" by "Jean de Bruges" Oan van Eyck) . 

Seventeenth-century painters thus considered oil painting a "new" method 

of painting, and, judging from the source books, the preparation of the medium 

was obviously a vexing problem. De La Fontaine suggests nut oil with 

lead white thickened by sunlight (6). He follows this with the recommendation 

that in order to dry a layer oflead white, or a grisaille layer, one should 

mix the white pigment with oil of turpentine, which will cause the layer to 

dry as the turpentine evaporates (7). This is followed by advice on how to 

prevent layers of drapery from running. He suggests placing the painting flat 

on the floor or on a table and scattering small pieces of absorbent paper onto 

the surface, especially onto the shadows of the drapery. When the paper has 

drawn out the oil, but before the paint layer has dried, one should pick up 

the painting and let it fall gently on one corner, so that the paper comes away 

(8) . 

Although it is not possible to go into detail about early technical practice 

here, I will mention briefly De La Fontaine's method of preparing supports 

for painting, methods used well into the eighteenth century. In the case of 

canvas, the material was stretched onto a loom, smoothed with a pumice 

stone, sized with glue, and then given a double ground, the first colored with 

raw umber and red brown spread on with a knife, the second with lead white 

and just enough carbon black to make gray. For panels, three layers of a fine 

chalk ground ("blanc d'Espagne, comme on Ie vend chez les Chandeliers") 

was recommended, with either another layer of glue on top or a gray oil 

layer (9). 

The Acadernie Royale and the theoretical treatises 

Before 1800 the history of oil painting in France, and consequently the history 

of its technique, was dominated by Italian influence. In the sixteenth 

Massing 21

century, the royal patronage for the School of Fontainebleau was crucial. Later, 

most ambitious French painters traveled to Italy, often spending much of their 

professional lives there. The influence of the French Academy, the Academie 

Royale de Peinture et de Sculpture founded in 1648, was also of central 

importance for any aspect of the history of painting in France. By 1655 it 

had become a royal enterprise and was soon the most powerful art institution 

in Europe. The purpose of the Academy was to convey the principles of art 

to its members by means of lectures and to instruct students through life 

classes. In 1673 the Academy also began organizing exhibitions fo r its members. 

These exhibitions were not opened to a wider public until 1791 (10). 

With the noble aim of nurturing good craftsmanship, the Academy defined 

the principles from which painters were not allowed to deviate. The goal was 

to train students in one particular style of drawing. The necessity of copying 

from the ancients was stressed, and this emphasis continued until the French 

Revolution. Nowhere outside the Academy was a life class allowed, even in 

an artist's private studio. The Academy did not provide fo r the whole of the 

professional education of a young painter, however; the student painter still 

learned the basics of his craft in the workshop of his master, in whose house 

he often lived, just as in the Middle Ages. Thus, craft "secrets" were still 

essential to a painter's success, and only gradually were detailed accounts of 

methods and materials forced into print. The publication of Diderot's encyclopedia 

in the mid-eighteenth century and the French Revolution a few 

years later provide evidence of the culmination of a trend toward the dissemination 

of knowledge which had proved impossible to stem. 

The personalities of Jean-Baptiste Colbert (1619-1683), the elected "Protector" 

of the Academy, and Charles LeBrun (1619-1690), "Premier Peintre du 

Roi," were crucial in the development of the Academy and thus to the history 

of French painting. Since it was a royal academy, the king's (i.e., Colbert's) 

intentions were imposed, and Colbert's tight dictatorship led to a centralization 

from which France has still not fully emerged. The approval of the 

Royal Academy was a necessity fo r every endeavor within its sphere of influence. 

Colbert was also responsible fo r the creation of many other academies 

in addition to the Academie de la Peinture, including the Academie des 

Sciences in 1666; the importance of this fact to the subject will be indicated 

later. 

The lectures of the Royal Academy were the basis fo r many of the published 

treatises of the seventeenth and even the eighteenth centuries. Aesthetics, 

beauty and proportion, lighting and perspective, and the expressions of the 

figures portrayed were discussed at length by seventeenth-century authors. 

Charles LeBrun's famous lectures on human expression (1668) influenced 

generations of artists. Most of the academic authors, such as Roland Freart 

de Chambray, Andre Felibien, Charles Alphonse Du Fresnoy, and Roger De 

Piles, were not professional painters, however, and were more concerned with 

theory than practice. The dispute between the Ancients (Poussinistes) and the 

Moderns (Rubenistes) dominated the Academy lectures for more than twenty 

years. The study of the lives of ancient Greek and Roman as well as Renaissance 

painters was also a popular subject, as was the history of the origin 

and "rediscovery" of painting. Although the academic lectures did not usually 

discuss techniques, Jean Baptiste Oudry's (1686-1 755) lectures were an important 

exception. As president of the Academy, Oudry gave a lecture in 1752 

on painting techniques as practiced by the members (1 1). 

The spokesman of the Poussinistes, Roger De Piles (1635-1709), an amateur 

painter, connoisseur, and member of the Academy, was a prolific writer on 

the theory of painting. He did, however, compose one book on technique in 

1684, Les premiers Clemens de la peinture practique. As edited and augmented by 

Charles Antoine Jombert in 1776, the text became the most important and 

informative collection of recipes on French painting technique of the seventeenth 

and early eighteenth centuries (Figs. 1,2). Another important author 

fo r the history of technique was Philippe de La Hire of the Academie Royale 

22 

Histon·cal Painting Techniques, Materials, and Studio Practice

DE 

PEJN'FURE PRATIQUE. 

Par 

demit:: 

M. DE. 

Royale 

PILES, 

de Pcinture 

de l'Aca 

Sculpture. 

& 

NOUVELLE EDITION 

Entierement refonduc. & augmende con· 

lidefablement. 

Par CHARL!S-ANTOna J olinItY; , 

-I$ 

A AMSTERDAM lIT 

.. 

did sink, however, he suggested applying with a small sponge a layer of nut 

oil mixed with a siccative before continuing to paint. He added that although 

this layer should dry before adding fu rther details, it was possible to paint fine 

lines of architectural elements while the layer is still wet (12). 

Learned journals and periodicals 

l'! ENCYCLOP BDIE, 

ou 

DICTIONNAIRE RAISONNE 

DES SCIENCES, 

PAR. UNE SOCIETE DE GENS DE LETTRES. 

DES ARTS ET DES METIERS. 

Nio"'onIr

In spite of the laments of certain people who complain if the taste of the 

century fo r dictionaries, this predilection is increasing, a proof if the benifrt 

the public is receiving. We are inundated, and if this torrent is not stopped, 

we will be learning only from dictionaries . ... We want to know everything-or 

rather speak if every thing and pretend to be ignorant if nothing. 

We must submit, therifore, to this "gout du siecle. JJ 

Encaustic painting 

The history of oil painting was an eighteenth-century fascination. In the mideighteenth 

century Diderot still believed, as De La Fontaine previously asserted 

in his seventeenth-century treatise, that the art of painting was reborn 

when "Jean de Bruges" Gan van Eyck) discovered the "secret" of oil painting. 

It was generally accepted that the Ancient painters possessed superior techniques, 

and that it should be possible to "rediscover" their methods. By the 

eighteenth century there was also concern about the state of seventeenthcentury 

paintings, which no longer retained the freshness of those more recently 

painted (23). 

MEMOIRE 

U. PIINTURE A L'ENCAUSTIQUE 

SU k U. PiINTUl.t A LA CIIE.o 

,.. JI. C-::.:..:::.. *r"' ... 

f>JI._.nr.e-"Io.F_.M6iI

ived from the Greek words fo r oil and water, the two main components of 

the method. The method follows (27): 

The support, usually of fine cloth primed with poppy oil, is placed under 

a thin film

While cooking their eggs, two young Spartans quarreled over a piece if 

cheese which had just disappeared in the frying pan. Both young men armed 

themselves with a spoon and searched the scrambled eggs, but in vain. The 

cheese was gone. They came to blows, one claiming the cheese was still 

there, the other that it had disappeared, carried off by a genie. The noise 

if this battle drew a crowd, and the cause if the dispute became known. 

It was suggested that the experience be repeated. The ability if egg yolk 

to dissolve cheese was recognized. 

The author consoled himself that he was not the first to have discovered this 

secret, and so made it public, congratulating himself on his generosity. Others 

might have kept the critical ingredient secret in an attempt to profit from it. 

Selling his secret would have been understandable, after all, as reimbursement 

fo r the heavy expenses in eggs and cheese. Rouquet continues his description 

with all the seriousness of a technical treatise: 

Mix gruyere cheese cut in f ine strips with two beaten egg yolks over a bainmarie 

until the cheese has melted. For the question of a support fo r the 

paintinsince cheese does not adhere well to panel or canvas, it is better 

to follow its natural association with bread. Therifore, take flour and make 

dough with a little milk. Finally, it would be bemificial to add a bitter 

substance to discourage worms, mice, and children from eating the paintings. 

However, by leaving it out, poor painters could at least dine on their own 

paintings. 

Conclusion 

As we approach the twenty-first century, our curiosity increases about how 

paintings of past generations were created. More and more, we look toward 

modern methods of scientific analysis to answer our questions about historical 

painting materials, but a return to the written sources on painting techniques 

is an important first step toward a proper understanding. By definition, painters 

are practically oriented, however, and have rarely composed with the pen 

as well as the brush; records of their materials and techniques must be plucked 

from various publications. Undoubtedly during the coming decades, historians 

of painting technique will be recovering more information from the 

source books about the history of studio practice, painting materials, and 

techniques. 

Notes 

1. Patent rights existed in Europe in the mid-eighteenth century, but enforcement 

came only much later; in order to benefit from a discovery, the inventor still 

relied on secrecy. Singer gives an example involving an improvement to the 

system of production of liquid bleach made in 1789 fo r which the patent owner 

enjoyed the protection of his patent for only four years. Singer, c., et al. 1958. 

A history oj technology (4): Oxford, 247. 

2. A book of secrets, Essay des merveilles de nature et des plus nobles entifices (Rouen, 

1622), proved so popular that by 1657, it was already in the thirteenth edition. 

The author was Etienne Binet, the pseudonym of Rene Franyois, Predicateur 

du Roy. 

3. Boutet, C. 1672. Traite de la mignature, pour apprender aisement a peindre sans martre, 

et Ie secret de Ja ire les plus belles couleurs, I'or bruny, et I'or en coquille, Paris. Other 

editions were published with slight variations in the title and contents. Some 

editions appeared anonymously as Escole de la mignature. At least twenty-five 

editions were published between 1674 and 1800. 

4. De La Fontaine. 1679. L'Academie de la peinture [etc.], Paris. De La Fontaine 

dedicates his treatise to "Mesire Charles de Sainte Maure, . .. Gouverneur de 

Monseigneur Ie Dauphin," for the education of the future king. Two other important 

treatises of the seventeenth century are now available in facsimile editions. 

These are Bernard Dupuy du Grez's 1699 Traite sur la peinture pour en apprendre 

la theorie, & se peifectionner dans la pratique, To ulouse: J & A. Pech; and Le Blond 

de la Tour's 1669 Lettre a un de ses amis, contenant quelques instructions touchant la 

peinture. Bourges et Bordeaux. By the eighteenth century, the number of treatises 

Massing 27

increased considerably; the reader is best referred to Massing's 1990 bibliography 

(available from the author) . 

5. De La Fontaine, op. cit., epistre aiiii: "Je pretend les instruire Des jeunes] au 

melange des couleurs, a la force des teintes, & a la union des jours & des ombres 

que les Romains nomment clair-obscur." 

6. De La Fontaine, op. cit., 15: "Autre huille grasse pour faire seicher les couleurs. 

Vous prendres de I'huile de noix dedans une fiolle, vous y mettres de la mine 

de plomb, & du blanc de plomb pille ensemble, la mettres au soleil, cela s'engraisse, 

& sera fo rt clair: on en mesle avec des couleurs, quand on est presse pour 

les faire." 

7. De La Fontaine, op. cit. "Beau secret pour faire seicher Ie blanc de plomb sans 

changer. Quand vous voudres faire seicher du blanc de plomb, couche entiere, 

ou bien avec du gris, en grisaille, vous prendres de l'huile de terebentine meslee 

avec vostre blanc, il Ie fera seicher: notes que I'huile de terebentine fait tout 

seicher, parcequ'elles'en va en fu mee. " [Author's emphasis.] 

8. De La Fontaine, op. cit., 17. 

9. De La Fontaine, op. cit., 27-28. For the preparation of a copper plate: ''Lon 

prend une plantaine de cuivre bien poly, apres vous prendrez du blanc de plomb 

bien broye avec de la terre d'ombre, & noir de charbon mesle ensemble, avec 

une brosse vous frotterez par dessus la plataine bien unie, & avec un linge & du 

cotton dedans vous frapperez dessus, pour la rendre mieux unie, quand la couleur 

sera seiche vous prendrez un cousteau & passerez par dessus pour unir davantage, 

apres vous la chargerez encore une fois, & ferez encore de mesme; Ie laissant 

seicher, vous pouvez travailler apres pour peindre sur une pierre ou piastre ...." 

10. The history of the French Academy has been written by several authors. See 

Pevsner, N. 1940. Academies of Art Past and Present. Cambridge; and more recently, 

Tradition and Revolution in French Art 1700- 1880. 1993. London: National Gallery. 

11. Oudry. 1753. Discours sur la pratique de la peinture et ses procedes principaux: ebaucher, 

peindre a fo nd, et retoucher. Paper presented to the French Academy, 2 December 

1753, published from the manuscript in the Ecole des Beaux-Arts. In Le Cabinet 

de l'amateur, 1861-1862. 1863. Ed. E. Piot. Paris, 107-17. For fu rther discussion 

of Oudry's lecture, especially as it relates to his use of varnish, see Swicklik, M. 

1993. French painting and the use of varnish, 1750-1900. In Conservation Research. 

Washington, D.c.: National Gallery of Art, 157-74. 

12. La Hire, Philippe de. 1730. Traite de la practique de la peinture. Paris, 710-14. 

13. PJ. Macquer published several small brochures printed by the Imprimerie Royale 

containing information on pigments used both in the dyeing industry and for 

painting. In the series Description des arts et metiers fa ites 014 approuvees par Messieurs 

de l'Academie Royale des Sciences, many articles were published which were publications 

in their own right; subjects include pigments used fo r dyeing, pigments 

for painting on porcelain, different types of glues, and the making of parchment. 

14. For example, a translation of a publication of the Royal Society of London on 

minerals by E. Hussey with reference to their use in painting was published by 

Dijonval. See Delaval. 1778. Recherches experimentales sur fa cause des changemens de 

couleurs dans les corps opaques et naturellement colores. Translated from the English 

by Dijonval. Quatremere: Paris de l'imprimerie de Monsieur. 

15. Mengs, A.-R. 1786. Oeuvres completes d'Antoine-Raphael Mengs, premier peintre du 

roi d'Espagne, &c. Contenant d!ff erens, traites sur la theorie de la peinture. Traduit de 

l'italien. Vols. I-II. Paris. Mengs wrote various treatises on the theory of art. See 

also Lairesse, Gerard de. 1787. Le grand livre des peintres, 014 f'art de la peinture 

considere dans toutes ses parties, et de,nontre par principes; avec des riflexions sur les 

ouvrages de quelques bons ma/tres, et sur les difauts qui s'y trouvent. Auquel on a joint 

les principes du dessin du meme auteur. Traduit du Hollandois sur la seconde edition. 

Vols. I-II: Paris. (Dutch original: Groot Schilderboek. Haarlem. 1740) . Lairesse was 

considered among the best painters of his time. His treatise was well known to 

eighteenth-century writers and artists, and his work is representative of practice 

in the early eighteenth century. Book IV is entirely devoted to colors and coloring; 

there are also chapters on drawing, composition, portraits, ceiling painting, 

engraving, etc. 

16. Observations periodiques sur fa physique, et les beaux-arts continued from 1756 to 

1823, although the title changed to Journal de Physique et chemie d'histoire naturelle 

et des arts. 

17. In the Journal Oeconomique, 1751-1772, in addition to various articles of interest, 

the tables of merchandise arriving into the ports of London, Amsterdam, etc., 

were published, with prices and exchange rațes for various materials including 

28 


some used for painting (i.e., amber, brazilwood and cochineal, gum arabic, indigo, 

saffron, sandarac, shellac, verdigris, vermilion, etc.). 

18. The Royal Academy in London was founded in 1768. By that period, the exchange 

of information all over Europe was so rapid that translations of new 

methods and techniques happened very quickly. See also Carlyle, L. 1991. A 

critical analysis of artists' handbooks, manuals and treatises on oil painting published 

in Britain between 1800-1900: with reference to selected eighteenthcentury 

sources. PhD. diss. Courtauld Institute of Art, University of London. 

Also, compare early German source books, including Schiessl, U. 1989. Die 

deutschsprachige Literatur zu Werkstoifen und Techniken der Malerei von 1530 bis ca. 

1950. Worms. 

19. The layout of Diderot's great work is not alphabetical but systematic, after the 

division of the sciences following the scheme of Francis Bacon (1561-1626) fo r 

an Encyclopaedia oj Nature and Art, which was published in 1620 (the encyclopedia 

itself was never published). Painting is listed under two categories: philosophy 

and imagination. Alphabetical arrangement of the contents of an encyclopedia 

came into general use in the eighteenth century, and indexing was not employed 

until the 1830s. For fu rther information of the dramatic history of Diderot's 

encyclopedia, see Collison, R. 1964. Encyclopaedias: Their History throughout the 

Ages. London. 

20. Copying from previous authors was commonplace in the seventeenth and eighteenth 

centuries; everyone did it, and it was unusual when a source was cited. 

One splendid example is the recipe for making drying oil by boiling nut oil 

with litharge, water, and one onion. None of the authors really believed in the 

efficacy of the onion, but dutifully copied it nonetheless. See Pernety, D. A,-J. 

1757. Dictionnaire portatif de peinture, sculpture et gravure. Paris, Ixxxvii-Ixxxviii. 

Pernety copied the recipe from La Hire (La Hire, op. cit., 708): "II y en a qui 

font cuire avec l'huile un oignon coupe en plusieurs morceaux pour la degraisser 

& pour la rendre plus coulante & moins gluante, a ce qu'ils pretendent." De Piles 

copied it again with similar doubts and suggested using a piece of bread instead 

(De Piles. 1776. Elemens de peinture pratique. Charles-Antoine Jombert, 141-42). 

21. Lacombe, J. 1752. Dictionnaire portatif des beaux-arts. Paris. See also Macquer, 

p. J. 1766-1767. Dictiormaire portatif des arts et metiers. Paris. For information on 

the many dictionairies published in the later half of the eighteenth century, see 

Massing's 1990 bibliography (available from the author). 

22. Pernety, op. cit. (note 20), iii-iv. 

23. Several authors refer to the imperfections of the oil medium. De Piles. 1766, op. 

cit. (note 20), 97-98. De Piles writes of the deterioration of oil painting, "Cette 

espece de peinture est moderne en comparison des autres . ... II n'y a pas de 

donte qu'elle seroit la plus parfaite de toutes les manieres de peindre, si les 

couleurs ne se ternissoient point par la suite des tems; mais elles brunissent 

toujours de plus en plus & tirent sur un jaune brun, ce qui vient de I'huile avec 

laquelle to utes les couleurs sont broyees & incorporees." 

24. Caylus, M. Ie Comte de, and M. Majault. 1755. Memoire sur la peinture a l'encaustique 

et sur la peinture a la eire. Geneva and Paris. 

25. Eighteenth-century authors such as the Count of Caylus believed that the Ancient 

Greek painters used an encaustic painting technique. The wax was "fixed" 

to the wooden support by placing the completed painting next to an open fire. 

Since the only painting materials Pliny mentioned were wax, pigments, fire, and 

brushes, Caylus dismissed the use of solvents such as turpentine to soften wax 

and make it "paintable." To paint as the ancients did, Caylus considered that the 

wax needed to be heated. The watertight metal boxes depicted here could be 

filled with boiling water, which kept the colored waxes at a constant temperature. 

The box on the left has a roughened glass top used to grind the colors. A similar 

box with an opaque top would be a palette. 

26. Fratrel,J. 1770. La cire alliee avec l'/wile ou la peinture a /wile-eire. Trouvee a Manheim 

par M. Charles Baron de Taubenheim experimerttee decrite & dedie a l'electeur. Manheim. 

27. For further discussion of this method of painting, which was taken quite seriously 

not only by the inventor himself, but by most contemporary source books, see 

Massing, A. 1993. Arnaud Vincent de Montpetit and eludoric painting. Zeitschrifi 

jur Kunsttechnologie 7 (2):359-68. 

28. For a discussion of this popular technique, see Massing A. 1989. From print to 

painting. The technique of glass transfer painting. Print Quarterly 6 (4):382-93. 

29. Rouquet,J. A. 1755. L'art nouveau de la peinture en jromage, ou en remequin, inventee 

pour suivre Ie louable projet de trouver graduellement des jafons de peindre injerieures a 

celles qui existent. Marolles. 

Massing 29

Abstract 

The Roberson Archive, the archive 

of the artists' colorman Charles 

Roberson & Co. (1819-1985) is 

housed at the Hamilton Kerr Institute 

in Cambridge. Roberson was 

one of the most important colormen 

in nineteenth-century London. The 

material in the archive gives detailed 

information about the internal 

workings of the company and its relations 

to both its customers and 

suppliers. It is the largest artists' colorman 

archive in the United Kingdom 

and covers the period from 

1815 to 1944, providing a record of 

the materials and techniques of 

many of the leading nineteenth- and 

twentieth-century British artists. In 

addition to catalogues, sample books, 

and more than three hundred account 

ledgers, there is a collection of 

pigments and objects sold by Roberson 

that provides material fo r analysis 

and study. A three-year research 

project is currently being undertaken 

to catalogue the archive, compile a 

database of the most important documents, 

and make the archive accessible 

for research. 

Figure 1. Roberson's shop at 154 Piccadilly, 

ca. 189{}-1911. Fitzwilliam Museum, University 

of Cambridge. 

The Roberson Archive: Content and Significance 

Sally A. Woodcock 


University of Cambridge 

Whittlesford 



The company's history 

Charles Roberson opened his first recorded shop to sell artists' materials at 

54 Long Acre, London. Although accounts of the firm's history give the date 

as 1810, Henry Matley is listed as "colourman to artists" at this address in a 

trade directory of 1817, and a note in a recipe book dating from around the 

turn of the century states that Roberson did not succeed Matley until 1819 

(1). At that time, the shop was in the heart of the artistic area of London, 

with the Royal Academy Schools being based nearby in Somerset House on 

the Strand. In 1828 Roberson and his assistant Thomas Miller became partners 

and moved to 51 Long Acre (2). The partnership dissolved in 1840. 

The company continued to trade on Long Acre, moving to the premises they 

built at number 99 in 1853. In 1868 the Royal Academy Schools moved to 

Burlington House, Piccadilly, and Roberson opened a retail branch on Piccadilly 

in 1890 (Fig. 1) (3). Over the following years, however, many artisans 

and manufacturing trades moved away from the center of London to be 

replaced by retailers and offices; in 1937 Roberson closed both the Piccadilly 

and Long Acre branches, moving to Parkway, Camden (4). They retained a 

West End presence for an unspecified length of time, arranging with the 

Medici Galleries in Grafton Street to move a "representative stock of Artists 

materials" from their closed Piccadilly branch in January 1940, to be sold 

from the galleries on commission (5). Two successive French addresses appeared 

in their catalogues after 1870, which they described as their depots in 

Paris, but which were in fact the premises of Parisian colormen (6). 

The company remained in the Roberson family until sold to a Dutch firm 

in the 1970s. G. F. Roberson Park remained a board member (7). In 1985 it 

went into liquidation. At liquidation the name was bought by the London 

colorman Cornelissen and a small range of high-quality materials bearing the 

Roberson name is still produced (8). 

The arrival of the archive at the Hamilton Kerr Institute 

In 1975 Roberson's historic accounts ledgers were moved to the Hamilton 

Kerr Institute (9). They were loaned to the Institute for safekeeping and 

research and, when the company was dissolved, became part of the Fitzwilliam 

Museum's manuscript collection. The archive is now housed in the Institute 

and will be available for research once cataloguing is complete. 

The archive consists of a collection of objects, including a large number of 

pigment samples, and more than three hundred ledgers dating from 1815 to 

1944. There are several continuous series of different types of ledgers, but 

there are many gaps, and some sets appear to be incomplete (10). Despite 

this, it is the most extensive artists' colorman archive in the United Kingdom. 

The manuscripts in the Roberson Archive give a detailed picture of the 

company's activities. Only the most important categories are discussed here, 

as many of the warehouse, order, day, and sundries books contain information 

duplicated in the main accounts books. 

Catalogues and sample books 

The catalogues preserved in the archive date from 1840-1 853 to 1926-1933 

and illustrate goods sold by Roberson. They offered a fairly conventional 

30 


ange of artists' materials, books on art, and plaster casts of antique statues. 

Although they sold some patented equipment (such as a sketching stool that 

converted to a walking stick and the "Eiffel," a stool made of steel), Roberson 

avoided the more innovative artists' equipment patented in the nineteenth 

century, such as easels that also served as bicycles or rowing machines (11). 

Most products were sold under the Roberson name, although a few materials 

such as paper were usually identified by the name of the manufacturer. It is 

therefore difficult to establish which materials were made by Roberson and 

which brought in to be made up and labeled on the premises. It appears that 

the company prepared their own canvas throughout the company's history, a 

room used fo r this was still extant in the Parkway workshop until liquidation 

(12). They also mixed their own paints, buying pigments and other materials 

from wholesale suppliers, in common with most other artists' colormen, not 

manufacturing the raw pigments themselves. In general they only bought 

small quantities of prepared paints from other artists' colormen, sometimes as 

little as one tube or cake bought for a specific customer. However, the popularity 

of Roberson's Medium, used fo r oil painting, was such that both British 

and foreign colormen bought wholesale quantities of it from Roberson. 

After the First World War, when Roberson's business was beginning a gradual 

decline, trading practices began to change and by the period following the 

Second World War, a reciprocal arrangement with the London colorman 

Rowney was made to divide trade into two spheres, Roberson supplying 

Rowney with canvas and Rowney supplying Roberson with paints (13). 

The catalogues indicate the cost of artists' materials and demonstrate prices 

were stable for much of the nineteenth century. There is a clear differentiation 

between luxurious pigments such as ultramarine or carmine and other, cheaper 

colors; the catalogues show that an ounce (28.35 g) of genuine ultramarine 

was twenty-eight times as expensive as its artificial substitute between around 

1840 and 1911, with prices almost completely stable in that period. The 

catalogues also reflect the introduction of new pigments and demonstrate that 

there could be a considerable time lag between the discovery of a pigment 

and its commercial application. Roberson & Co. appears to have been comprised 

of rather reactionary colormen, who contributed little to technical 

innovation, emphasizing the hand-prepared nature of the company's products 

until long after many of the other colormen had introduced a degree of 

mechanization; their reaction to new developments, therefore, may not be 

characteristic of the market as a whole. 

Three different prices were given in the catalogues until 1920: wholesale, 

retail, and professional. After 1920, discounts offered by artists' colormen to 

professional artists were abolished after complaints of abuse (14). 

In addition to catalogues, the archive contains a number of sample books of 

canvas, paper, and paints, both from Roberson and other companies, that are 

of great use in providing identified, untreated material fo r analysis. Projects 

have been carried out to establish the composition of the ground on the 

canvas samples and the pigments in the paint samples. 

Price books. recipe books. and notebooks 

Price books indicate both the cost to Roberson of a range of materials and 

the resale price. Many of the cost prices are in an alphabetical code, illustrating 

the importance of secrecy to the company; an alphabetical code is also used 

in the recipe books. 

The recipe books in the archive contain a number of formulae for paints, 

media, and grounds, as would be expected from an artists' colorman, but also 

have recipes for trifle, lemon pickle, wine, boot polish, and blacking fo r harnesses. 

This illustrates the early connections between colormen, grocers, and 

apothecaries, professions which were separating into distinct trades in the 

nineteenth century (15). 

Woodcock 31

Figure 2. Charles Roberson's recipe fo r Roberson's Medium, originally in sealed envelope. Fitzwilliam 

Museum, University of Cambridge. 

The books also demonstrate a degree of experimentation in the formulation 

of artists' colors; the composition of mixed colors in particular, such as olive 

green or neutral tint, was not absolutely fixed. Notes in the recipe books' 

margins indicate that formulae were not always successful, and in the later 

recipes a degree of substitution was effected (16). 

The recipe books show the importance given to secret recipes and trade 

secrets. Many of the recipes are written in code or shorthand or have certain 

portions in Greek. The fo rmula of Roberson's Medium, the company's most 

successful product, has been struck out in a recipe book (17). It is contained 

in a once-sealed envelope from Charles Roberson to his nephew Charles 

Park that is marked, "To be opened when I am dead," and dated 1868, ten 

years before Roberson's death (Fig. 2) (18). Roberson had reason fo r secrecy. 

When a Mr. Ellis advertised in the Daily Telegraph of 1900 that he would sell 

"the fo rmula fo r making this famous medium fo r imparting permanency to 

oil paintings" at a price of seventy guineas, Roberson quickly threatened 

prosecution. Ellis's solicitors responded the next day stating that he had "no 

intention ... of selling the medium as 'Roberson's Medium,' " had effected 

no sale of either the fo rmula or the medium, and would discontinue the 

advertisement (19). 

Personal accounts 

The personal account books are among the most important records in the 

archive. They are incomplete, but provide detailed information about the 

materials and techniques of many of the most prominent nineteenth- and 

twentieth-century artists in Britain. They list in most cases what an artist 

bought and how much, and when he or she paid (Fig. 3) . They also demonstrate 

the range of services Roberson would perform, such as sending 

workers to artists' studios to carry out a variety of tasks, transporting paintings 

to and from exhibitions and loaning equipment to artists. William Holman 

Hunt's accounts show that he frequently changed his mind as to the proportions 

of his paintings, and that Roberson regularly sent workers to add strips 

of canvas and enlarge stretchers, particularly toward the end of the artist's life. 

Account holders were not only confined to artists living in Britain. Jozef 

Israels in the Hague, Louis Raemaekers in Brussels, Fritz Voellmy in Basel, 

and the American artist Wilton Lockwood all bought from Roberson. 

The accounts record extensive use oflay figures, usually rented by artists from 

Roberson, as their purchase price was very high (Fig. 4) There are a number 

32 


Figure 3. Personal account book, 1881-1894, MS 250- 1993, 37-8. Fitzwilliam Museum, University 

of Cambridge. 

of both full- and child-size figures in the archive; their composition-stuffed 

knitted textile over a wooden and metal frame with papier-mache and gesso 

heads-helps to explain why they were so frequently sent fo r repair and why 

Roberson entered into disputes with artists over damage sustained in use (20) . 

Despite this, they were probably cheaper and more reliable than living models, 

of which Roberson kept a register to assist customers (21). 

The accounts also show the range of Roberson's customers from wealthy, 

well-established painters to professional graphic artists, scene painters, and 

others. Many amateur artists also held accounts with Roberson, including 

several members of royalty, both in England and abroad. Roberson appears 

to have been supplying the upper end of the market and although it did not 

hold the royal warrant fo r Queen Victoria, Roberson supplied the Queen 

with sketching materials during her reign and was granted the royal warrant 

to Queen Alexandra in 1912 and to King Vittorio Emanuele III of Italy in 

1903. In 1949, Gerald Kelly remembered in a letter to Roberson, "When I 

first started painting I could not afford to deal with what I was then told was 

the best firm in London, but as soon as I got a little money I got my things 

from you, and gradually I have bought more and more ..." (22). 

Figure 4. Lay figure: Child no. 98. Fitzwilliam 

Museum, University of Cambridge. 

Some artists had their materials specially made to their own recipes or to a 

formula supplied in the past which they particularly liked. Leighton asked for 

his paints to be made "extra stiff," a preference alluded to in a letter from a 

fellow artist in 1888: "I have used your colours fo r several years, and I can 

with the greatest confidence say that they have given me every satisfaction. 

There is one quality which is greatly esteemed-and this Sir Frederick Leighton 

says is a great desideratum, that is stiffness" (23). 

Roberson, however, did not always think the customers' demands were advisable. 

A notebook entry dated 27 November 1882 records, "Sir F. Leighton 

P. R. A. in 1881, 82, had canvases prepared with different grounds made to 

his order & warranted not to stand; plaster of Paris &c. with Lac Varnish over 

&c. &c." (24). In June 1888, Millais is also recorded as having "canvases prepared 

thickly with one coat of turpentine color to his order, also prepared 

his own canvases. Not warranted to stand" (25). 

Woodcock 33

Bought ledgers 

The bought ledgers help to complete the picture of Roberson's trading practices. 

They date from 1854 to 1931 and indicate the limited extent of Roberson's 

manufacturing activities and its widespread trade with manufacturers 

and wholesale suppliers, both in Britain and abroad. From this, it is possible 

to reconstruct the chain of supply from raw material to artist. The ledgers 

also show how often Roberson & Co. printed its catalogues and where the 

company advertised; the ledgers even give details of the fixtures and fittings 

of their shops and workshops. 

One of the primary uses of the bought ledgers is to indicate the dating and 

extent of use of a variety of pigments or pigment mixtures introduced 

throughout the period in which Roberson traded. The popularity of a new 

pigment and the decline of its predecessor can be gauged from Roberson's 

purchases from the wholesale supplier, providing a fu ndamental record of the 

dissemination of new colors and materials. 

Letters 

The letters preserved in the archive are, by their nature, a disparate group of 

materials, but provide detailed and wide-ranging information about the company. 

Roberson kept much of its more interesting correspondence, particularly 

letters expressing approval of its products or restoration treatments. From 

1919 to 1940 they also kept handwritten (and later typed) copies of their 

answers to correspondence. Some letters were later used in their catalogues 

as testimonials; for example, the letter from W P. Frith of January 1897 referring 

to his painting in the Tate Gallery, London: " . .. you would like to 

know that it is just fifty years since I painted the Derby Day with your colours 

and medium, and that in so long a time there is no change whatsoever in 

the materials used" (26) . 

The letters also record advice both to and from artists concerning technique 

and the fo rmulation and application of materials. This can often be most 

informative, such as Thomas Gambier Parry's letter in 1874 from Ely, where 

he was working on the cathedral; the letter provides both a recipe and instructions 

for making up the "usual common encaustic." 

The letters also help to confirm information about developments within the 

company, such as a letter to the Factories Inspector announcing the installation 

of electric color grinding machinery in 1919, contradicting its catalogue 

note of around 1926 in which the company states that it is convinced that 

paints "which are ground by hand under the muller give results superior to 

those ground by machinery" and that the company, therefore, continued to 

use the former system (27). They also give some details of the conditions of 

the work force Roberson employed, mentioning industrial accidents and the 

possibility of redundancy due to ill health. 

The accounts books demonstrate how long customers were prepared to delay 

payment of their bills and many of the letters refer to attempts to recover 

bad debts, using an investigation agency if necessary. This demonstrates why 

Roberson asked fo r customers to provide references from reputable London 

businesses before opening an account. The bad debts were not always the 

result of fraud. Before the establishment of the welfare state in Britain, medical 

treatment could be ruinously expensive; therefore illness or death often left 

dependents at the mercy of creditors. 

Throughout the nineteenth century, artists and writers expressed unease about 

both the quality of artists' materials and the soundness of contemporary technique. 

This erupted into a public debate in 1880 when William Holman 

Hunt, a member of the Pre-Raphaelite Brotherhood and a long-standing 

customer of Roberson, gave a paper to the Royal Society of Arts that criticized 

the quality of artists' materials and accused the color men of either being 

deceived by their suppliers or of adulterating the products themselves. Hunt's 

34 


paper resulted in a volume of correspondence and a leading article in the 

Times. The background to this can be traced in the Roberson Archive, as 

the anonymous colorman cited by Hunt as supplying him with adulterated 

paint was, in fact, Roberson. 

It is significant that Hunt did not transfer his account from Roberson to 

another colorman and continued to buy from the company until shortly 

before his death. Roberson took his accusations seriously and painted out 

samples of pigment and made up a chart of a number of aging tests they had 

carried out (Plate 5). The company's concern for the quality of its pigments 

is evident in the twentieth century when it used a chemical laboratory to 

analyze samples of pigment from the wholesale suppliers. Roberson also sent 

materials to artists fo r testing; and he corresponded with both Professor 

Church, professor of chemistry at the Royal Academy, and Helmut Ruhemann 

at the National Gallery in developing and assessing materials. A project 

to analyze pigments in the archive fo r adulterants has been carried out at the 

Hamilton Kerr Institute with mixed results: a small number of spectacularly 

adulterated pigments (with much use of fillers and extenders) were found, 

but there was less adulteration than the literature of the time and Hunt's 

accusations would suggest (28). 

Information from objects 

The archive contains a number of objects, most of which can be identified 

in the company's catalogues. These objects provide a useful set of references 

against which to check the documentary information. For example, analysis 

of the pigment collection has revealed the unreliability of many of the recipe 

books, with adulteration and substitution being common. The objects in the 

collection also illustrate the development in the nineteenth century of paint 

containers, including both a paint bladder and the ivory pins used fo r piercing 

it, a set of brass paint syringes, the forerunner of the collapsible tube, and 

tubes adopted in the 1840s after they were first patented and advertised (Plate 

6) . 

Other activities of the company 

In addition to supplying artists' materials, Roberson was involved in a number 

of other activities that can be traced in various sources in the archive. These 

included restoration, dealing, and publishing. Trade directories show that it 

was not uncommon fo r color men to be dealers and restorers in the nineteenth 

century, but gradually the professors came to be listed separately, although 

Roberson's accounts indicate that the company continued all three 

pursuits until its records end in the 1940s. 

Historical context 

Many aspects of the history of the Victorian and Edwardian period can be 

gleaned from the Roberson Archive. The presence of an extensive empire is 

felt both in terms of the materials bought from the colonies and in the 

products supplied, which were designed to withstand extremes of climate not 

experienced in Britain. Stylistic movements were also reflected in Roberson's 

catalogues; materials for illumination, missal painting, and heraldic hatchments 

became popular during the Gothic revival. Roberson also responded to the 

Victorian expansion in public building and the popularity of painted interior 

decoration by fo rmulating a number of media and specially prepared canvases 

to imitate fresco. The company was involved in both supplying the materials 

and erecting the canvases in a number of public buildings in Britain. 

Roberson also supplied materials for the great explorations of the period 

since, even though photography was used by this time, artists were still often 

sent on expeditions to record the results. A tube of yellow ochre taken on 

the 1912 Shackleton South Pole expedition was given to Roberson by the 

expedition artist George Marston on his return. When the expedition was 

Woodcock 35

stranded by pack ice that crushed their ship, Marston mixed his paints with 

lamp wick to caulk the seams of the open boat in order to successfully fe tch 

help. The Egyptologist Howard Carter was also a customer; he bought artists' 

materials, using Roberson as an agent, and sent cases of drawings of his finds 

to the company fo r distribution. 

The effect of both world wars is seen, to some extent, mainly as a series of 

problems of supply, raw materials being difficult to obtain and materials such 

as paper subject to government restrictions. Some pigments were also in short 

supply, vermilion increasing in price by 120 percent in 1940 because of shortages 

of mercury, which was used fo r shell and mine detonators, and because 

difficulties with chrome colors were being anticipated as they were used fo r 

dyes fo r khaki and in the manufacture of poison gas (29). Roberson was 

fortunate in sustaining only slight damage from an incendiary bomb in 1940 

and avoiding the more extensive damage experienced by other colormen. 

Post-war disruption is also evident; letters from Roberson in 1919 showed 

the difficulties of supply and delivery facing the colormen. Canvas was in 

short supply, materials usually coming from Russia were unobtainable, and 

turpentine was scarce and expensive; even tin tubes in which to pack the 

colors were difficult to find (30). Similar difficulties were experienced during 

World War II. The world wars, the depression of the 1930s, and the rise of 

new reproduction methods contributed to Roberson's decline in the twentieth 

century. 

The archive as a research resource 

The Roberson Archive is chiefly relevant to the study of the materials and 

techniques of British artists. Its relevance outside the United Kingdom is 

confined to the small number of foreign artists who bought from the company 

and the larger number of suppliers and retailers who traded with them. 

These contacts include firms in all five continents, but the most numerous 

contacts were in France, Germany, and the United States. 

The archive has been used in the last year to follow up a number of queries 

from both the United Kingdom and abroad. Although access will be restricted 

until the end of 1996 (when the cataloguing project ends), thereafter it will 

be available for study. A database of account holders, regarded as the most 

informative part of the archive, is being compiled, along with details of their 

purchases, and it is hoped that a checklist of account holders will be published. 

The database will make it possible to search for both particular artists and 

specific materials and, when used in conjunction with information from recipe 

books and catalogues, should provide a clearer picture of what paintings 

were made of and what artists really used in the nineteenth and early twentieth 

centuries. 

Acknowledgments 

The Roberson Archive is published by permission of the Syndics of the Fitzwilliam 

Museum, University of Cambridge and is generously funded by the Leverhulme 

Trust. 

Notes 

1. Gallagher, M. 1991. The Roberson Archive: inventory oj written and printed material. 

Unpublished typescript. Cambridge: Hamilton Kerr Institute, 2. See also Kent's 

Original London Directory. 1817. Hamilton Kerr Institute (HKI) MS 785-1993, 

fol. 54v, 223. See also C. Roberson & Co. Ltd. 1969-1970 Catalogue, ii. Charles 

Roberson does not appear in trade directories at this address until Pigot's directory 

was published (Pigot & Co. 's Metropolitan new alphabetical directory. 1922- 

1923, 68). A letter to the National Gallery in London dated 1923 states that 

Charles Roberson founded the company in 1819 (HKI MS 862-1993, 293). 

2. Kent's Original London Directory. Op. cit. (note 1). 

3. HKI MS 204-1993, 638. 

36 


4. HKI MS 865-1993, 497. The move is attributed to increasing congestion caused 

by the growth of Covent Garden Market in a note on the firm's history in a 

Roberson catalogue of 1969-1970, ii. A letter of 18 January 1940, however, states 

that they were not going to renew their lease on the Piccadilly branch "owing 

to conditions caused by the war," HKI MS 866-1993, 47. 

5. HKI MS 866-1993, 51-2. 

6. The first address is 26 Rue Chaptal, the premises of E. Mary & Fils, later Geo. 

Mary, who gave up business in 1920. The second address is 3 Quai Voltaire, 

where G. Sennelier & Fils are listed between 1920 and 1935. 

7. Gallagher, op. cit., 3. 

8. Walt, N. 1993. Personal communication. L. Cornelissen & Son, 105 Great Russell 

Street, London WC1B 3RY 

9. Cob be, A. 1994. Personal communication. A full account of the arrival of the 

archive at the Institute is recorded in Gallagher, op. cit., 3-4. 

10. The archive is listed in the Cambridge Union Catalogue of Departmental and 

College Libraries, and was compiled by Paul Woudhuysen, Keeper of Manuscripts 

and Printed Books, Fitzwilliam Museum. 

11. McGrath, M. 1973. A catalogue of some eighteenth and nineteenth century patents 

dealing with artists' instruments. Unpublished typescript. London: Courtauld Institute 

of Art. 

12. Walt, N. 1994. Personal communication. 

13. Walt, N. 1993. Personal communication. 

14. HKI MS 865-1993, 191. 

15. This is demonstrated to even greater effect by H. Hodder, described as "Chemist 

& Druggist, Oil and Colorman, Tallow Chandler, Tea Dealer, and Tobacconist," 

whose recipe book (1823-1858) covers fo od and drink, household cleaning 

agents, cures fo r both human and animal ailments, and recipes fo r varnishes, oils, 

and paints (Guildhall Library MS 16262, by permission of the Guildhall Library, 

Corporation of London) . The preface to a 1875 Post Office Directory clarifies 

the separation of the trades: "Groceries in early times consisted chiefly of spices 

. . . but subsequently came to comprise confectionery, dyes, drugs, chemicals, 

whale oil, &c.; the confectioner, the druggist, and the oilman have now branched 

off into separate and distinct trades" (The 1875 post office directory of the grocery 

and oil and color trades. 1875, iv) . 

16. Townsend, J. H., L. Carlyle, S. Woodcock, and N. Khandekar. n.d. Later nineteenth 

century pigments: analyses of reference samples, and evidence fo r adulterants 

and substitutions. The Conservator. Forthcoming. 

17. HKI MS 788-1993, fo1. 45v. 

18. HKI MS 892-1 993. Charles Roberson died and was succeeded by his nephew 

in 1878, according to the note on the firm's history in a 1969-1970 catalogue, 

11. 

19. HKI MS 891-1993. 

20. HKI MS 864-1993, 142. 

21. HKI MS 865-1993, 373. 

22. HKI MS 527-1993. 

23. HKI MS 500-1993. 

24. HKI MS. 785-1993, fo1. 37r. 

25. Ibid., fo1. 41r. 

26. HKI MS 597-1993. 

27. HKI MS 860-1993, 30. 

28. Townsend, J. H., et aI., op. cit. See also Woodcock, S., and N. Khandekar, n.d. 

The Adulteration of Twentieth Century Pigments in the Roberson Archive. Forthcommg. 

29. HKI MS 866-1993, 164, 219. 

30. HKI MS 860-1993, 6, 15, 22, 83, 85. 

Woodcock 37

Abstract 

Two texts from Simone de Monte 

Dante's medieval Italian manuscript 

are presented here in English translation. 

With their many recipes, these 

texts shed light on the technique 

and materials of fifteenth-century 

Italian manuscript illumination and 

the historical development of the illuminator's 

craft. Studying the relationship 

of this manuscript to other 

treatises may provide a better understanding 

of the mechanisms involved 

in the adaptations and transmission 

of medieval art technology. 

Figure 1. Rome, Biblioteca Casanatense, 

MS 1793, cover. 


MS 1793, jol. 10v. 

38 

Libro Secondo de Diversi Colori e Sise da Mettere a Oro: 

A Fifteenth-Century Technical Treatise on Manuscript 

Illumination 

Arie Wallert 


Museum Services Laboratory 

The J. Paul Getty Museum 

17985 Pacific Coast Highway 

Malibu, California 90265 

USA 

Introduction 

The Biblioteca Casanatense in Rome holds a very interesting fifteenth-century 

technical art manuscript. It has the shelf mark "MS 1793" and shows on 

folio 2r ("r" for recto) the author's name and the date of the manuscript: 

"Questo libro et ne di Simone de Monte Dante dela Zazera . .. 10 chopiai 

el mese di novembre 1422" (Fig. 1). 

The manuscript contains two closely related treatises on folios 10v-13v ("v" 

for verso) and 15v-20v. Written in different hands, these texts contain recipes 

devoted to the practice of manuscript illumination. Various aspects of the art 

are described, such as ink production, parchment preparation, gold leaf application, 

and the preparation and application of pigments and various transparent 

organic colorants. 

The recipes in the Simone manuscript are not related to the well-known 

Mappae Clavicula tradition, the texts from Theophilus' Schedula de Diversarum 

artium, the Eraclius texts, or the texts of the Liber de Coloribus. What makes 

this manuscript so important is the fact that it appears to be related to a 

particular group of technical art treatises comprising a number of littleknown, 

mostly unpublished recipe texts that are specifically devoted to the 

miniaturist's craft. 

The relationships of parts in Simone's treatise with those manuscripts in the 

Siena Biblioteca Comunale, the Modena Biblioteca Estense, the Bologna Biblioteca 

Universitaria, the Oxford Bodleian Library, and the Florence Biblioteca 

Nazionale are quite intriguing. The recipes of both treatises in Simone's 

manuscript are given here in translation, with notes indicating the relationships 

to other recipe texts, as well as references to relevant technical literature 

(1) . 

Relationships to recipes in the well-known Bolognese manuscript are referred 

to according to the numbering and pagination of Merrifield's edition (2). 

Relationships with two treatises in the Siena Biblioteca Comunale follow the 

numbering in Wallert's edition (3). To facilitate the understanding of these 

relationships, Simone's recipes-originally not numbered-are given numbers 

in brackets in this translation. 

The Simone manuscript is clearly not written by a person with a firsthand 

knowledge about the products and techniques he describes. Occasionally, 

Simone prescribes the wrong ingredients, gives only part of a recipe, or 

gives some recipes in which the final outcome clearly will not be the result 

promised by the heading; obviously, he was not a craftsman. Sometimes his 

descriptions are too succinct to be fully understood by the uninitiated or 

modern reader. In those cases I have added to the Simone text, placing these 

additions in brackets to distinguish them from the original text and to make 

the recipes more intelligible. 

With folio 10v, Simone begins a book on various colors and grounds to lay 

gold (Fig. 2). Following is the text of folios 10v-13v (page changes in the 

original will be noted here with the symbol {}): 


[1] How to make a gold ground. Take fine, slaked gypsum, as much as the size 

of a nut, and grind it with a little bit of clear water. Then take Armenian 

bole, as much as a chickpea or a bean, and grind this separately, and also with 

water. Having done this, mix them together, take some dissolved glue and 

add it. Next have a bit of candy or sugar and some earwax and grind it gently 

together. The glue must be so strong that the whole matter sticks a little bit 

to the porphyry stone when you grind it. Then you keep it warm above 

some hot coals. When you work with it you can add some glue so that it 

will be a bit fluid. Let the dissolved glue stand a few days so that it comes 

from the fire with better strength. Now you can work with it. When the 

work is dry, scrape it until it is smooth. Then wet it with clear water and lay 

the gold on it. Press on it when it is dry. If the gold ground would be too 

weak for you, distemper it: you can put a bit of glair in the water (4). 

[2] To make lines of burnished gold. Take slaked gypsum, grind it with glair and 

a bit of honey, some earwax, and a bit of glue. Let it dry [when applied], then 

huff on it and lay the gold. Burnish it (5). 

[3] To make lines of gold in another way. Take [gum] ammoniac and grind it 

without water or anything else. Then take garlic juice and grind it with it. 

Next add a little bit of Armenian bole to it. Work with it. When this is done; 

huff on it and lay the gold on it (6). 

[4] To make the blue clothlets. Take a plant called turnsole, of which you pick 

one-by-one the flowers and clustering {llr} berries. Then you put these in 

a mortar, according to the quantity you want to make. Crush them and press 

the juice out of the aforesaid berries. You put this juice in a glazed bowl and 

then you take pieces of cloth that are white and coarse. Then you put these 

into the aforementioned bowl [with the juice] and let them soak well in it. 

Do this three times, and every time that you take it out of the mentioned 

bowl, you let it dry well on a wooden bench. Before you reach this stage, 

make sure you find a remote quiet hole in the ground, and urinate much in 

it, six days before you come to make the said turnsole. Take straw and make 

a screen over the urine and place the cloths above the humidity of this urine. 

Thus it will become darker and more beautiful, and let it stand there for at 

least twenty days above the aforesaid urine and then the turnsole will have 

become beautiful. 

You prepare it as follows: Take glair according to the quantity you think this 

turnsole needs to have to become deeply colored. When you want to work 

with it, take this glair in a shell or some glazed small pot and then take the 

turns ole and put some of it in according to the right amount and let it stand 

[in the glair] for an hour and then press it out with your finger and the juice 

that comes out of the clothlet is the turnsole. With this you can work, using 

a little bit at a time (7). 

[5] To make a fine ink. If you want to make a fine ink, take one pound of 

galls. Break it into smaller pieces and put it in ten pounds of rainwater or 

water from the pot. Let it boil in a pot that has not previously been used, 

until the volume of this water is reduced to half. Next sieve it and throw out 

the substance of the galls and put the water back in after you have washed 

the jar. {11 v} Having this done, put in five pounds of the softest white wine 

you can find, i.e. replacing the amount of water that has evaporated. Let it 

cook and when it starts to boil, put in half a pound of perfectly pulverized 

gum Arabic, adding this bit by bit while stirring with an iron cane until the 

gum is dissolved. When that is done, take the aforesaid vase from the fire and 

then put in half a pound of perfectly pulverized Roman vitriol. Then cover 

it, so that no dust can get into it and let it stand for three or fo ur days on a 

quiet place, because when you leave it alone it becomes more beautiful and 

blacker. Then without forcing it, sieve it through a cloth into a glazed pot. 

And if you want this to be right, you sieve it after 15 days three times. Know 

that the vitriol loses its quality through smoke, that is why one does not heat 

it on the fire. If you cannot find white wine, you may take red wine, which 

Wallert 39

should be fresh and clear and of little color. Know that the quality of the 

right gum can be recognized in that it is difficult to break. The gum that 

comes in small grains is much better. The good quality Roman vitriol can 

be recognized by its blue color and [its texture], in that it is hard and grainy, 

like coarse salt. It is sufficiently good when the rain water or river water is 

soft, but better is that which comes from the fo untain. Good galls are recognized 

by their being small and wrinkled, and hard inside (8). 

[6] To make a good yellow. Take an egg and take out the white of it and put 

it in a glazed pot. Put in as much saffron as seems right to you. Take a wooden 

mixer and beat it so much that the egg white breaks and becomes nicely 

yellow. Then take a fine clothlet and sieve [the glair mixture] through it in a 

glazed dish and leave it {12r} in the sun to dry. Then take it out, keep it dry, 

and when you want to work with it, grind it with a little bit of water and a 

few drops of glair (9). 

[7] To make a fine lac. Take a man's urine, as much as you need, put it in a 

bowl and let it stand for eight days. Then put it in a pot and let it boil so 

long that there is no foam anymore. Then percolate it through strong ashes, 

so that the liquid that comes through is like a lye. When you have sieved it; 

take coarse gum lac and crush it until gets the appearance of bread crumbs, 

put it in a new pot, and put in the said ... (10). 

[8] To make indigo. Take flower of woad and flour of grain. Make a dough 

of it with urine and with vinegar. Make a cake of it and dry it in the sun. 

And if it is too light, take more flower of woad and mix it again until it takes 

on the color you want (11). 

[9] How to make vermilion. Take one part of mercury and one of white sulfur, 

as much of one as of the other. Put it in a glass bottle, thoroughly clad with 

clay. Put it on a moderate fire and cover the mouth of the bottle with a tile. 

Close it when you see yellow smoke coming out of the bottle, until you see 

the red and almost vermilion-colored smoke. Then take it from the fire and 

the vermilion will be ready (12). 

[10] To make burnished gold lines. Take slaked gypsum, ground with glair and 

a bit of honey, and some earwax subtly ground together with a little bit of 

bone glue. Let it dry when you have applied it. Then huff on it, lay on the 

gold, and polish it (13). 

[11] A way to test whether the ultramarine blue is good. Take an iron knife and 

hold it in the fire. Then take a little bit of the blue and put it on this 

knife, and if it is good {12v} it will become more beautiful and of a tender 

color. If it is not good, it will become black like ink (14). 

[12] To make blue. Take a glass pot and put in it fresh sulfur and mercury, 

thoroughly mixed. And take two parts of sulfur to three parts of mercury. 

Clad the pot with white potters clay and make sure that this is thoroughly 

mixed with horse dung. Put this to heat well on the fire, until the smoke 

comes out of it. Then take it off the fire and it will be a perfect blue (15). 

[13] To make a blue water. Take from the elder tree the berries with the 

kernels, at the time when they are in the stage between green and ripe. Cook 

them in a kettle and when they are boiled enough, take a cloth and squeeze 

the juice that comes out of them in a glass or glazed pot. Then temper it 

with urine and use it as a painter's color. When it is dry you can lightly paint 

with the juice oflilies with a brush on it, and it will be a blue color. Similarly, 

you can dye cloths with it or paper to write on with letters in gold. 

[14] How to make red lead. Heat litharge together with rasped lead in the fire. 

This makes the red lead. 

[15] How to make the dark "verzino. " If you wish to make the dark verzino, 

take about a "quarto" of rasped brazilwood. Put it in a beaker. Take white of 

egg, thoroughly beaten fine with a sponge and put it fo r eight days in a bottle 

with some realgar. Mix it together after those days. Then sieve this verzino 

40 


into a glass, and put in a bit of well ground alum, of about the size of a 

chickpea. Then take a piece of white linen and sieve the verzino and put it 

{ 13r} in a glazed bowl and put it to dry in a place where there is no sun, 

until it is dry. With the verzino you can make red. When you work with it, 

put a little bit of it in the shell and add a little bit of clear water (16). 

[16] How to make green. Take arzica and grind it and then you add pale azurite 

blue, so much that it takes the color you want. Know that the greens that are 

made with arzica need only a little bit of medium. 

[17] To make "p urpurina. " Take similar amounts of mercury and tin and have 

them amalgamated together and when they thus are thoroughly combined, 

let it cool down and grind it together. But first take "caschina" [?], which is 

separately ground very well (17). Then take an oven with red hot coals and 

put [the ingredients] in a bottle or pot on the lit coals, and let it stand until 

smoke comes off, and make sure that mouth of the bottle stays open. 

[18] To take letters off from the parchment. Take one pound of saltpeter, one 

pound of vitriol, and distill it. Then take a sponge, dip it into this water, rub 

it on the parchment and the letters will disappear (18). 

[19] To make "origno" (19). Take lead and sulfur and nitric salt and mercury 

in the same quantities together and put it in a bottle of which the mouth is 

wide (20, 21). Put it on a moderate fire and when it starts to look glittery 

like gold on the mouth of the bottle, take it off the fire and it will be done 

(22). 

[20] To make red. Take very well ground eggshells, treat them on the stone 

and put them in a new, hollow brick. And take verzino and boil it with a 

little bit of roche alum and pour it on the eggshells. Mix it and of this you 

add until is becomes as light as you wish (23). {13v} 

[21] To make the yellow "arzicha. " Take very well ground eggshells and put 

them in the hollow of a new brick. Take the weld herb of the textile dyers 

and let it boil in water with a little bit of alum. Pour it on the eggshells and 

thus make it as light or as dark as you wish (24). 

[22] To make green. Take arzicha and azurite blue and mix it together and if 

you want to make a darker green, you must add more blue. Distemper it 

with gum water. 

[23] To make shades on the arzicha. Take saffron, tempered with white of egg. 

[24] To make gold ground. Take lead white and Armenian bole, some earwax 

and saffron. Temper it with white of egg a little bit of saffron and a bit of 

"bandoli"(?). Grind it with water. 

[25] To make letters in gold with the pen. Take lead white, and make eight parts 

of it, and take an amount of honey of one of these eight parts and grind it 

with white of egg and a pinch of saffron. 

[26] To lay bole on gesso. You want to mix it with glair and then grind it with 

clear water. 

[27] To make gold letters. Take a bit of bole and some burned chalk, about 

half the amount of the bole and take three or four times as much slaked 

gypsum and grind it with glair. 

[28] To make lines in gold. Take slaked gypsum, ground with glair in a bit of 

honey, a little bit of earwax, and a bit of glue. Let it dry. Then huff on it, lay 

the gold leaf on it and burnish it (25). 

[29] To make gold you can apply with the pen. If you want to prepare gold that 

you can work with the pen, take beaten gold leaves-Praise God. 

End. {15v} 

The following text is from folios 15v-20v, one page of which is shown in 

Figure 3: 

Wallert 41

[30] To make gum water for paints. Take of the white gum Arabic, the quantity 

you need and put it in a beaker. Next take clear water and put it in this 

beaker, so much that it covers the gum about two fingers wide. Place it in 

the sun. If there is no sun, put it near the fire and let it stand there for three 

or fo ur hours, until the gum dissolves. Then take it from the fire, and let it 

stand for two days, but {16r} make sure that you stir it every fo ur hours. 

Next sieve it through a cloth and cover it until you will work with it. This 

IS proven. 

[31] To make egg medium to temper the colors. Take as much white of egg as 

you want and put it in a very clean pot. Then take fo ur sponges bound 

together at the top and knead with both hands so that the white of eggs 

completely turns into such a froth that it will not even fall out if you turn 

the pot upside down. Let it stand from the morning until the evening until 

it settles and then put it in a bottle. And put some camphor in this bottle so 

that, not even after a year, will it go bad. 


MS 1793, fol. 17r. 

[32] To make a watercolor like gold. Take this previously described glair and put 

a little bit of saffron in it; and if you put in more yellow it will all be like 

gold. 

[33] To grind and clean vermilion for writing and miniature painting. Take as much 

vermilion as you think you need, grind it very fine with water on the porphyry 

stone. Then let it dry on the stone. Then grind it very fine fo r the 

second time. Next put it in a glass horn and temper it with the aforesaid 

glair and let it stand fo r three or fo ur hours until the vermilion has settled 

on the bottom. Then carefully decant the supernatant liquid and replace it 

fo r new glair. And keep on doing this for two or three times. Thereafter write 

with this on your paper. When you think it is not shiny enough, make the 

letters you want, large or small, let them dry and then lightly lay with the 

brush some of the previously described yellow on it and it will be very shiny 

(26). 

[34] To make green for writing and miniature painting. Take purple lilies and pick 

only the petals of these. Pound them and take of the juice. You should know 

that this juice is of a purple color. When you add a bit of roche alum, that 

is, to a beaker of juice about a chestnut of well pulverized alum, it will turn 

blue. Then take pieces of linen cloth and put them to soak very well in this 

juice. Then lay it [the cloth] on the grid [to dry], and then it must be soaked 

another time so that it will better take on the color. And in between you let 

it dry on the grid. 

[35] To make blue clothlet colors. Take certain flowers, that grow between the 

corn, which have five petals and have the color of a purple lily. Take the 

petals and rub these leaves in a piece of old linen. Let it dry and if you feel 

that the linen does not take enough color, repeat the treatment fo r a second 

time. You can also take the flowers that are called "fior di lebio" or "ibio," 

which also grow between the corn. Pick, when you begin, the flowers that 

have a sky blue color and have lots of petals. Pick the petals of these flowers 

and squeeze them bit-by-bit with your fingers. Next press them well together, 

take out the juice and then soak the linen in it and let it dry on the grid. 

{17v} Also: take the petals of the "fior di lino" and crush them and put the 

pieces of linen in this juice (27). Let them dry in the same way as described 

before and if the pieces of linen would not take up enough color the first 

time, dip them in again (28). 

[36] To make red clothlets fo r writing and miniature painting. Take the leaves of 

"bietoli," also called "gelosia," and let them dry a bit and then rub these leaves 

in a piece of linen cloth and if it does not have enough color for you, rub 

in more until you think it is right (29, 30). You should know that the leaves 

need to be dried because they have a certain humidity of water in them. 

Therefore, one lets them dry, which consumes this humidity, so that only the 

color remains. 

42 


[37] To make crude verzino for writing and miniature painting. Take the brazilwood 

and finely scrape of the quantity you think you need, with a piece of glass. 

Put these scrapings in a new glass. Add so much vinegar to it, that it covers 

the brazilwood half a fmger's width, and not more. The vinegar must be 

strong and clear. Add to the substance in this glass the quantity of about a 

bean of gum Arabic and the same amount of well-pulverized roche alum. 

Place it in the sun and make sure that the glass is well covered. Keep it this 

way for three or four days. If there is no sun, then keep it for four days so 

close to the fire that it gets lukewarm. When it is warmed this way, put it in 

shells as shown here, and let it stand in these shells, until it seems done to 

you. Know that the longer it remains, the better it is. Next put it in a bottle. 

{17v} Write with this on your work. If you want to make red clothlet colors 

of it, dip in to soak the pieces of cloth and let them dry in the previously 

described way, except that in the making of the clothIets [neither] the alum 

nor the gum Arabic should be used. 

[38] To make a purple clothlet for writing and miniature painting. Take the petals 

of the "papatello," which is a sort of wild poppy. Crush them and take out 

the juice. To one glass of juice it needs to have about the size of a bean of 

well-pulverized roche alum. Then dip in the piece of cloth, once or twice, 

until you think that it has taken enough of the color. 

[39] To make a beautiful blue at little cost. Take quicklime and green and ground 

verdigris and sal armoniac, as much of one as of the other. Grind these all 

together with urine and you will see a beautiful blue. Temper it with the 

previously described glair, when you want to work with it (31). 

[40] To make invisible letters, which you cannot see unless near the fire. Take sal 

armoniac, the amount of it is about the size of a chestnut to half a glass of 

water. Write with this water when it is dissolved, and if you want to read it 

you heat the parchment near the fire and the letters will appear as if they 

were written in ink. 

[41] To make black letters with every water you write. Take galls and vitriol, as 

much of one as of the other. Grind it and sieve it and put it on the parchment 

like a varnish coating. Next write on this with any water you want and while 

you are writing the letters will appear black. 

[42] To erase letters from goatskin parchment without a scraping iron. Take the juice 

of a lemon or of a strong orange and immerse {18r} a sponge in it. Rub 

with this sponge on the letters and they will be taken off as if they never had 

been written. 

[43] To make a ground fo r gold. Take slaked gypsum, as much as you need, and 

Armenian bole about one third of the gypsum, and of aloe one third of the 

bole. Then grind everything together very fine with water on the porphyry 

stone. Let it dry on the porphyry stone and then grind it again very fine. 

Add a little bit of candy sugar and when it is ground enough, gather it in a 

glazed horn and temper it with equal quantities of the aforesaid glair and 

water of animal glue, that is hide glue. Make this gold ground so liquid that 

it flows from the pen and write whatever you want. And if it appears not to 

have enough volume, write over it a second time, like before. Scrape it carefully 

with a scraping iron when it is dry, so that the letter will be plane and 

smooth. When you want to lay the gold, huff on the letter and immediately 

lay the leaf of gold or silver. Press on it with the calf's tooth. Then clean it 

with some cotton wool so that what sticks out from the letter will be wiped 

off. If you want to make a head of a frame with this gold ground with the 

brush like described, paint it over the first application and then scrape it and 

lay the gold. 

[44] To make French glue for many purposes. Take animal glue, i.e. bone glue, 

and put it to soak in so much water that it covers the glue. Put it to soak in 

the evening and put it on the fire in the morning. When you soak it, it does 

not matter {18v} if there is not enough water to cover it completely, because 

Wallert 43

when the lather covers it about the width of a finger it is good enough. 

When you have had it cooking fo r a while, take it from the fire and immediately 

add four or five hard and fresh pieces of chicken dung. Then you 

sieve it through a straining cloth, without wringing it. Then put as much as 

you need of it back in the water and use it as it pleases you. 

[45] To color and lay gold on oranges and other fruits. Take previously described 

glue and douse your hands in it. Rub with these washed hands the orange 

or any another apple very well. Let it dry. Then wet your hands again and 

rub this fruit again until it has taken on the glue well. Then give it the color 

you want it to take. But if you wish to be it in gold, be it as letters or in 

gold leaves, do as follows: Take armoniac, very well ground with urine on 

the porphyry stone. Distemper it with urine, but not so much that it flows 

from the pen, because the letters or the leaves want to be done with the 

brush over the colors with this armoniac. Let it dry. Next take the gold leaf 

and put it on the letters. But you should know that the letters first need to 

be warmed up with the knife. Then, when the gold is laid, immediately press 

lightly on it with cottonwool. Then rub with the cottonwool over the letters 

so that the gold that does not adhere to the letters is taken off. You can also 

give it beautiful, nice colors. 

[46] To make a tragacanth gum paste oj any color you want, to make reliifs. Take 

as much tragacanth gum as you want and put it in a glass or a dish, and add 

so much vinegar that it covers the tragacanth. But the vinegar you put in 

should be strong. Let it stand to soak fo r two days and two nights. Then take 

slaked gypsum which is dry. Next take lead white and triturate it very well 

with the tragacanth in a mortar. But put on the bottom of the mortar {19r} 

a bit of the gypsum so that the lead white and the tragacanth will not stick 

to the mortar. To prevent it from sticking to the pestle, put also some gypsum 

on that. Pound it well and it will become a paste as white as cottonwool. If 

you want the paste to become black, do as follows: Take a lit candle and over 

the flame you hold a piece of tinned iron so close that it touches the flame, 

whose smoke will deposit in a black spot. Take this deposit and rub it together 

with the paste. And the more you put into it the better it will be. Grind it 

with indigo if you want it to be blue, and if you put in little, it will be light 

blue, and it will be dark blue if you put much more indigo in it. If you want 

it to be red, put in vermilion which has been finely ground on the porphyry 

stone. The more you put in, the redder it will be. If you add little of it, it 

will be of a flesh tone. Put in a bit of well pulverized verdigris if you want 

.it to be green. If you want it to be yellow, put in some well ground orpiment. 

If you want it to be a purple color, put in a bit of lac. Know that the more 

of these colors you put in, the stronger colored it will be. The paste you thus 

make should be a little bit harder and you make it in such a way that you 

grind it again many times with the previously mentioned things. To make a 

good paste and to prevent it from sticking, either to the mortar, or to the 

pestle, knead a piece of it in your hand. When you want make it in a mold, 

take some of the soft paste that fits on the blade of a knife, or even less. But 

if you want to let it harden in a lead or clay mold, clad the form with some 

cotton so that the paste will not get stuck to the mold. Apply some warm 

bone glue in the areas where you want to do some modeling or make foliage. 

That will do it. 

[47] Papier mache to make reliifs. Take scraps of the paper cuttings from books 

and put them to boil in a kettle. When it has boiled, take the scraps out and 

pound them fine in a bronze {19v} mortar. Then let it boil a second time. 

Take the kettle from the fire once it has boiled and let it cool down. When 

it is cold enough to have your hands in it, take this paper pulp out. Squeeze 

the water out, make balls of it and let them dry. When you want to make 

reliefs with it, put these balls to soak in hot water during one or two days. 

When the water is cooled down enough to have your hands in it, break the 

balls and stir. It will become like a dough. This paste spreads smoothly. Apply 

44 


it on what you are working, with a scuilie, so that it spreads. Press everywhere 

on it with a sponge, as you work with the paper pulp without glue. 

[48] To make cork board fo r making reliif. Take good writing paper, which is 

smooth and with a little glue. Place them between washcloths. Always fold 

the cloths in four quarters, placing the paper in between. Put it in boiling 

water, making sure that it does not become brownish or stained by lye. Then 

take it out and let it dry on the grid. When it is a bit dry, you can work 

with it. It is best to work with water, so warm that you can just stand to have 

your hand in it. 

[49] To make letters, big or small, in gold. Take the milky juice of a fig tree 

and put as much as you think you need in a dish. Next take good black ink, 

which is rich in gum, and put so much of it in that the juice looks black. 

Write with it. When the letters are written, you huff on it and immediately 

lay the gold on it. Burnish with the tooth, and then rub the letters with 

cottonwool and where you have huffed on the ground, the gold will stick 

and the rest will be taken off (32). 

[50] Another way. Take Armenian bole and grind it with clear water very fine 

on the porphyry stone. Let it dry on the stone. Then grind it again with 

water and let it again dry on the stone. Then grind it again with the aforesaid 

milky juice. When ground, {20r} gather what you have in a horn of glass, 

or of glazed earthenware. Temper this so much that it flows from the pen. 

[51] To make secret writings. Take well pounded and sieved galls and put it to 

soak in water. Let it stand in the water until it settles to the bottom. Four 

glasses of this powder go to one glass of water. Write with this water whatever 

you want. Take, when you want to read it, vitriol and put it in water, with 

which you wet these letters. The letters, which you will be able to read, will 

now appear. 

[52] To make the blue clothlet. Take a plant which is called turnsole whose 

flowers and clustering berries you pick one-by-one. Then you put these in 

a mortar, according to the quantity you want to make of it. Crush them and 

take the juice out. Put this juice in a glazed bowl and then you take pieces 

of cloth that are white and coarse. Then you immerse them in the bowl [with 

the juice] and let them soak well in it. 

Do this three times, and every time that you that you take it out of this bowl, 

you let it dry well on a wooden bench. Before you reach this stage, you have 

to find a remote quiet place in the ground, and urinate much in it, six days 

before you come to make the said turnsole. Then, to place it above the 

humidity of this urine, take straw and make a fabric over the urine, and place 

the cloths on it. Let it stand fo r at least twenty days above the aforesaid urine 

to make it darker and more beautiful. 

The best turnsole is made in this way: First, take glair according to the quantity 

you want to make. You may wish to use amounts small enough that the 

said turns ole will be {20v} strongly colored. When you want to work with 

it, take the glair and put it in a shell or in a small glazed jar. Then take the 

turnsole and put some of it in, according to the right proportion and let it 

stand [in the glair] fo r an hour. Then press it out with your finger and the 

juice that comes out of the clothlet is the turnsole. Always prepare a little bit 

at a time because that is better. 

Notes 

1. Constraints of space in the present volume made it impossible to publish a true 

bilingual text edition (with a transcript of the original printed next to the translation). 

A copy of my transcript, however, is available upon request. 

2. Merrifield, M. P. 1849. Original Treatises, Dating from the XIIth to the XVIIlth 

Centuries on the Arts of Painting in Oil, Miniature, Mosaic, and on Glass; of Gilding, 

Dyeing, and the Preparation of Colours and Artificial Gems, Vol. II. London: John 

Murray, 325-547. References to the so-called Bolognese manuscript (Segreti per 

Wallert 45

Colori, Biblioteca Universitaria MS 1536) are given as Bo1. MS, followed by 

recipe number and page number in Merrifield's publication. 

3. Wallert, A. 1991. Kookboeken en Koorboeken, Techniek en productie van boekverluchting 

in koorboeken van het klooster Monte Oliveto Maggiore. Groningen: Groningse Universiteits 

Drukkerij. References to Ambruogio da Sienas Ricepte daffare pili colori, 

Siena Biblioteca Comunale MS 1.JI.19, fols. 99r-106r, are given as Siena 

MS I, followed by recipe number and page number in Wallert's edition. The 

indication Siena MS II with recipe and page number refers to the edition in the 

same publication of Bartolomeo da Sienas Experientie, Siena Biblioteca Comunale, 

MS L.XI.41, fols. 34v-39v. 

4. This recipe is an abbrieviated version of Siena MS II, XVIII, p. 45. A version in 

Latin of the same recipe is in Bo1. MS, 160, p. 467. 

5. Alexander, S. M. 1965. Medieval recipes describing the use of metals in manuscripts. 

Marsyas (XII) :34-51. 

6. An identical recipe is in Siena MS I, recipe XXIV, p. 35. An abbreviated version 

of this recipe is in Oxford (Bodleian Library, MS Conon. Misc. 128, fo1. 2v): "A 

mettere oro in carta." 

7. Wallert, A. 1990. Chrozophora tinctoria Juss.: Problems in identifying an illumination 

colorant. Restaurator (1 1):141-55. 

8. Identical recipe in Siena MS I, XVIII, p. 31. Similar descriptions of ingredients 

are given in Siena MS II, III, p. 37; Bo1. MS 389; and Bo1. MS 390, p. 597. Also 

see Zerdoun Bat-Jehouda, M. 1980. La fabrication des encres noires d'apres les 

textes, Codicologica 5; Les materiaux du livre manuscrit, Litterae textuales Leiden. 

9. Kuhn, H., 1961, Safran und dessen Nachweis durch Infrarotspektrographie in 

Malerei und Kunsthandwerk, Leitz-Mitteilungen fUr Wissenschaft und Technik, 

Bd 11, (1):24-28. See also Straub, R. E. 1965. Der Traktat "De Clarea" in der 

Burgerbibliothek Bern, Schweizerisches Institut ju r Kunstwissenschajt, jahresbericht 

(12):89-1 14. 

10. The same but more complete recipe can be found in Bo1. MS 129, p. 447. 

11. Identical recipe Siena MS I, XXIII, p. 35. This recipe recurs in a slightly more 

elaborate form in Oxford, Bodleian Library, MS Conon. Mise. 128, fo1. 21r and 

also in Bo1. MS, 77, p. 415, and Bo1. MS, 80, p. 417. 

12. Brachert, E and T. 1980. Zinnober. Maltechnik / Restauro (86):145-59. 

13. Siena MS II, XlV, p. 49.; Siena MS I, Xv, p. 29. 

14. Siena MS II, XII, p. 41. Also see Kurella, A., and I. Strauss. 1983. Lapis lazuli und 

naturliches Ultramarin. Maltechnik / Restauro (89):34-55. 

15. This recipe will not result in the promised blue pigment, but rather in the 

formation of the red mercuric sulphide. On the "azure-vermilion tangle," see 

Thompson, D. V 1956. The Materials and Techniques oj Medieval Painting. New 

York: Dover Reprints, 155. 

16. An identical recipe can be fo und in Siena MS I, XII, p. 29. Also similarities with 

Modena, Biblioteca Estense, MS ex.T.7.3., fo1. lOv. Also see Wallert, A. 1986. 

Verzino and Roseta Colours in 15th Century Italian Manuscripts. Maltechnik / 

Restauro (92):52-70. 

17. The recipe requires at this point the use of sulphur to be practically applicable. 

It is not clear what is meant by "cashina." 

18. Similar in Siena MS I, XXI, p. 32, and Siena MS II, XXI, p. 46. 

19. Probably a corruption of "porporino," a golden glistening tin (II) sulphide. 

20. The recipe calls for lead. Any useful pigment, however, can only be obtained if 

this lead were replaced with tin. 

21. The nitric salt, "sal nitro," should have been "sal armonic," i.e., al11110nium chloride. 

22. This recipe is slightly more intelligible than Simone's recipe 17. See Schiessl, U. 

1981. Musivgold, eine pigmentgeschichtliche Skizze. Maltechnik / Restauro (87): 

219-30. 

23. Florence, Biblioteca Nazionale, MS Magliabecchiana CL XV 8bis, fo1. 14v. More 

elaborate variations are in Siena MS I, III, p. 23, and in Florence, Biblioteca 

Laurenziana, MS Ashburnhamia 349, fols. 80v-81r. 

24. The weld herb of the dyers is Reseda luteola L. See Stoll, A. M. 1981. Gelbe 

Pflanzenlacke, Schuttgelb und Safran. Maltechnik / Restauro (87):73-1 11. 

25. An identical recipe can be found in Florence, Biblioteca Nazionale, MS Palatina 

916, fo1. 110r, "Affare profili d'oro brunito." A related recipe occurs in Bo1. MS, 

162, p. 469. 

26. This recipe occurs in a slighly more elaborate version in Siena MS II, VI, p. 39. 

27. Fior di lino is the cornflower, Centaurea cyanus L. 

46 


28. On these "clothlet colors," see Wallert, A. 1993. Natural organic colorants on 

medieval parchment: anthocyanins. ICOM Committee fo r Conservation Preprints, 

516-23; and Willhauk, N. 1981. Farbprobleme spatmittelalterlicher Buchmalerei. 

Restaurator (1-2):103-34. 

29. Bietola is a beet or garden beet; bietolina = weld (Reseda luteola) (?). 

30. A plant with the same name, gilosia (Amaranthus tricolor?), is mentioned in Bol. 

MS, 117, p. 439, for making a purple clothlet color. 

31. Gettens R. J., and E. West Fitzhugh. 1966. Azurite and blue verditer. Studies in 

Conservation (2): 54-61. 

32. Quite similar is Siena MS I, IV, p. 22. In Simone's manuscript, however, the 

vermilion prescribed by Ambruogio is replaced for a black ink. The recipe recurs 

in a rather abbreviated version in Florence, Biblioteca Laurenziana, MS Ashburnhamia 

349, fol. 84v. 

Wallert 47

Abstract 

Primary sources can be extremely 

valuable for the study of historical 

painting techniques, materials, and 

studio practice. Most authors of 

these treatises and manuals are either 

anonymous or little known. Quite 

often, therefore, no context for these 

books can be traced. In some cases, 

however, research on the author(s), if 

known, can reveal information that 

helps to interpret these sources, 

shedding light on the contents and 

their fu nction in studio practice. An 

example can be fo und in a seventeenth-century 

Italian treatise on 

miniature painting which, according 

to its title, was written by the miniaturist 

Valerio Mariani da Pesaro in 

1620. Research on this treatise has 

revealed some unexpected aspects of 

its authorship. 

A Seventeenth-Century Italian Treatise on Miniature 

Painting and Its Author(s) 

Erma Hermens 

Kunsthistorisch Instituut 

Rijksuniversiteit Leiden 

Doelensteeg 16 

Postbus 9515 

2300 RA Leiden 

The Netherlands 

The manuscripts 

The Leiden University Library owns a seventeenth-century Italian manuscript 

titled Della Miniatura, del Signor Valerio Mariani da Pesaro, Miniatore del serenissimo 

Signor Duca d' Urbino, del Signore Capitano Giorgio Maynwaringe, inglese, 

l'anno del Signore 1620, in Padova (sign. Voss. Ger. Gall. 5q). The manuscript, 

consisting of three sections, contains a treatise on miniature painting technique. 

The first section covers technical recipes for the preparation of pigments, 

dyestuffs and inks, binding media, and utensils such as pencils and 

brushes. The second section gives instructions on how to paint landscapes, 

and describes the mixtures of colors to use and the build-up in layers fo r the 

execution of different types of trees, weather conditions, water, sea, rocks, 

plants, mountains, villages, close-up and distant views, and so fo rth. The third 

and smallest section describes the mixtures of colors and the layering of colors 

used to depict skin and cloth in portrait painting. 

The recipes in the first section seem to be quite accurate, a result in part of 

the continuation of traditional methods, and in part of personal experiments 

by the writer. In general, one can say that the technique described is a watercolor 

technique fo r paper and parchment, in which the final effect is 

reached by using a mixture of pigments and dyestuffs or, as is more often and 

more typical fo r this treatise, by several layers of washes of colors. 

During the research on the treatise, two more manuscripts were found to be 

connected with the Leiden manuscript. One manuscript, owned by the Beinecke 

Library, Yale University (sign. MS 372), is entitled Della miniatura del 

Signore Valerio Mariani da Pesaro, miniatore del Duca d'Urbino con aggiunte d'altre 

cose per l'istessa professione dal Signor D. Antonello Bertozzi scrittore e miniatore in 

Padova, per me Francesco Manlio Romano, l'anno MDCXX. The second manuscript, 

entitled Ricordi di belli colon:, is part of the Urbinati Latini collection in 

the Biblioteca Apostolica Vaticana, Citta del Vaticano, Italy (sign. Urb. Lat. 

1280). The three manuscripts will be referred to here as the Leiden, Yale, and 

Rome manuscripts. 

The Yale manuscript, despite the differences in its title, is very similar to the 

Leiden manuscript, with only some minor variations in language. It contains, 

however, some extra pages at the end with additional recipes fo r pigments, 

paper dyeing, etching materials, as well as recipes fo r gunpowder, payments 

for books, a list of the quarters of Venice and other notes, some of them 

dated between 1614 and 1628. The extra recipes could be the "aggiunte 

d' altre cose per l'istessa professione dal Signor D. Antonello Bertozzi scrittore 

e miniatore in Padova." 

The Leiden manuscript contains a price list of pigments and dyestuffs and 

also a list of all materials needed by the miniatore. Both are lacking in the 

Yale manuscript (see Appendix doc. 1). 

The Rome manuscript covers only the second part of the Leiden and Yale 

manuscripts, that is, the part on landscape painting. We find here many differences 

in language and also in the sequence and length of the chapters. The 

Landscape section in the Leiden and Yale manuscripts is more extensive, as 

48 


it also includes instructions fo r composition, the build-up of a miniature from 

a pencil drawing up to the final execution in paint, and so on. Some chapters 

are a combination of two or three short chapters from the Rome manuscript, 

covering more or less the same subjects. It seems that the writer of the treatise 

rearranged and expanded the text from the Rome manuscript to improve the 

contents (1). 

Valerio Mariani da Pesaro 

The miniaturist Valerio Mariani da Pesaro-the writer of the treatise, according 

to its title-is mentioned by Thieme-Becker as being active from 

1560 to 1600, and by Zani as living from 1565 to 1611, but further information 

on the treatise and its author seemed difficult to fmd (2). Archival 

research in Italy however proved otherwise, and the treatise can now be provided 

with an ample context. 

Mariani is described in Lancellotti's L'Hoggidi (1636) as a pupil of Giovanni 

Maria Boduino, a miniaturist who worked in Friuli (3). In a document from 

1582, Mariani is mentioned as the twelve-year-old pupil of Boduino, establishing 

Mariani's date of birth as 1570 (4). About Mariani's working environment 

after his training in Friuli, little information has been fo und so far. He 

worked, as he mentions in the treatise (Chapter XVI: Giallo Santo), for the 

Duke of Savoy: "Giallo santo is a colour which is extracted from the flower 

of the broom, as a painter in Borgo in Brescia taught me, while I stayed in 

that place in the service of the honourable memory of Duke Emanuele Filiberto, 

Duke of Savoy ..." (5). And, as can be read in the title of the treatise, 

he was also employed by the Duke of Urbino. According to the date of the 

treatise, this must have been Francesco Maria II della Rovere. The last known 

dates in Mariani's chronology are 1618, when he is mentioned as heir in his 

brother's will, and 1625, when he is mentioned as a debitore in the duke's 

bookkeeping (6). 

So far no documents on his work for the Savoy court or other employers 

have been fo und; therefore, we concentrate on Mariani's activities fo r the 

Urbino court, which provides us, as will become clear, with the context in 

which the treatise was written. 

Francesco Maria II della Rovere began his reign in 1574, after his father 

Guidobaldo II della Rovere died, leaving him with a bankrupt state. The 

young duke, who had a strict and sober upbringing at the court of Philip II, 

started a period of severe economizing, including trimming the court's cultural 

expenses. It was 1580 before the duke regained interest and renewed 

his financial backing fo r cultural and artistic activities. Many documents, as 

well as artworks he commissioned, survive the period of his reign, creating a 

clear image of the cultural policy and interests of the duke (7). 

As a contemporary chronicle states, "When Francesco Maria, the last Duke, 

had paid the large debts of his father Guidobaldo by stopping the many arts 

and famous crafts that were executed here, he changed his mind after seeing 

the severe consequences and ordered the fo undation of several workshops at 

his court where very excellent masters in every art and profession were put 

to work" (8). One such master was the miniaturist Valerio Mariani da Pesaro. 

These workshops fo rmed a well-organized business, providing artworks "to 

order" for the duke. As not only masters but apprentices were present, the 

workshops seem to provide the appropriate milieu for writing a technical 

manual. A sketch of the workshops will, therefore, be given. 

The workshops 

The duke made personal notes during the years 1580-1620 of all his "artistic 

expenses," including monthly payments to the botteghini (workshops) (9). He 

employed painters, sculptors, watchmakers, ebony workers, bookbinders, goldsmiths, 

and miniaturists in a lively and businesslike organization (10). 

Hermens 

49

The actual working areas of the botteghini were situated on the ground floor 

in the left wing of the ducal palace in Pesaro. There seems to have been an 

internal entrance from the palace (11). The botteghini also opened onto the 

street like small shops; indeed, the artists were permitted by the duke to accept 

orders from other clients. Some artists had their living quarters above the 

workshops; others who had families received money to rent housing elsewhere 

in Pesaro. They also received provisions, candles, and firewood, as can 

be read in a ledger (12). 

A supervisor was appointed with responsibility for the daily organization, 

provisions, and the supply of materials. Bills with lists of ordered materials, 

including pigments and brushes, are among the many surviving documents 

concerning the workshops (see Appendix doc. 2). The duke's commissions 

were not given by the duke personally but passed on by intermediaries, mostly 

noblemen employed as his secretaries or persons suitable because of their 

profession. These intermediaries also acted as the duke's representatives in the 

search for artists and artisans for the workshops. For example, the duke's ambassadors 

were ordered to search in Rome, Venice, and Florence fo r artists of 

the highest quality. The proposed artists had to send proof of their capacities 

which, according to the documents, was quite often rejected by the duke. 

In 1581 the duke asked his ambassador in Rome, Baldo Falcucci, to find him 

a miniaturist. From 1582 on, a "miniatore" is mentioned in the lists of workshop 

employees (13). In many instances, the name of the artist is not given 

but a "maestro Valerio miniatore" appears from 1603 to 1605. Although employee 

lists from 1605 on are missing, Mariani's activities at the court of 

Urbino most likely continued. 

Figure 1. Detail from the Battle of San Fabiano, 

illustrated in Plate 7a, showing the 

center scene where the sketchy underdrawing is 

clearly visible. Photograph by E. Buzzegoli, 

LAboratorio di Restauro. Courtesy of the UjJ 

izi Gallery, Florence. 

Mariani's activities can be traced not only in documents but also in many 

miniatures. The Biblioteca Apostolica Vaticana has three illustrated manuscripts 

attributed to Mariani, in addition to a part of the Purgatorio and all 

the miniatures of the famous Dante Urbinate's Paradiso (14). The Galleria 

Palatina in Florence owns two series of miniatures, framed in black ebony 

frames, with scenes from the lives of Christ and Mary, miracles, and martyred 

saints. The Uffizi owns the only miniature with Mariani's signature, a battle 

scene from the life of Federico da Montefeltro (15) (Plate 7a). This miniature 

is of great interest as we can compare the techniques used with those as 

described in the treatise, thus comparing practice with theory. The miniature 

is kept in a dark cabinet; the colors are, therefore, in extremely good condition. 

A study of the miniature using microscopy made clear that the standard 

technique of washes was used. Specific comparisons (i.e., the build-up of the 

trees, background with the village, bluish mountains and sky, etc.) with chapters 

in the landscape section of the treatise, show that Mariani is following 

the instructions given there. The miniature has a very sketchy underdrawing-which 

can be seen with the naked eye but is even more clearly apparent 

with an infrared camera-that has been traced in some places with a brownish 

ink (Fig. 1) (16). A blue-black ink is also used to emphasize contours and 

shadows. On the underdrawing, the paint is most often added in transparent 

layers using the techniques described in the treatise. 

Mariani was known for his technical abilities; Lancellotti described him as a 

pupil of Boduino who "surpassed his master in patience, and his miniatures 

were owned by the most important princes in the world." They both "kept 

a secret of how to grind gold in the Persian way for miniature painting and 

writing, that stayed stable as the antique [gold] " (17). 

Although the preceding paragraphs comprise a short resume of the research 

results so far, it is clear that we are considering a treatise that can be placed 

in an interesting context, offering a unique possibility of comparing the theory 

of the treatise with the practice as laid down in Mariani's work. The 

context of the workshops, and Mariani's obvious technical skills, make it plausible 

that the treatise was written as a manual to be used in the workshop. 

50 


Ricordi di belli colori 

Initially it seemed reasonable to assume that the Rome manuscript was writter. 

Sy Mariani as a draft version of the landscape section. The manuscript 

wa, then reworked and expanded for the larger treatise of which, presumably, 

both the Leiden and Yale manuscripts are copies. However, there are some 

remarkable features that made us continue the search fo r the origin of the 

Rome manuscript, leading to interesting results. 

The manuscript's text covers most of the chapters on specific landscape elements 

that are present in the Leiden and Yale manuscripts, but lacks the 

chapters on composition and underdrawing. There are also considerable differences 

in the order and composition of the chapters and in the language, 

as well as some repetitions and crossing out of text. First, a comparison with 

other Italian treatises on painting technique from the sixteenth and early 

seventeenth century shows that no specific attention is given to landscape 

painting, especially not in such great detail as we see in the Rome manuscript. 

Landscape painting of that period was simply a Flemish specialty. It reached 

its peak after Pieter Breughel the Elder introduced the new spatial concept 

of several planes leading to a distant vista in which the religious theme became 

a detail instead of the main, foreground scene. The landscapes produced by 

the many Flemish artists working in Italy were enormously popular, and 

landscape print series circulated widely (18). The Rome treatise seems to 

reflect a Flemish sense of detail concerning landscape elements, an idea supported 

by the presence of Flemish painter Michiel Gast's name in a chapter 

on how to paint villages in the background. Gast's methods are found to be 

exemplary. We know little about Michiel Gast today except that he came to 

Rome as a pupil of Lorenzo of Rotterdam and was known for his paintings 

of ruins (19). The citation of Gast's methods and the detailed attention to 

landscape elements suggests a strong Flemish influence on the author of Ricordi 

di belli colori. 

Another remarkable feature of the Rome manuscript are the chapters on 

flowers and plants. Although the plants described are quite common in Italy, 

the names used are often botanical ones, probably most familiar to a person 

with a special botanical interest. In these chapters some personal remarks are 

present. For example, the writer says, "I have made a yellow flower in my 

book with herbs, the biggest one without landscapes" and "to represent well 

the fruit of the Jusaina or otherwise nocella as that is how it is called in the 

village of Rocca Contrada." It is striking that all such personal remarks made 

by the author are omitted in the landscape section in the Leiden and Yale 

manuscripts, while personal remarks are retained in the sections of those 

manuscripts containing recipes for pigments and dyestuffs . 

A reading of the manuscript suggests that the writer lived in Rocca Contrada, 

painted landscapes, came from a Flemish background (or was strongly influenced 

by Flemish landscape painting), had botanical interests, and probably 

illustrated a herbarium. Knowing this, should we still consider Valerio Mariani 

da Pesaro as the writer of the Rome manuscript or should we look elsewhere? 

It was in fact the citation of Michie! Gast's name which helped us to answer 

this question. 

Gherardo Cibo 

One of the very few known works by Gast can be found in the collection 

of Gherardo Cibo, an Italian nobleman-artist-botanist (1512-1600). Cibo 

lived for most of his life in the small village of Rocca Contrada, now called 

Arcevio, in the Marche, the region of the Duchy of Urbino. Cibo started an 

ecclesiastic career in Rome but he did not pursue it, settling in 1539 in Rocca 

Contrada where he occupied himself with artistic and botanical activities. 

Although in obscurity until the 1980s, research has revealed that Cibo was 

responsible fo r many landscape drawings in major collections that had been 

anonymous or wrongly attributed, often to Flemish artists (Fig. 2). A 1989 

Hermens 51

Figure 2. Gherardo Cibo, Landscape. Courtesy of the Printrooltl if the University of Leiden, the 

Netherlands (il1v. nr. A W 441). 

exhibition and catalogue dedicated to Gherardo Cibo shed much light on 

the artist's career and personality (20). A large group of landscape drawings, 

several herbaria, and illustrated editions of Mattioli's translated version of 

Dioscorides can now be attributed to Cibo (Plate 7b, Fig. 3). According to 

contemporary testimonies, Cibo must have been an interesting personality 

who came from a wealthy family. Praised by the local population for his 

virtues and charity work, Cibo occupied himself with botanical expeditions 

and painting. A local seventeenth-century historian alludes to Cibo's high 

artistic qualities and his method of preparing colors by extracting the dyestuffs 

from herbs, fruits, and seeds (21). 

Figure 3. Gherardo Cibo, "Hell/ionite, " 

from Herbarium (MS ADD 22332), 

fol. 143r, ca. 1570. Courtesy of the British 

Library. 

Could Gherardo Cibo be the writer of Ricordi di belli colori? Proof was to 

be found by comparing Cibo's works with the Rome manuscript. In the 

manuscript we not only find instructions on the mixtures of colors and the 

build-up of transparent washes of paint to achieve certain effects but also 

instructions specifically directed toward working on pen drawings or prints 

(i.e., Chapter XXIX: "Beautiful green to use on prints and on plants drawn 

with pen"; Chapter XXXVI: "To colour herbs, printed or drawn with pen"). 

Cibo's works show that he did both. Many of his landscape drawings were 

executed in pen and ink and then colored with transparently applied colors. 

He also colored prints in several editions of Mattioli's translation of Dioscorides 

and added landscape backgrounds as a personal touch. 

Additional proof can be found in the text. The personal remarks cited above 

concerning the chapters on plants point to the herbaria illustrated by Cibo. 

Two of these drawings are kept in the British Library (22). The largest herbarium 

illustrated contains only plants, drawn in ink and colored with watercolor, 

without landscapes in the background: " . .. my largest book on herbs 

without landscapes." The other smaller herbarium does have landscapes; on 

folio 183v, one reads, "Fusaina salva[ti]ca, nocella qui chiamata a Roccha 

C[ontrada]," which corresponds fully with the first lines of Chapter XLV of 

the manuscript, in which the instructions on how to paint this flower are 

given. The author describes several color mixtures for the Jusaina or nocella 

flowers and folio 184v in the herbarium can be fo und many color samples 

and some fusaina flowers where these mixtures are apparently tested (Figs. 4, 

5). 

Finally, examination reveals that the Rome manuscript is written in two 

hands. One author wrote only two pages; the script on these pages can be 

52 


Figure 4. Gherardo Cibo, "Fusaina [. .. J 

nocella qui chiamata a Roccha C [ontrada J, " 

Herbarium, (MS ADD 22332), Jo l. 

183v. Courtesy of the British Library. 

Figure 5. Gherardo Cibo, several proofs oj 

colors and "Fusaina" flowers, Jrom Herbarium 

(MS ADD 22332), 101. 184v. Courtesy 

of the British Library. 

identified as Cibo's handwriting. The conclusion must be that Cibo-not 

Mariani-was the author of the Rome manuscript. 

Gherardo Cibo and Valerio Mariani da Pesaro 

Although Gherardo Cibo's part in the origin of the treatise is established, we 

may still assume that Valerio Mariani composed the treatise, using Cibo's 

specific knowledge and thus incorporating the section on landscape painting. 

There is no reason, thus far, not to assume that Mariani wrote the first section 

with technical recipes, especially as some personal notes present cannot be 

attributed to Cibo, pointing to a professional miniaturist who was working 

on a commission basis. 

Did Cibo and Mariani meet? Cibo died in 1600 when Mariani was thirty 

years old. His name appears in the employee lists for the first time in 1603, 

which does not indicate he was not working in Pesaro before that date, as 

most of the time the workshop employees are only indicated by their function 

(i.e., miniatore) and the duke had started his search for capable miniaturists 

in 1581. If they did meet, Cibo must have been in his late seventies or early 

eighties but still active, as a letter he wrote to the Duke Francesco Maria II 

della Rovere indicates. In the letter, dated 1580, Cibo tells the duke he is 

very honored by his request to illustrate an edition of Mattioli's Dioscorides; 

Cibo finished the work at nearly seventy years of age (23). This letter, however, 

also indicates a clear contact between Cibo and the duke, but there were 

apparently more contacts with the Urbino court. 

In the Biblioteca Comunale in Jesi (Marche, Italy), an album is kept with 

landscape drawings, mainly by Cibo. The text on the cover of the album is 

in Cibo's handwriting and says that the album contains "a little landscape on 

paper from the hand of the Flemish painter who serves our illustrious Duke 

of Urbino, which Sir Cavaliere Ardoino sent me, in April I think. 1591. And 

he names himself M[aestro] Giovanne. There are here two drawings of 

M[aestro] giovanne fiame[n]go from landscapes on coloured paper ...." It 

also mentions drawings of "the Painter from ForB" (24). 

Cavaliere (knight) Ardoino can be identified as Girolamo Ardovino (also Ardoino 

or Arduini), the duke's architect (25). In many documents, he figures 

Hermens 53

as an intermediary between the duke and the workshops. Maestro Giovanne 

Fiamengo and the Pittore da ForIi both appear in the many documents concerning 

the workshops as being employed by the duke. Cibo was obviously 

acquainted with Ardovino and, therefore, must have been familiar with the 

workshops and the artists working there. He might have known the young 

Mariani if the latter did arrive in Pesaro before 1600. In any case, he probably 

knew the miniaturists' workshop. 

The connection 

As mentioned earlier, the Rome manuscript is written in two hands. The 

two pages in Cibo's handwriting comprise two chapters, one chapter on the 

mixture for meadows and fields (with two color samples in the margin), and 

one chapter with more general instructions on mixtures of colors (see Appendix 

doc. 3). At the end of the latter chapter, the following is written: " . .. 

such that if you Sir, will try to be a little less lazy and exercise more often, 

the exercises will work out very well." This last sentence makes clear that the 

manuscript was intended as a manual for someone who wanted to learn, or 

was ordered to learn, to paint and draw landscapes like Cibo did, that is, in 

the Flemish way. Although the manuscript is part of the Urbinati Latini collection 

of the Vatican Library, it seems unlikely that it was written fo r the 

duke himself, as the final remark is not very suitable fo r addressing a duke. It 

is reasonable to assume, however, that the duke, obviously impressed by Cibo's 

artistic capabilities, asked him to write down his techniques for the execution 

of landscapes in the Flemish manner in a manual that could be used by the 

duke's miniaturists' workshop. The different handwriting, corrections, and 

repetitions may indicate that Cibo dictated the text to someone, except for 

the chapter in which he addresses the person for whom the manuscript was 

meant. This person might have been the young Mariani. If not, then Mariani 

found the little book on landscape painting, reworked it, and included it in 

his treatise. The pigments and dyestuffs that appear in the landscape sectionsome 

of them rather uncommon, such as giallo de' vasari and bruno d'Inghilterra--are 

all described in Mariani's recipes (in the first section of the Leiden 

and Yale manuscripts) . Cibo really did experiment with extracting colors from 

plants and fruits, as well as with mixtures of pigments, as similar proofs of the 

color samples described previously can be found throughout all Cibo's works, 

even in the margin of the text of the Rome manuscript. These ricordi (reminders) 

may have inspired Mariani to describe exactly those pigments and 

dyestuffs Cibo used. Interesting is the use of the name giallo de' vasari (potter's 

yellow) for lead-antimone yellow. In the Marche, the most important and 

famous majolica industry of Italy was flourishing in the sixteenth century. 

Cibo presumably obtained this pigment directly from the potters. In the recipe 

for bruno de Inghilterra, Mariani says the color was known to spetiali et 

pittori (pharmacists and painters) but that so far he had not been able to 

discover the exact composition. Most of the other recipes are traditional, 

except for those using plants and fruits to extract dyestuffs; the latter recipes 

may also come from Cibo. The extra chapters on composition and underdrawing, 

although not present in Ricordi di belli colori, seem to reflect Cibo's 

techniques when compared with his work. 

Although no definite proof can be given, a personal exchange of information 

between Mariani and Cibo cannot be excluded. 

A comparison between practice and theory is possible, as ample material can 

be researched. Not only Mariani's miniatures, especially the signed one, but 

also Cibo's many landscape drawings and the herbaria and illustrated Dioscorides 

editions, provide this opportunity. It is clear that Mariani used the 

same method of building up transparent layers in the background landscape 

of his signed miniature as is described in Cibo's instructions (Plate 8). Mariani's 

technique, although clearly guided by a personal interpretation, is very 

similar to that used in Cibo's landscape drawings executed in ink and colored 

with transparent watercolors. 

54 


The treatise is typical for the late sixteenth and early seventeenth century in 

describing the techniques for the execution of two extremely important genres 

of painting in that period: landscapes and miniature portraits. Both authors 

provide us with valuable information on the materials used, the techniques, 

and the studio practice, as well as the setting and the possible cooperation 

that provided the inspiration to write the treatise. 

Appendix 

Document 1. List of materials needed by the miniaturist, most described in 

the recipes in the first section of the treatise (Leiden, University Library, Voss. 

Ger. Gall. Sq. c.2v.): 

Biacca, Cinabro, Minio, Biadetto chiaro e scuro, Azurro oltramarino, 

Azurro grosso, Giallolino f ino e sottile, Giallo santo, Terra negra, Terra 

rossa, Terra gialla, Terra d'ombra, Acqua verde, Tomasole, Terra gialla 

brugnata, Lacca di grana, Lacca di verzino, Verde di giglio paonazzo, 

Fuligine, Anaraccio, Negro Ju me, Negro d'osso di persico, Terra di campana, 

Orpimento, Terra verde 0 verde porro, Verd'azzurro e verdetto, 

Indico, ZaJfarano, Verde rame orientale, Bol armenio 0 ocria lemnia, Sandracca, 

Bruno d'Inghilterra 0 color di sale, Sangue di drago, Acqua gialla 

di spino cavino, Marchesita arsa, Giallo de' vasari, Ocria, Ocria abrugiata, 

Acqua di verzino, Carbone, Acqua e aceto gommato e calcinato, Inchiostro, 

Liscia, Lapis rosso e negro, Piombino per disegnare. 

Document 2. List of materials ordered by the supervisor Felice Antonio di 

Letiero for the workshops (Archivio di Stato, Florence, Fondo Urbino, Appendice, 

Filza 54, c.n.n.): 

A dl 15 9bre 1607. [ ... ] In per ordine di Maestro Felice Antonio di 

Letiero a di 24 detto per Ii botteghini di S[ua] A{ltezza]: Libra 2 112 

Azurro di Spagna fino [ ... ]; Biadetto libre 4 [ ... ]; Verdetto 3 112 [ ... ] 

Giallolino di Fiandra libre 4 1/2 [ ... ]; Verdetto libre 6 112 [ ... ]; Azurro 

di Spagna chiaro libre 4 oncie 3 [ . .. ]; Lacca di grana fina libre 2 [ ... ]; 

Smaltino di Fiandra libre 10 1/2 [ ... ]; Sma/tine diverse con cor po libra 1 

oncie 9 [ ... ]; Penelli di sedoli con asta [ ... ]; Penelli grossi di vano [ ... ]; 

Penelli mezani [ ... ]; Penelli piccoli [ ... ]; Cocciole di madre perla no. l00 

[ ... ]; Jacchino per portare detti robbe [ ... ]. A di 14 di Xbre 1607. 10 Felice 

Antonio di Lettiero ho ricevuto Ii sopradetti robbe per servitio di botteghini 

di Sua Altezza. Umilmente Felice Antonio. 

Document 3. Rome, Biblioteca Apostolica Vaticana, Urb. Lat. 1280, c.18v- 

19r. 

Avertimenti sopra Ie mestiche. 

Bisogna di avertire che tutte Ie sorte de mach ie, tanto per arbori quanta 

per sassi e per greppi et prati, che e necessario di Ja rle piu oscure, 0 piu 

chiare secondo la qualita del paese, anchora che tirino piu al verde, 0 piu 

al giallo, 0 piu all' azzurro, 0 piu al rossigno, 0 piu al negro secondo che 

piu piacera, 0 secondo che ricercara il disegno. La prima coperta non vol 

mai essere troppo chiara per potere dar poi i lumi secondo il bisogno, pero 

sempre sara bene mesticare con la prima machia quei colori che haveranno 

da s f oleggiare, poi illuminarle secondo si conviene e che satiffacci all'occhio 

et tanto si havera da Ja re per prati, campi et sassi quanto per Ii arbori. 

Non si po Jar questo senza un po' de Ja tica et perdimento de' colori. Pero 

chi vorra Ja re che Ie cose passino secondo il dovere, bisognara che componendo 

Ie mestiche vadi provando 0 sopra carte, 0 sopra tele, secondo dove 

si havera da des igna re, et haver tanto di pazienza che si asciughi et parendole 

poi 0 troppo chiara 0 troppo scura, 0 troppo negra, 0 troppo azurra, 

o troppo gialla, 0 troppo rossegiante, potra sempre rimediare con giongere 

piu di quel colore che sara necessario. Et sempre e bene Jar tanta mesticha 

che piu presto possi avanzare che manchare, perche non sola la persona se 

ne potra servire ad ogni hora per quel disegno, ma perche mancando havera 

grandissima Ja tica de poterla rifare come Jece la prima volta. Questo e piu 

Hermens 55

necessario di fa re che di haver 10 esempio delle machie perche si potra fa re 

qui una machia che volendola usar simile 0 sara gran fatica farla dello 

istesso colore, avera sara 0 troppo chiara 0 troppo scura 0 de altra qualita 

che non fara a proposito per il disegno. Si che Vostro Signore se risolvi de' 

spigrirsi un poco et de esercitarse pitA spesso et cosi I esercitio passara benissimo. 


This research has been carried out with the financial support of The Netherlands 

Organization fo r Scientific research (NWO). I especially want to thank Tim Bedford 

for a critical reading and correction of the manuscript and fo r his encouragement. 

Notes 

1. From the results of linguistic research of the Leiden and Yale manuscripts, we 

may assume that both manuscripts are copies from the same, presumably original, 

manuscript. The handwriting as well as the paper quality and watermarks of the 

Leiden manuscript seem to point at an early date, possibly early or mid-seventeenth 

century. The Yale manuscript contains extra notes, some of them dated 

between 1614 and 1628, suggesting the early 1620s as the date of the copy. A 

complete edition of the treatise, with technical and art historical comments by 

the present author, is forthcoming. 

2. Thieme, U, and F. Becker. 1930. Algemeines Lexicon der hildende Kunstler von der 

Antike his der Gegenwart, XXXI, Leipzig, 64-65. See also Zani, P. 1823. Enciclopedia 

metodico-critico-ragionata delle helle arti, part I, Vo!' XIII. Parma, 44, 457, note 28. 

3. Lancellotti, S. 1636. L'Hoggid£. Venezia, 310. 

4. Zuccolo Padrono, G. M. 1969. Miniature manieristiche nelle comn"lissioni dogali 

del 110 cinquecento presso il Museo Correr. Bolletino dei Musei Civici Veneziani 

72 (XIV):6, 16, note 6. 

5. "II giallo santo e un colore che si cava dal fiore della ginestrella, come m'insegno 

un pittore a borgo in Brescia, mentre stava in qual luogo nel servitio del serenissima 

memoria del Duca Emanuele Filiberto di Savoia ..." Leiden: University 

Library (Voss. Germ. Gall. 5q, fo!' 16r) 

6. For the will, see Pesaro, Biblioteca Oliveriana (MS 455), Spogli Almerici, fo!' 

338v. For Mariani's debt, see Florence, Archivio di Stato, Fondo Urbino, Classe 

III, Filza XXIV, fo!' 242v. 

7. Most documents are kept in the Archivio di Stato, Florence and the Biblioteca 

Oliveriana in Pesaro. See also Gronau, G. 1935. Documenti Artistici Urhinati. Firenze. 

8. Biblioteca Oliveriana, Pesaro (MS Olivo 1009) Cavaliere Domenico Bonamini, 

Biografie degli Uomini Illustri Pesaresi. fo!' 290: ". .. l'ultimo Duca Francesco Maria 

che avendo prima risecate la grossa spesa del Duca Guidobaldo suo padre col 

far cessar tanti arti e celebri manifatte che qui si professavano, che vedendone in 

seguito il gravissimo derivo, si ricredette e comando che si esigessero sotto la di 

lui corte, varie ollicine, aile quali deputo eccellentissimi maestri in ogni professione 

ed arti." 

9. These lists can be fo und in the Archivio di Stato, Florence, Fondo Urbino, Classe 

III, Filza XXIII (1581-1610 and 1613-1620), Classe I, Div. G., Filza CVI (161 1- 

1612). 

10. Archivio di Stato, Florence, Fondo Urbino, Classe III, Filza XXIII, fo!' 474r: 

"Lista della famiglia e salariati di S.A.S. e provisioni del'anno 1603: Maestro 

Propertio Grilli orologgiero, bocche 1; Maestro Cesare Maggiero pittore, bocche 

2; Maestro Giovanni Fiamengo, bocche 2; Maestro Simuntio miniatore, bocche 

2; Maestro Valerio miniatore, bocche 2; Maestro Nicolo argentiero, bocche 2; 

Maestro Boniforte orefici, bocche 2; Maestro Jacopo Cossa lavora d'ebbano, bocche 

2; Maestro Gervasio libraro ...." The name of the artist and the number of 

people working in the workshop are mentioned: the artist, plus an assistant or 

apprentice, equals hocche 2 (two mouths). 

11. Archivio di Stato, Florence, Fondo Urbino, Appendice, filza 48, not numbered. 

In this document the decoration of some rooms on the first floor of the palace 

is described: "Dalla porta che entra nella galleria nuova alia porta che va aile 

botteghe ...." 

12. Archivio di Stato, Florence, Fondo Urbino, Appendice, Filza 32: fo!' 13r a.f. 

13. The letters from Falcucci to the duke concerning the search fo r a miniaturist 

can be fo und in the Archivio di Stato, Florence, Fondo Urbino, Classe I, Div. G, 

56 


Filza CXLIII, fols. 1122r, 1125v, 1129v, 1133r; and Biblioteca Oliveriana, Pesaro 

(MS 375), vol. XXIV, fols. 227v, 212r, 207r, 179r. On the miniaturists, see also 

Meloni Trkulja, S. 1981. I miniatori di Francesco Maria II della Rovere. In 

Omaggio ai Della Rovere, ed. P Dal Poggetto. Pesaro. 

14. Michelini-Tocci, L. 1965. II Dante Urbinate della Biblioteca Vaticana. Facsimile. 

Milano. 

15. Meloni Trkulja, op. cit. (note 13). From the same author: 1980. Firenze e la 

Toscana dei Medici nell'europa del Cinquecento, vol. Palazzo Vecchio: commitenza e 

collezionismo medicei, Florence: 200-205. See fo r more information on Mariani 

and his miniatures: Hermens, E. 1993. Valerio Mariani da Pesaro: a seventeenthcentury 

Italian miniaturist and his treatise. Miniatura (3/4):80-93. 

16. See Hermens, E. 1993. Valerio Mariani da Pesaro and his treatise: theory and 

practice. Technologia Artis (3):109-12. The infrared research was executed by E. 

Buzzegoli and D. Kunzelman, Laboratorio di Restauro, Uffizi, Florence. 

17. Lancellotti, op. cit., 310. "Valerio Mariani di Urbino fu suo discepolo e avanzo 

il maestro di patienza e delle sue miniature hebbero i maggiori Prencipi del 

mondo .... Havevano un degreto di macinar l' oro alia Persiana per miniare e 

per iscrivere, che stava saldo come I'antico." 

18. Tongiorgi Tomasi, L. 1989. Gherardo Cibo: Visions of landscape and the botanical 

sciences in a sixteenth-century artist. Journal oj Carden History 9 (4) :208. 

19. Thieme and Becker, op. cit., XIII, 240. See also Bertolotti, A. 1880. Artisti Belgi 

ed Olandesi, Roma, 41-4. See also Hoogewerff, G. 1. 1912. Nederlandsche schilders 

in Italie in de XIVe eeuw, Utrecht, 142-43. 

20. Nesselrath, A., 1989. Gherardo Cibo: "qui non e cognito." In cherardo Cibo alias 

"Ulisse Severino da Cingoli, " ed. A. Nesselrath. Comune di San Severino Marche, 

5-36. 

21. Celani, E. 1902. Sopra un erbario di Gherardo Cibo conservato nella R. Biblioteca 

Angelica di Roma. Malphigia XVI:I0. 

22. British Library, London, (MS Add 22333 and 22332). See Tongiorgi, op. cit. 

23. Cibo's letter can be fo und in the Archivio di Stato, Florence, Fondo Urbino, 

Classe I, div.G, Filza CCLX, fol. 6r. For the edition of Diose or ides, see Nesselrath, 

op. cit., catalogue no. 20, 100. Roma, Biblioteca Alessandrina (inv. Rari 278). 

24. Idem, catalogue no.40: 123-27. Jesi, Biblioteca Communale, Album A: "Qui dentro 

sta un paesetto in carta reale, de mano del pittor[e] Fiamengo ch[e] serve il 

n[ostro] Duca Sereniss[im]o di Urbino: che mi mano il S[ignor]e Cavalier Ardoino. 

Di Aprile mi pare. 1591. E chiamase M Giovanne." He also mentions 

drawings "quelli de mano d[e]l pittor[e] da ForE ...." 

25. Thieme and Becker, op. cit., II, 78, under the name Girolamo Arduini. See also 

Zani op.cit. (note 3), II, 181, 320, note 138. Arduini is said to have written a 

treatise titled Trattato del modo di piantare e Jortificare una cittCt. 

Hermens 57

Abstract 

The problems of interpretation of 

written sources on painting technique 

are well known. Through loss 

of the technical tradition, within 

which details of information were 

well understood at the time of writing, 

technical information is obscured 

fo r later generations. In 

courses on historical techniques of 

painting at the School of Conservation 

in Copenhagen, attempts at reconstructing 

the kind of gesso 

ground used in early Italian painting 

have prompted investigation into the 

actual meaning of the "giesso uolteriano" 

mentioned by Cennino Cennini 

in his treatise. This paper examines 

the problem from three angles: 

(1) the possible meanings of Cennini's 

text on this point; (2) the preparation 

of gesso grounds from the 

possible fo rms of gesso resulting 

from the first point (dihydrate, hemihydrate, 

anhydrite); and (3) technical 

evaluation of the reconstructions and 

comparison with the results of scientific 

examination of grounds in early 

Italian painting. 

Questions about Medieval Gesso Grounds 

Beate Federspiel 

Konservators Skolen 

Danske Kunst Akademie 

Esplanaden 34 

1263 Kopenhagen-K 

Denmark 

Introduction 

The problems in interpreting written sources on painting techniques are well 

known. Besides the paintings themselves, the written sources are the only 

testimony of materials and techniques used in former times. Advanced methods 

of scientific analysis employed in the examination of paintings do not 

always answer questions about materials and techniques. And the written 

sources do not always provide easy access to painting techniques. Time has 

obscured the comprehension of the texts. The pure linguistic translation of a 

written source is often far from sufficient, but may be greatly aided by reconstructing 

the technical details described and comparing the results of the 

scientific analyses. 

Understanding Cennini's text 

During the courses in historical techniques of painting at the School of Conservation 

in Copenhagen, it was increasingly dissatisfying for the author and 

others to reconstruct the gesso grounds as described by Cennino Cennini, 

and it was eventually necessary to scrutinize his text concerning grounds. 

Merely reading Cennini's instructions in the English translation by D. V 

Thompson in 1933 was not sufficient; it was obvious that things were not as 

easy as they may have seemed (1). Not that Cennini is imprecise in his instructions 

on this point; he is more thorough in his instructions on ground 

than in his description of paint application. But how can certain important 

passages be interpreted 600 years later? 

The original manuscript by Cennini being lost, the question of which surviving 

copy to use as a source remains, of course, a central one. That aspect 

will not be addressed in this article; the source used here is Lindberg's Swedish 

version of Cennini's Codex Laurentianus. Lindberg's translation shows semantic 

details, absent in previous translations, that are important for the understanding 

of decisive technical details (2). 

In Chapter CXV of his treatise, for example, Cennini describes the preparation 

of the ground for painting on panel. Painting in the Middle Ages 

included gilding. Gilding was the main reason for the great efforts invested 

in creating a perfect ground. Gypsum was the material used in the preparations 

of grounds for painting and gilding throughout the whole Mediterranean 

area as far back as the first millennium B.C.E. (3). The first written 

evidence of a ground for painting made of gypsum appears in the ninthcentury 

Lucca manuscript, which mentions a ground consisting of gypsum 

and glue for gilding on wood (4) . Cennini clearly distinguishes between gesso 

grosso and gesso sotille; that is, a double-structured ground consisting of several 

layers of a coarse ground on top of which are applied several layers of a finer 

ground. In both structures, the medium is animal glue (5). Such grounds, with 

local variations, were found in fourteenth- and fifteenth-century paintings 

from Florence and Siena (6) . 

Lindberg's translation, here translated from Swedish into English by the author 

(7), differs in several crucial passages from the 1933 English translation by D. 

V Thompson. The passage concerned is the following: The Italian text says, 

"poi abbi giesso grosso cioe uolteriano che e purghato ede tamigiato amodo 

di farina, ..." which Thompson translates as, "then take some gesso grosso, 

58 


that is, plaster of Paris, which has been purified and sifted like flour" (8, 9). 

The Swedish translation says: "Tag sedan grovgips, det vill saga Volterra-gips, 

som er renad och siktad som mj ol. ..." [Then take coarse gypsum, that is to 

say Volterra gypsum that has been purified and sifted like flour. ... J 

The question is, apart from the obvious fact that the gypsum mentioned was 

mined in the quarries at Volterra, what was then understood by this statement? 

Initial attempts by the author at reconstructions were based on 

Thompson's translation. The use of his "plaster of Paris," a gypsum burned 

to the hemihydrate fo rm, resulted in immediate setting in water (10). 

The next chapter, which describes the preparation of the gesso sottile, reveals 

details about the gesso volteriano not mentioned in Chapter CXV Chapter 

CXVI says, "Ora si uuole chettu abbi dun giesso elquale sichiama giesso 

sottile elquale e diques to medesimo giesso mae purghato perbene unmese 

tenuto in molle innun mastello rinuoua ogni di laqua chesquasi siinarsiscie 

edesciene fuori ogni fo chor di fuocho e uiene morbido chome seta ...." (11). 

In Lindberg's translation: "Nu will man att du skall ha en gips som man kallar 

fingips, vilken besrar av denna samma gips, men den ar renad i gott och val 

en manad, lagd i blot i en balja. Byt varje dag vattnet tills den ar nastan torr, 

och varje glad av eld gar ut den, och den blir mj uk som silke ...." [Now 

you should have a gesso, which is called fine gesso, which is this same gesso, 

but it has been purified for a good month, soaked in a basin. Change the 

water every day until it is almost dry, and every glow of fire leaves it, and it 

will be soft as silk .... 

J 

There are two things to be noticed here: Lindberg argues that the remark 

about the glow of fire leaving the gypsum-the same gypsum that was used 

for the gesso grosso-can only be understood in the sense that the gypsum 

was indeed burned. The question is, what form resulted from the process? 

The remark about the soaking in water constitutes another important point: 

What kind of procedure is meant? The existing translations are not very clear 

about this point. It seems quite conceivable that a double purpose-a washing 

process and a process of changing the morphology and chemical composition 

of the material-was served by this treatment. 

The preparation of gesso grounds 

The naturally occurring gypsum, calcium sulfate dihydrate CaS04'2H20, can 

be burned at various temperatures. Burning at 128 °C produces CaS04'H20, 

the hemihydrate form. Burning at 130-160 °c creates an anhydrite and hemihydrate 

mixture (12). This is the so-called plaster of Paris or stucco plaster, 

which sets quickly with water and thus returns to the dihydrate form. Between 

163 °C and 300 DC, soluble anhydrite, CaS04, is formed, which also 

reacts quickly with water. According to Mora, et aI., the dihydrate form will, 

at temperatures above 250 °C, turn into insoluble anhydrite that is no longer 

able to set with water (13). Experiments at the School of Conservation have 

shown, however, that anhydrite burned at 300 °C, 400 DC, and even 500 °C 

is still able to react with water, forming dihydrate again (Figs. 1 , 2) (14). The 

Merck Index even gives 650 °C as the limit above which the insoluble anhydrite 

is formed (15). Above 900 DC, the so-called Estrich gypsum (a combination 

of anhydrite and calcium oxide) is fo rmed; this material sets very 

slowly with water and becomes extremely hard. Gypsum can also occur in 

nature as insoluble anhydrite, CaS04. 

In Cennini's time, the mined gypsum was burned in rather primitive kilns. 

Gettens describes a very ancient kiln with little temperature control, in which 

blocks of gypsum are stacked and a fire lighted at the base of the kiln (16). 

The result of this process must have been a mixture of anhydrite, hemihydrate, 

and even dihydrate forms. Overburned or dead-burned insoluble anhydrite 

lumps must have been prevalent at the bottom of the stack near the source 

of heat. In the upper tiers, the lumps burned at lower temperatures, hemihydrate/ 

anhydrite forms must have been present in larger relative amounts. 

Federspiel 59

70 

60 

50 

40 

a 3500 3000 

Transmittance I Wavenumber (cm-1) 

2500 2000 1500 1000 500 

80 

b 3500 3000 

Transmittance I Wavenumber (em-1) 

2500 2000 1500 1000 500 

60 

40 

2 

c 3500 

Transmittance I 

Wavenumber 

II 

3000 

(em-1) 

2500 2000 1500 1000 500 

Figure 1. Fourier transform infrared spectra of calcium sulfa te Lilith d!fJerent amounts if crystal water: 

(a) anhydrite; (b) hemihydrite, dihydrate baked at 135 °Cfor two hOLm; (c) dihydrite. 

It is even possible, as suggested by Gettens and Mrose, that if the lumps were 

fairly large, that the outside would have been burned to anhydrite while the 

inside would be only partially dehydrated (17). Some of the inside of the 

gypsum lumps in the primitive kiln stack could have been heated below 100 

°C so that the gypsum remained hydrated. It is interesting to note here that 

Theophilus gives the following instructions: " ... take some gypsum burned 

in the fashion of lime ...." which must be understood to have been burned 

at very high temperatures, that is, above 900 °C (18). The resulting substance 

will be quicklime (calcium oxide), which is slaked in water and transformed 

into lime (calcium hydroxide). 

Gettens describes another kind of kiln that worked in the same way as ancient 

bread-baking ovens; after a fire had preheated the inside of the brick oven 

until the interior was red hot, the fire was extinguished, the coals removed, 

and the bread-or, in this case, lumps of gypsum-were placed in the oven 

overnight. This kind of operation allowed for a more uniform heating. This 

60 


100 

80 

60 

40 

a 3500 30.00 2500 2000 1500 1000 500 

Transmittance J Wavenumber (cm-1) 

20 

b 3500 3000 2500 2000 1500 1000 500 

Transmittance I Wavenumber (cm-1) 

70 

60 

50 

20 

10 

c 3500 

Transmittance I 

Wavenumber 

3000 2500 2000 1500 1000 500 

(cm-1) 

Figure 2. Fourier traniform irifrared spectra of gypsum bumed at 500 ° C: (a) anhydrite sample before 

treatment with water; (b) slurred in water for two days, the sample still consists of anhydrite; (c) slurred 

with water for ten days, the sample now consists of dihydrite. 

kind of oven had been in use in the Mediterranean civilizations since early 

Roman times (19). The painter's manual of Mount Athos actually describes 

the baking of gypsum in a similar type of kiln (20). This manual suggests that 

the product from the burning process would be soluble anhydrite, which, 

soaked in water, would form the dihydrate product again. 

In Cennini's directions for the preparation of the gesso sottile, the burnt 

gypsum should be soaked in water for about a month. Lindberg states that 

Thompson's translation of the word rinuoua as "stir up," referring to stirring 

the water every day, is not correct. Thompson may have mistaken the "n" 

for "m" (21). Rinuoua literally means renew or change. Lindberg further 

argues that siinarsiscie means "to dry," dismissing Thompson's translation of 

"rots away" as incorrect. Lindberg offers the following interpretation: the 

water should be poured off the soaked gesso every day until the gesso is 

almost dry, which simply means that as much water as possible should be 

Federspiel 61

poured off each time; such a process is actually a washing process. Watersoluble 

impurities, such as salts, possibly present in the gypsum would be 

washed away in this process. Such impurities might cause discoloration of the 

ground or efflorescence of salts from it. Apart from this, soaking the burned 

gypsum in water obviously also had the function of changing the texture and 

the chemical composition of the material, supposing the point of departure 

is soluble anhydrite. 

Technical evaluation and comparison of results 

The recent examination of grounds in Italian paintings by the Laboratoire 

de Recherches des Musees de France elaborates on the double structure of 

the Italian gesso grounds (22). This double structure was also shown in the 

examinations by the National Gallery's laboratories in London (23). Unfortunately, 

this point was not addressed in the otherwise excellent 1954 examination 

by Gettens and Mrose, which makes their results of somewhat 

limited value in this context. 

Concluding from the results of examinations of grounds, the gesso grosso 

consists of mainly anhydrite, sometimes with dihydrate present. I am here 

referring to the French examination, which even states the ratio of anhydrite 

to dihydrate (24). The numerous Florentine and Sienese examples show the 

following compositions for the gesso grosso: 100% anhydrite or 75:25 anhydrite:dihydrate 

or 50:50 anhydrite:dihydrate. For the gesso sottile, two ratios: 

100% dihydrate or dihydrate containing 25% anhydrite. 

Considering the chemical changes of soluble anhydrite in contact with water 

(the binding medium of the ground is animal glue, which always contains a 

certain amount of water), it seems puzzling that the gypsum in the layers of 

gesso grosso kept its anhydrous fo rm. A possible explanation could be that 

the conversion process became slower as the burning temperature increased 

(25). As the anhydrite did not change into dihydrate, the evaporation time of 

the water in the binding medium must have been shorter than the hydration 

time of the anhydrite. It would be extremely improbable that the gesso grosso 

material would be a stable anhydrite, which is not able to react with water, 

or a dead-burned anhydrite, which is neither form, because the examples 

concerned show the double-structured grounds of gesso grosso and gesso 

sottile. It would be very unlikely that two different kinds of gypsum would 

be employed in the process. The analyses of the sottile layers show dihydrate, 

which is the soluble anhydrite soaked in water. It must be assumed that the 

point of departure in both the grosso and the sottile is the same compound: 

soluble anhydrite. The presence in the gesso grosso layers of a mixture of 

anhydrite and dihydrate would be explained by the sometimes poorly controlled 

burning process. 

In the samples from Umbria, Latium, the Marches, Venice, and Ferrara showing 

single-structured grounds, the dihydrate present (100%) is claimed to be 

the natural unburned dihydrate (26). The raw gypsum (calcium sulfate dihydrate) 

does not differ either in chemical composition or crystal structure 

from a dihydrate that has gone through the process applied to a gesso sottile 

of burning and soaking in water (27). 

The question remains, is it possible that the dihydrate present in the examples 

mentioned above could be the processed dihydrate? The author's reconstructions 

indicate that the raw calcium sulfate dihydrate can only with great 

difficulty be triturated to a degree that will make it usable as a ground. Even 

then, the resulting surface will not facilitate a satisfactory base fo r gilding. It 

must be admitted, however, that raw gypsum can vary considerably with 

regard to texture (28). Considering this, we can exclude the possibility that 

raw gypsum could have been pulverized to yield a satisfactory product. It 

seems likely that in most cases the process would be greatly facilitated by first 

burning the calcium sulfate to an anhydrite, followed by grinding and processing 

with a final soak in water to produce the dihydrate form. This is the 

62 


process described by Cennini for gesso sottile as furnishing the best ground 

for gilding (29). In fact, Cennini describes in Chapter CXVIII how certain 

panels can be grounded only with gesso sottile. 

Conclusion 

Finally, returning to the meaning of the giesso volteriano in Cennini's text, 

we presume that the material must have been calcium sulfate dihydrate 

burned at 300-650 °c, thus forming soluble anhydrite, which was used fo r 

the gesso grosso. The preparation of the gesso sottile would be a process of 

changing the chemical composition, altering the texture, and washing. 

The author does not agree with the point of view put forward by E. Martin, 

et al. that Cennini did not know whether the giesso volteriano he discussed 

was unburned gypsum directly from the quarries of Volterra or if it was 

processed by being burned (30). Cennini must have had a good understanding 

of these materials. That we do not understand the exact meaning of his words 

600 years later is another matter. In his own time and geographical sphere, it 

was most likely very well understood, at least by other craftspersons. 


It was extremely difficult to obtain raw calcium sulfate dihydrate in Denmark. I am 

very much indebted to the following persons who made the practical experiments 

possible: V Meyer (Building Materials, Kalkbr;enderihavnsgade 20, 2100 Copenhagen), 

who readily took the trouble to import unburned gypsum for my research 

purposes, and to Niels Erik Jensen (CTO, Aiborg Portland, 9220 Aiborg 0st), who 

demonstrated a keen interest and readiness to help with this project by providing me 

with raw gypsum in various forms. I wish to thank my colleague, chemist Mads Chr. 

Christensen (School of Conservation, Copenhagen) fo r his help in carrying out the 

analyses, and for fruitful discussions. My colleague, geologist Nicoline Kalsbeek, 

(School of Conservation), has kindly furnished me with information from the Cambridge 

Structural Database. Finally, I wish to thank Professor Bo Ossian Lindberg 

(Institutionen fOr Konstvetenskap, University of Lund, Sweden) for reading and commenting 

on the text. 

Notes 

1. Cennini, C. 1960. The Craftsman's Handbook: II libro de/l'arte. Translated by D. V 

Thompson, Jr.. New Yo rk, Dover Reprint (Yale University Press, 1933). 

2. Cennini, C. Boken om nullarkonsten. Oversiittning och notkommentar av Ossian Lindberg. 

Unpublished manuscript provided by Professor Lindberg. Part of the text 

has been published in Lindberg, B. 0. 1991. Antologi om milleriteknik. Lund: Institutionen 

fo r Konstvetenskap, Lunds Universitet. As shown by Lindberg, the 

Codex Laurentianus (Florence, Biblioteca Laurentiana, MS 78 P 23) must be considered 

closest to the lost original manuscript. The Codex Riccardianus (Florence, 

Biblioteca Riccardiana, MS 2190), however, is our only source for several passages 

which are missing in the Laurentianus text. Therefore, Lindberg based his edition 

on both the Laurentianus and the Riccardianus codices, as Thompson also did. 

For the discussion of the surviving transcripts of Cennini's text and the various 

editions and translations, see Lindberg, B. 0. 1991. Cennino Cennini: den obnde 

masteren. In Miliningens anatomi. Exhibition catalogue. Kuituren, Lund, 39-48. 

3. Oddy, W A. 1981. Gilding through the ages: An outline history of the process 

in the Old World. Gold Bulletin and Gold Patent Digest (14): 75-79, in particular 

page 77. See also, Gilded wood: conservation and history. 1991. Gilding Conservation 

Symposium, Philadelphia Museum oj Art. Madison, Sound View Press, 34. 

4. Berger, E. 1973. Quellen und Technik der Fresko-, Oel-und Temperamalerei des Mittelalters. 

Liechtenstein Sandig reprint (Munich 1912), 13. 

5. The medieval sources mentioning the preparation of grounds also mention (animal) 

glue as the medium. The only exception, to my knowledge, is the fifteenthcentury 

Bolognese manuscript, which mentions a ground that apparently has no 

medium but depends on the setting with water of the hernihydrite. See Merrifield, 

M. P 1849. Segreti per colori, Original treatises on the arts oj painting, 

Vol. II. London, John Murray, 595. 

6. Martin, E., N. Sonoda, and A. R. Duval. 1992. Contribution a l'etude des preparations 

blanches des tableaux italiens sur bois. Studies in Conservation (37): 82- 

Federspiel 63

92. The fifty paintings examined by the authors reveal a clear trend. Paintings 

belonging to the Florentine and Sienese schools showed double-structured 

grounds of gesso grosso and gesso sottile, identified respectively as anhydrite and 

dihydrate. Venetian examples showed a tendency towards single-structured 

grounds consisting only of dihydrate, as also observed in Gettens, R. E. and M. 

E. Mrose. 1954. Calcium sulphate minerals in the grounds of Italian paintings. 

Studies in Conservation (1):174-90. See in particular pages 180-83. The recent 

French examination finds that the examined paintings representing schools outside 

Tuscany are not numerous enough in their material to conclude anything 

general about the treatment of grounds in the rest of Italy. 

7. See note 2. 

8. The wording of the Codex Laurentianus is quoted from Lindberg, B. 0. 1990. 

Feta och magra limmer enligt Cennino Cennini. In Meddelelser om Konservering, 

165-87. Lindberg notes that the Codex Riccardianus in this passage has a slightly 

different, but not clearer wording than the Laurentianus. Lindberg points to the 

fact that the manuscript lacks interior punctuation. 

9. Cennini, 1960. Op. cit., 70. 

10. Unfortunately, this suggestion by Thompson, which in the author's opinion is 

wrong, is apparently reflected in numerous texts on medieval grounds. See, for 

instance, Straub, R. E. 1984. Tafel-und Tiichleinmalerei des Mittelalters. In Reclams 

Handbuch der kiinstlerischen Techniken, Vol. I. Stuttgart Philipp Reclam, 156- 

57. 


12. The various sources on gypsum are obviously not identical in their information 

on the relationship between the chemical composition of the gypsum and burning 

temperatures. Rompps Chemielexicon. 1979. Stuttgart, 575-76. See also, Mora, 

L., P. Mora, and P. Philippot. 1984. The Conservation of Wallpaintings. London, 

Butterworth, 39-47. Also see The Merck Index, 9th ed. 1976, 1707. 

13. Mora, et aI., op. cit. (note 12), 42. 

14. Lindberg, B. 0. , and B. Skans, in their experiments carried out at Institutionen 

fo r Konstvetenskap i Lund, have also observed that gypsum burned at 300 °C 

fo r two hours is still able to react with water. Lindberg, B. 0. , and B. Skans. 1990. 

Feta och magra limmar enligt Cennino. In Meddelelser om konservering, 184. 


16. Gettens, R. J. 1954. A visit to an ancient gypsum quarry in Tuscany. Studies in 

Conservation (1):4, 190-92. 

17. Gettens and Mrose, 1954. Op. cit., (note 6), 185. 

18. Theophilus on Divers Arts. 1979. Translated by G. Hawthorne and C. S. Smith. 

New York, Dover Reprints, 27. 

19. A History of Technology, Vol. II. 1957. Oxford, The Clarendon Press, 118. 

20. The Painter's Manual of Dionysius of Fourna. 1978. Translated by P. Hetherington. 

London, 6. It appears from the directions in the Athos book, that right from the 

mining, it is important to choose the right kind of gypsum: " . .. see that you 

use only what is white and glistening." 

21. Lindberg, 1990. Op. cit. (note 8), 184. 


23. Bomford, D., J. Dunkerton, D. Gordon, A. Roy, and J. Kirby. 1990. Art in the 

Making. London, National Gallery. See in particular pages 17-19. 

24. Martin, Sonoda, and Duval, 1992. Op. cit., 86-89, Table 2. 

25. It must, of course, be below 650 °C, which is the limit beyond which the anhydrite 

is no longer able to react with water. 

26. Martin, Sonoda, and Duval, 1992. Op. cit., 90. See also Dunkerton, J., and A. 

Roy. 1986. The Technique and Restoration on Cima's The Incredulity if S. Thomas. 

National Gallery Technical Bulletin (10) :4-27, especially page 5. See also Dunkerton, 

J., A. Roy, and A. Smith, A. 1987. The unmasking of Tura's "allegorical 

figure": A painting and its concealed image. National Gallery Technical Bulletin 

(11):5-35, especially 19. 

27. If a different crystal structure had existed fo r the dihydrate it would most likely 

have been included in the CSD (Cambridge Structural Database). This is not 

the case. 

28. It was quite obvious that not only was the burned gypsum (the anhydrite) much 

easier to grind than the raw gypsum, but the anhydrite burned at 700 °C was 

much easier to grind than the anhydrite burned at 400 0c. 

29. Cennini, 1960. Op. cit., 73. 

30. Martin, Sonoda, and Duval, 1992. Op. cit., 90 

64 


Abstract 

The frequent use of verdigris in 

paintings from the early Gothic 

period to the eighteenth century is 

reflected by the citation of the pigment 

in treatises. The many warnings 

against the use of verdigris and 

precautions recommended to prevent 

it from discoloring are set against the 

fact that verdigris and copper resinate 

have survived in many cases as an 

intensely green color. It is suggested 

that the method of preparation, dissolving 

verdigris in warm oil, and its 

application with a piece of canvas 

contribute to the stability of the 

green glaze. Colored underpainting 

and yellow glazes on top of the 

green layer were used to modify the 

green tone. 

Aspects of Painting Technique in the Use of Verdigris 

and Copper Resinate 

Renate Woudhuysen-Keller 


Fitzwilliam Museum 

University of Cambridge, Whittlesford 



Introduction 

Verdigris and copper resinate have been fo und in paintings dating from medieval 

times to well into the eighteenth century. Among the green substances 

available for painting, verdigris had the most intense color but little covering 

power compared with malachite. Verdigris was made by exposing strips of 

copper to vinegar vapor as described by Theophilus at the beginning of the 

twelfth century (1). Many later treatises and artists' handbooks contain recipes 

on how to make verdigris, how to distill it by dissolving the material in 

vinegar and allowing it to recrystallize, and how to turn it into the transparent 

green color that today is called copper resinate. 

The use of verdigris 

Verdigris was used ground in oil, in cherry gum, and also in egg, but it had 

a bad reputation for turning brown or black. Cennino Cennini, at the end 

of the fourteenth century, wrote that "it is nice to the eye but it does not 

last," and in the fifteenth-century Strasburg Manuscript, a note states that 

verdigris does not agree with orpiment (2, 3) . Leonardo da Vinci wrote at 

the end of the fifteenth century that the beauty of verdigris "vanishes into 

thin air if it is not varnished immediately" (4) . The warnings become even 

clearer in the Brussels Manuscript, written by Pierre Lebrun in 1635: "Verdigris 

is added to charcoal black, or lamp black, to make these colours dry, 

but it is used only with the shadows, for it is a poison in painting, and it kills 

all the colours with which it is mixed" (5). At the same time, between 1620 

and 1640, De Mayerne in London wrote in his notebook (6): 

Le verd de gris} which is only used Jo r glazing} is an enemy oj all colours} 

so much so} that it kills them all} especially azurite. Even if you work 

with a brush that has been cleaned with oil that has come into contact with 

verdigris before} as clean as it may seem} or if you put the colours on a 

palette on which there has been verdigris before} it spoils everything. In this 

way} he who wishes to work with verdigris has to keep brushes} palette 

and oil Jor cleaning separate. 

In 1757 Pernety wrote disdainfully, "Ver de gris: this is poisonous for all 

animals as well as for colours; if one wants to make use of it in oil painting, 

one has to use it on its own or at most mixed with black. It ruins all the 

colours, and even if there was only a little bit in the grain of the canvas it 

would spoil all the colours that were laid on it in the process of painting" 

(7). 

Chemical analysis, however, has shown that the beautiful greens and green 

glazes that have survived in paintings are indeed very often verdigris and 

copper resinate. Numerous reports published in the National Gallery Technical 

Bulletin and in the Bulletin de I } Institut Royal du Patrimoine Artistique bear witness 

to this fact. In his investigation of verdigris and copper resinate, Hermann 

Kuhn came to the conclusion that "experiments with paint specimens and 

observations on paintings ... show that the properties of verdigris are not 

nearly as harmful as suggested by the literature" (8). 

Woudhuysen-Keller 65

Why have some green paint layers of verdigris and copper resinate turned 

brown while others remained intact? After a search through the treatises for 

instructions on how to use this dangerous green pigment and a comparison 

of the written instructions with the actual methods of applying the pigment 

on the paintings, results seemed to suggest that three interesting factors determine 

appearance: (1) verdigris has to be thoroughly embedded in oil or 

an oil-resin varnish to be protected from air and humidity; (2) colored underlayers 

contribute to the beauty of the green glaze; and (3) admixtures of 

yellow lakes or yellow glazes on top of the transparent green layer were 

applied to soften the sharpness of the bluish-green tone of verdigris and 

copper resinate. 

The instructions for the use of verdigris emphasize that it should be thoroughly 

incorporated into the medium and covered with varnish as soon as 

possible. The early use of verdigris for pictura transludda on gold leaf and tinfoil 

described by Theophilus, Cennini, and in the Tegernsee Manuscript (ca. 

1500) points to this fact (9, 10, 11). The Strasburg Manuscript simply states 

that all colors should be ground in oil. It proceeds to explain how to mix 

colors and how to achieve good results by painting in several layers, the 

painting technique generally found in fifteenth-century paintings (12). 

In On the true precepts if the art of painting (1587), Armenini gives very detailed 

instructions on painting technique, particularly on how to paint green drapery 

(13): 

if the drape is to be green, one does as follows: After the sketch is made 

using somewhat coarse green, black and white, it is lightly painted with a 

mixture if verdigris, a little common varnish, and some giallo santo. With 

a coarse brush of miniver, one veils the sketch uniformly; next one pats it 

either with the palm if the hand or with a little wad if cotton wool covered 

with linen, unta the given colour is uniform and no brush strokes can be 

detected. And if the result is not to one's satisfaction, after the veiling is 

dry one repaints with the same mixture and then pats in the prescribed 

way. 

He also explains that before veiling, the thoroughly dried sketch has to be 

oiled out very thinly to stop the glaze from being repelled by the underlayer. 

A painting by Palmezzano depicting the Mystic Marriage if Saint Catherine, 

signed and dated 1537, shows exactly this technique in the green drapery of 

the throne and St. Catherine's green garment (Plate 9). The pattern of the 

textile used for dabbing on the glaze is clearly visible (Plate 10). The buildup 

shown in the cross section also corresponds to Armenini's instructions 

(Plate 11); even the oiling-out layer is visible in ultraviolet light (Plate 12). 

The glaze was apparently too viscous to be spread out evenly with a brush, 

therefore the glaze was spread by dabbing it with a rag. Traces of textile 

pattern are also visible in a Flemish altarpiece, painted in Antwerp around 

1520, now at Oxburgh Hall in Norfolk. In the process of dabbing on the 

green glaze the artist could not always keep within the outline of his green 

drapery, so he had to retouch the background in some places. Minor overlaps 

were simply left; one can see the weave pattern of the rag, the glaze is partly 

discolored and some green particles are still visible. 

Experimentation 

As an experiment, some neutral recrystallized verdigris was ground in linseed 

oil and the mixture was heated very gently to approximately 50 °C, until the 

copper acetate had dissolved and the pigment grain had disappeared. The 

mixture was intensely green and quite viscous. It could be spread with a 

brush while warm, but congealed very quickly, making the brush strokes very 

coarse and imprecise. However, it was quite easy to spread the glaze by dabbing 

it on with a piece of canvas. After a few hours, the glaze dried. Within 

a few days the thick brush strokes of green glaze showed drying wrinkles, 

whereas the dabbed area looked the same as when it was applied, demon- 

66 


strating that spreading this glaze thinly in several layers is not merely a matter 

of style but also a technical necessity. 

The green tablecloth in Jan Davidsz de Heem's Still Life with Fruit (Cambridge, 

Fitzwilliam Museum), painted in 1650, consists of a green glaze over 

a dark brown underlayer with yellow-green highlights. Clearly this green 

glaze has been applied with a brush, as the fairly clumsy brush marks are 

visible. Along the contours of the vine leaves and on either side of the thin 

stalks of the cherries, the green glaze leaves a gap showing the dark brown 

underlayer. It is quite clear that this glaze must have been a rather viscous 

liquid and therefore difficult to paint out with any precision. 

The recipes for making copper resinate, collected by the doctor Theodore 

Turquet de Mayerne in London between 1620 and 1640, are generally considered 

to be the earliest. They call for verdigris to be heated with Venetian 

turpentine and oil of turpentine as follows: "Beautiful green: take 2 ounces 

of Venetian turpentine, 1 112 ounce of oil of turpentine, mix and add 2 

ounces of verdigris in little pieces. Set it on hot ashes and let it boil gently. 

Try it on some glass to see if you like the colour; strain it through a cloth" 

(14) . 

Trying out the recipe, it was found that the verdigris did not dissolve in the 

mixture of Venetian turpentine and oil of turpentine because there was not 

enough resin present; also, the presence of oil of turpentine hampered the 

reaction of the copper acetate with the Venetian turpentine. When more resin 

in the fo rm of rosin was added, a dark green resinous substance resulted, 

which was liquid while hot, but hard and glassy as it cooled. This green glassy 

substance can be ground in oil like a pigment. When ground in oil, however, 

the color is no longer very intense. 

Another recipe in the De Mayerne manuscript asks for verdigris, ground in 

oil, to which hot common varnish is added: "Painters, i.e. those who paint 

as well as those who paint furniture and blinds, grind verdigris with linseed 

oil and then add common varnish, stirring it well. They allow the impurities 

to sink down and only use the clear liquid, which they apply warm" (15). 

With the term "common varnish" a solution of resins in oil is meant. This 

recipe was the base for the experiment described above, grinding verdigris 

in oil and heating it. 

The preparation and application of the green glaze is not the only secret of 

de Heem's tablecloth: the dark reddish-brown undermodeling gives the green 

glaze its velvety depth. This observation proved to be very useful during 

retouching. The only way to match this intense dark green color was to 

reconstruct the build-up of layers exactly. The verdigris in the green glaze 

was substituted with the transparent green pigment viridian (because of its 

stability) and a little synthetic Indian yellow to match the required tone. 

The reason a green glaze over a reddish-brown underlayer appears so very 

dark lies in the absorption of the waves of the spectrum: green absorbs all 

red waves, red absorbs all green waves. The two layers superimposed absorb 

practically the entire spectrum of visible light, so that the resulting color is 

almost black. 

The painter's use of green glazes and grisaille 

The green curtain in Titian's Tarquin and Lucretia (Fitzwilliam Museum) 

shows green glazes over an undermodeling in red with broad white highlights. 

The idea is ingenious. To start with, Titian underpainted the curtain in gray 

with some azurite, then he laid in the modeling with a brownish red containing 

some red lake and some very generous white highlights. As he applied 

the green glaze, the shadows in the folds appeared very dark green, the middle 

tones were light green because of the green glaze over white brush strokes, 

and the flickering highlights remained white from the undermodeling, partly 

emphasized with an extra brush stroke. In some places, Titian allowed the 

red to shine through, giving the material a wonderful shot-silk effect. 


In late Gothic painting, grisaille underpainting was fairly common for green 

garments or landscape elements. This was found in paintings by van Eyck as 

well as by Uccello, north of the Alps as well as in Italy (16, 17). The method 

is described in artists' treatises. Palomino, in the first quarter of the eighteenth 

century, describes the method of underpainting green cloths with grisaille 

(18). The same instructions are found in Croeker's treatise, published in 1743 

(19). This documents a long tradition. Grisaille undermodeling was not the 

only method recommended. In his Portrait if a Dominican Monk (Upton House 

near Banbury), painted around 1525, Lorenzo Lotto used undermodeling in 

azurite and lead white with the darkest shadows painted in an olive green 

mixture of black and lead-tin yellow. The next layer contains verdigris with 

very little lead-tin yellow and some black in the shadows, fo llowed by a green 

glaze. This green glaze is a substantial, absolutely homogenous layer, without 

any visible brushwork or textile pattern. 

In addition to colored undermodeling to modifY the green glaze, there is the 

possibility of adjusting the green color by means of yellow glazes. Cennini 

recommended saffron for this purpose (20). Leonardo suggested aloe dissolved 

in warm spirit of wine: "If you have finished a painting with this simple 

green [verdigris in oil] and if then you were to glaze it lightly with this aloe 

dissolved in spirit of wine, then it would be of a most beautiful colour. Also 

this aloe can be ground in oil, either on its own or together with the copper 

green and with any other colour you like" (21). 

De Mayerne noted a recipe fo r a green copper glaze that called fo r some 

terra merita, which is curcuma, to be added to improve the color (22). Goetghebeur 

and Kockaert drew attention to the existence of yellow and brown 

glazes on top of layers of verdigris or copper resinate (23). In cleaning green 

areas in paintings, therefore, it is not sufficient to test a brown layer for copper 

to find out whether the paint is discolored copper resinate. Even if it is 

not, it can still be an original yellow lake, now discolored. As these organic 

lakes are very difficult to identifY, their presence is often suspected without 

reassuring proof either way, a most disconcerting situation fo r the restorer. In 

addition to research by means of scientific analysis, it might be helpful to 

approach the problem via the written sources and by reconstructing some of 

the recipes. 

Notes 

1. Hawthorne, J. G. and C. S. Smith. 1979. Theophilus on Divers Arts. New York: 

Dover Publications, 41. See also Scholtka, A. 1992. Theophilus Presbyter: Die maltechnischen 

Anweisungen und ihre Gegenuberstellung wit naturwissenscha J tlichen Untersuchungsbifunden. 

Zeitschrift jur Kunsttechnologie und Konservierung 6 (1):1-54. 

2. Cennini, C. 1960. The Crcifisman's Handbook. Translated by D. V. Thompson. New 

York: Dover Publications, 33. 

3. Borradaile, V. and R. Borradaile. n.d. The Strasburg manuscript: a medieval painters' 

handbook. Munchen: Callwey, 55-7. Passage translated by the author. 

4. Ludwig, H. 1882. Lionardo da Vinci. Das such von der Malerei. Nach den Codes 

Vaticanus (Urbinas) 1270. Wien: Braumiiller (Quellenschriften fLir Kunstgeschichte 

und Kunsttechnik des Mittelalters und der Renaissance), 15:244. Passage 

translated by the author. 

5. Merrifield, M. P. 1967. Original Treatises on the Arts if Painting, Vol. 2. New York: 

Dover Publications (2):822. 

6. van de Graaf,J. A. 1958. Het De Mayeme Manuscript als bron voor de schildertechniek 

van de barak. British Museum, Sloane 2052. Proefschrift ... Mijdrecht: Drukkerij 

Verweij, 142. Passage translated by the author. 

7. Pernety, A. J. 1972. Dictionnaire portatif de peinture, sculpture et gravure. Geneva: 

Minkoff Reprint, 547. Passage translated by the author. 

8. Kuhn, H. 1993. Verdigris and copper resinate. In Artists' Pigments. Ed. A. Roy. 

Washington: National Gallery of Art, 136. For preparation and identification of 

green copper pigments, see Ellwanger-Eckel, E 1979. Herstellung und Verwendung 

kunstlicher gruner und blauer Kupjerpigmente in der Malerei. Stuttgart: Institut fu r 

Museumskunde an der Staatlichen Akademie der Bildenden Kunste. 

9. Hawthorne and Smith, op. cit., 33. 

68 


10. Thompson, op. cit., 61. 

11. Berger, E. 1975. Quellen und Technik der Fresco-, Oel-und Tempera-Malerei des Mittelalters 

. . . Walluf-Nendeln: Sandig Reprint, 195. 

12. Borradaile, op. cit., 55-7. 

13. Armenini, G. B. 1977. On the true precepts of the art of painting. Edited and translated 

by E. J. Olszewski. Burt Franklin, 194. 

14. van de Graaf, op. cit., 174. Passage translated by the author. 

15. van de Graaf, op. cit., 164. Passage translated by the author. 

16. Brinkmann, P W F., et al. 1984. Het Lam Godsretabel van van Eyck . .. Bulletin 

de l'Institut Royal du Patrimoine Artistique (20):164. 

17. Massing, A. and N. Christie. 1988. The Hunt in the Forest by Paolo Uccello. Bulletin 

of the Hamilton Kerr Institute (1):36. 

18. Veliz, Z. 1986. Artists' Techniques in Colden Age of Spain. Cambridge: Cambridge 

University Press, 168-69. 

19. Croeker, J. M. 1982. Der wohl anfuhrende Mahler. Rottenburg: Kremer-Reprint, 

110. 

20. Thompson, op. cit., 29-30. 

21. Ludwig, op. cit., 244-7. Passage translated by the author. See also Chastel, A. n.d. 

Leonardo da Vinci. Samtliche Cemalde und die Schriften zur Malerei. [Darmstadt] : 

Wissenschaftliche Buchgesellschaft, 351-52. 

22. van de Graaf, op. cit., 174. 

23. Goetghebeur, N., and L. Kockaert. 1980. Le grand Calvaire d'Albert Bouts au 

M usee des Beaux-Arts de Bruxelles. Bulletin de l'Institut Royal du Patrimoine Artistique 

(18):5-20. 


Abstract 

The discovery of three "new" pigments 

is described. Their history is 

traced through the literature of pigments, 

ceramics, and glass. Tin white, 

previously undiscovered, is followed 

from Iraqi ceramics to its occurrence 

as a pigment on Jain miniatures in 

India. Burnt green earth, mentioned 

by early Italian and nineteenth-century 

English writers, is identified on 

nineteenth-century oil paintings. 

Cobalt oxide colorant is followed 

from Persia to Europe and China 

and finally, in a new fo rm, to Venetian 

enamels and as smalt on Hindu 

miniatures. 

Connections and Coincidences: Three Pigments 

Josephine A. Darrah 

Science Group 

Conservation Department 

Victoria & Albert Museum 

London SW7 2RL 


Introduction 

There are certain advantages in being a generalist rather than a specialist. In 

this paper, the author hopes to show some of the connections and coincidences 

that occur when a wide range of materials and objects are analyzed 

in the same laboratory. 

Tin white 

The pigment known as tin white is elusive. It is mentioned occasionally in 

medieval and later European literature, but has never been identified on paintmgs. 

Tin oxide was used first as an opacifier in ceramic glazes to reproduce Chinese 

porcelains in ninth-century Iraq. From there it spread throughout the 

Near East (by the tenth century) and to Spain (by the thirteenth century) via 

North Africa (1). By the twelfth century it was being used to opacifY glass. 

It has been identified in twelfth-century Byzantine colored mosaic tesserae 

in Tchernigov, Ukraine (2). In 1612, in the first book devoted to glassmaking, 

Neri describes the preparation of enamel by adding calcined tin to calcined 

lead (3). The earliest dated European tin white glass is late fifteenth-century 

Venetian (4). White glass and enamel of the succeeding centuries is almost 

always opacified with tin (5). 

Writing in the late thirteenth century, Eraclius gives two very similar recipes 

for white glaze used for earthenware (6) . White glass was ground very finely, 

mixed with sulfur, painted onto the ceramic and fired. Tin is not mentioned 

but must have been the opacifier. The purpose of the sulfur is not known. It 

would not have survived the firing. 

The Paduan manuscript written in Venice in the seventeenth century (but 

copying sixteenth-century material), has another similar recipe, but here it is 

for a pigment: "Un bianco bellissmo-Si piglia cristallo di Venetia ..." [take 

some powdered Venice glass, add to it a third part of powdered sulfur] (7). 

The mixture was heated to red heat in a pipkin, cooled, and ground. Merrifield 

suggests that it was used for painting miniatures (8). 

Harley quotes an English source (ca. 1500): "For to make Ceruse. Take plates 

of tinne and beate them as thinne as thowe maist ..." (9). The tin was hung 

in a sealed barrel with vinegar for several weeks. The method is exactly the 

same as that used to produce lead or flake white. Indeed, the name ceruse 

was applied in England to both tin and lead white, and perhaps more correctly 

to a mixture of lead white and chalk. 

Harley says that documentary sources indicate that tin white was used in 

manuscripts. Her suggestion that it became obsolete because manuscript illumination 

declined cannot be substantiated, as the pigment has not been 

identified in any English or European miniatures. In the seventeenth century, 

Van Dyck tried it in oil and reported that it had insufficient body and was 

only useful for manuscript illumination (10). My tens fo und that it blackened 

in sunlight, spoiled white lead if the two were mixed, was useless in oil, and 

also in distemper if exposed to air (11). In the late eighteenth century, France 

Guyton de Morveau experimented with pigments and reported that tin white 

was unsuitable as it tended to yellow or blue (12). 

70 


Tingry, in 1830, gives a recipe for "Another Cremnitz White," which he 

describes as a beautiful pearly white, too expensive for house painting, but 

"it would, no doubt, be attended with great advantage in painting pictures" 

(13). This white was a mixture of tin white, one-fourth part zinc white and 

one-eighth part white clay separated from Briancon white. 

Linton describes tin white (oxide of tin) as, "too feeble in body ... to be of 

any service to the oil painter . .. " (14). He does, however, add that it may be 

unaffected by "injurious gases." Field also thinks the pigment to be poor, 

writing that it "dries badly and has almost no body in oil or in water, it is 

the basis of the best white in enamel painting" (15). In 1951, Mayer merely 

states that it is not a paint pigment at all (16). 

Figure 1. Tin white pigment, details on a 

sixteenth-century Jain miniature. Photograph 

by Paul Robins (Photo Studio), courtesy of 

the Victoria & Albert Museum (I. S.84- 

1963 1 15). 

Tin white, therefore, has two main uses: as an opacifier in glass, enamel, and 

ceramic glazes from the ninth to twentieth centuries; and as a possible pigment 

in manuscripts and miniatures from the fifteenth to seventeenth centuries. 

Even in the short review given here, three quite different recipes for 

its preparation are given from the following eras: (1) circa 1500, England, tin 

oxide; (2) circa 1580, Italy, white glass powder opacified with tin oxide; and 

(3) circa 1800, England, a mixture of tin oxide, zinc oxide, and white clay. 

During a research program in the laboratory of the Victoria & Albert Museum, 

methods for nondestructive identification of the pigments on Indian 

miniatures were studied (17). Incident light microscopy, energy-dispersive x 

ray fluorescence (EDXRF) spectroscopy, and ultraviolet and infrared color 

reversal photography were used. 

Several Jain miniatures were examined; these small, jewel-like, utterly distinctive 

paintings from Western India were of the fifteenth and sixteenth centuries. 

The areas of white paint were frequently restricted to details of textiles 

and jewelry (Figs. 1 , 2). Two of the miniatures had tin present as the major 

constituent in areas of white pigment. Because red lake and, in one case, gold 

leaf were beneath the white, this was regarded as interesting but not conclusive 

of tin white being present. 

Tin was then fo und on a third Jain miniature, possibly dating from the fifteenth 

century, with carbon black and verdigris in the same area. Three more 

paintings were chosen from the same manuscript as one of those examined 

earlier, as it was possible to focus on areas with no other pigment or only 

gold leaf. In all three, tin was the major constituent (Fig. 3) . It is possible, 

Figure 2. Tin white pigment, details on a 

sixteenth-century Jain miniature. Photograph 

by Paul Robins (Photo Studio), courtesy of 

the Victoria & Albert Museum (I. S. 46- 

19591 26). 

500 

450 

400 

350 

Sn 

300 

250 

200 

Sn 

Sn 

Au 

Au 

150 

Sn 

100 

50 

Au 

5 

10 

15 

20 

KeV 

25 

30 

35 

40 

Figure 3. EDXRF spectrum of tin white pigment on a sixteenth-century Jain miniature. Spectrum 

prepared by David Ford, Science Group, Victoria & Albert Museum (1. 5.46-19591 49). 

Darrah 71

therefore, to say that the pigment is a tin white (Table 1). A mixed tin-lead 

white was identified on a seventeenth-century Hindu miniature; much overpainting 

with zinc white confused the picture. 

It has not been possible to determine the constitution of the Indian tin white; 

the size and delicacy of the miniatures prohibits the removal of a sample. The 

pigment is a clean, brilliant white and appears reasonably opaque. It is probably 

tin oxide. 

Although the date of these miniatures falls into the same period in which 

the pigment is said to have been used in Europe, it seems unlikely that the 

Jain school was using the Western pigment. It is probable, however, that it 

developed from the use of tin oxide in ceramics, ultimately deriving, as it did 

in Europe, from the Near East or neighboring Persia. 

Burnt green earth 

Table 1. White pigments on jifleenth- to 

eighteenth-century Jain miniatures. 

Accession No. Da" Pigments 

IS 2-1972 c. 1460 Kaolin, mica 

Isacco 1 

15C Calcium white 

Isacco 2 15C Tin while 

IS 46-1959 f26 16C Tin white 


IS 46- 1959 f47 16C Tin white 


IS 84-1963 fl5 16C Tin white 

IS 82·1963 16C Hindu Calcium white, mica 

Private owner 17C Hindu Tin - lead white 

IS 2-1984 18C Lead white 

Green earth in its various fo rms has been used as a pigment through much 

of Eurasia fo r 2,000 years. Vitruvius wrote of it in the first century B.C.E. 

Burnt green earth seems to be mentioned first in the sixteenth or seventeenth 

centuries. The Paduan manuscript lists it as a color for miniature painting and 

also records that "the shadows of the flesh are made with terra 'ombra, terra 

verde burnt, and asphaltum" (18). The Volpato manuscript gives the method 

of preparation (19). Merrifield quotes Lomazzo, who directs that shadows on 

flesh should be made with burnt terra verde and nero di campana or umber 

(20). She says that "modern writers do not mention this colour, but the use 

of it has been revived by an eminent English artist, under the name of 'Verona 

Brown' "(21). 

Linton mentions it briefly with terre verte: "When calcined, it forms another 

beautiful pigment called Verona Brown" (22). 

To ch says Verona brown is a "fancy name" fo r a mixture of burnt umber and 

burnt or raw sienna (23). At the time of the Constable and Turner research 

projects in England, the author examined two paint boxes of the relevant 

period in the Victoria & Albert Museum collections. One was said to have 

Figure 4. Paint box, said to have belonged to William Turner, with bladders of oil paint, including 

Gebr. Terra di Verte and Gebr. Griine Erde. W. 65- 1920. Courtesy of the Victoria & Albert MuseU/t!. 

72 


elonged to William Mallord Turner (Fig. 4) . It contained seventeen bladders 

of solidified oil paint, all of which had been opened and sealed with a tack. 

Each was labeled in German or English. Two were named as Gebr. Terra di 

Verte and Gebr. Grune Erde, the term "Gebr." coming from the German 

gebrannte for burnt or roasted (24). 

EDXRF analysis of the paint on the surface of the bladders showed iron to 

be the major constituent with traces of manganese and titanium. The Terra 

di Verte contained a trace of calcium, the Griine Erde considerably more 

with traces of potassium, rubidium, and strontium (frequently present in calcium 

deposits). Green earth is a complex silicate colored by iron with a 

structure similar to mica (25). 

Dispersions of the two paints were made. The Gebr. Terra di Verte was a 

rich orange-brown comparable to burnt sienna but more translucent. The 

Gebr. Griine Erde was similar to raw sienna, but of a greener tone. This must 

have been burned at a lower temperature, as a few green particles remained. 

Under the microscope, the pigments were identical to green earth, with the 

overlapping plates of the crystals visible on the larger particles. Calcite was 

seen in both, but the Gebr. Griine Erde had, as indicated by EDXRF, a higher 

proportion of it. The labels suggest that both bladders were prepared in Germany, 

perhaps from two different sources of green earth. 

The appearance of the samples brought to mind the unidentified brown pigment 

seen by the author in several nineteenth-century landscape paintings, 

usually mixed with Prussian blue, ochres, and so forth, to produce greens and 

browns. It was suggested that it was the brown seen in the cross sections from 

a painting by Constable then being examined. SEM-EDX analysis of the latter 

at the National Gallery laboratory produced a spectrum identical to that for 

green earth. It was later identified in several Constable paintings dating from 

1811 to 1829 mixed in greens and browns (26, 27). Could Merrifield's "eminent 

English artist" be John Constable? 

The author has tentatively identified burnt green earth in paintings by Peter 

de Wint and J. F. Millet (28). 

Green earth was rarely used in England. The author has seen it only in 

seventeenth- and eighteenth-century wall paintings (oil) and cartoons by Verrio, 

Laguerre, Thornhill, and Robert Adam, all of whom trained in Italy or 

France. Verona brown seems to have been adopted in the early nineteenth 

century as a translucent addition to the earth and organic browns then available. 

Verona was a source of one of the better green earths, but was abandoned 

earlier this century (29). 

By coincidence, Constable Project researcher Sarah Cove visited Brussels and 

brought the author a bottle of pigment, Griine Erde Gebr., from an artists' 

suppliers. This modern sample is a darker, duller brown; perhaps burned at a 

higher temperature than the earlier examples, it too contains calcite. 

Brown pigments tend to be neglected, partly, no doubt, because of the difficulty 

in distinguishing the multitude of ochres, organic earths, and lakes. 

Burnt terra verte has a quite distinctive appearance, is easily identifiable by 

EDX, and may be more common than previously thought. 

Smalt 

The earliest blue glass colored with cobalt is from Eridu, Mesopotamia, circa 

2000 B.C.E. Recipes survive from Ashurbanipal's library in Nineveh, circa 

650 B.C.E. The Indians adopted Sumerian technology and were making cobalt 

blue glass by the sixth century B.C.E. (30). The Egyptians were using 

cobalt by circa 1400 B.C.E. The Romans were familiar with it; it was common 

in Western Europe in the seventh century and occurs in the Sassanian and 

Islamic periods (31). 

The first appearance of smalt is in a wall painting (ca. 1000-1200 C.E.) in 

Khara Khoto, Central Asia, and in the Church of Our Saviour of the Mon- 

Darrah 

73

Table 2. Unusual occurrences and constituents of smalt and blue enamel. 

Source Accession Blue pigment Date 

Number 

or enamel 

Minor 

EDXRF Analysis 

Trace 

BasohJi miniature 1.5.50-1953 smalt 1660-70 

1.5.51-1953 small 1660-70 

I.M.87-1930 small 1730-35 

I.M.88-1930 smalt 1730-35 

Bundi miniature 0.379-1889 small c.I770 

Venetian jug 273-1874 enamel 1472-1525 

tazza 5496-1859 enamel late 15th C. 

bowl C.170-1936 enamel 1521-23 

bowl C. I60-1936 enamel early 16th C. 

bowl 5489-1859 enamel 1500-1600 

S. German oil painting C.I.A. No 3021 small 15th - 16th C. 

Pb' 

As Ph· 

As 

As 

Pb' 

As 

As 

As 

As 

As 

KCa Fe Co Ni Cu As 

KCa Mn Fe Co Ni Cu Zn Bi 

KCa Fe Co Ni Ph Bi 

KCa Fe Co Ni Zn Pb Bi 

KCa Ti Fe Co Ni Cu As 

KCa Mn Fe Co Ni Cu Zn Pb Bi Sr Sn 

KCa Mn Fe Co Ni (Au) Ph Bi Sr 

KCa Ti Mn Fe Co Ni Cu Ph Bi Sr 

KCa Mn Fe Co Ni Cu Zn Ph As Bi Sr (So Sb)t 

KCa Mn Fe Co Ni Cu (Au) Ph Bi Sr Sn Sb 

Si K Fe Co Ni Pb. Bi 

• lead from lead white pigment 

t tin and antimony from yellow enamel applied over blue 

* Courtauld Institute of Art, SEM-EDX analysis 

astery of Chora (Kariye Camii), Constantinople (1325-1453 C.E.) (32, 33). It 

is likely that the smalt derived from Middle Eastern and Near Eastern blue 

glass. 

The earliest known source of cobalt ore is Khashan, Persia; the Mesopotamians 

and Egyptians probably obtained it from there. This suggests that the 

ore's coloring properties were already known to the Persians. Medieval European 

glass was tinted with Damascus pigment or zaffre, the Arabic name 

for cobalt oxide, again suggesting a Near Eastern source (34). 

Cobalt oxide, known as sulimani, was imported into China from Persia in the 

Tang period; by the fourteenth century it was transported by sea from the 

Persian Gulf via Sumatra. Muslim merchants residing in China influenced 

ceramic designs; much of the blue and white ware was produced fo r the 

Islamic market (35). 

Cobalt blue glass and glaze was known from China to Western Europe and 

yet throughout this period (3000 B.C.E. to ca. 700 C.E.) the most important 

blue pigment was Egyptian blue, a fr it colored with copper, at its best rivaling 

azurite but often appearing in paler turquoise shades. The terminal date is 

circa 850 C.E. on a fresco in the church of San Clemente in Rome (36). It 

seems unlikely that the secret of making Egyptian blue was lost to the 

Romans during the turmoil of the Teutonic invasions, as glass and enamel 

colored with copper continued to be made both in Italy and many other 

countries. 

In Europe there is an inexplicable hiatus in the use of blue pigments deriving 

from glass from about 850 to 1490 C.E. The Venetians were making cobalt 

glass by the mid-fifteenth century; the earliest references to smalt are by Leonardo 

da Vinci and Perugino in the 1490s (37). It has been identified on 

an altarpiece (1493) by Michael Pacher (38). Cobalt was discovered in Saxony 

in the mid-fifteenth century and fully exploited by around 1520. This may 

explain the greater utilization of smalt in the sixteenth century (39, 40). 

Cobalt ores combine iron with nickel and/or arsenic, the latter volatilized in 

the smelting process. Cobalt was also obtained from the residue in the separation 

of bismuth (41). The metal oxides were fused with sand and potash 

to produce zaffre (the Arabic name still being used) and sold to glassmakers. 

During research on Indian miniatures, the author found smalt on one Bundi 

example (ca. 1770 C.E., central India, Hindu). The pigment, not previously 

identified on Indian paintings, contained cobalt, iron, nickel, and a little arsenic 

(Table 2) (Plate 13). 

In a later project examining miniatures from the northern Hindu states, smalt 

was fo und on four miniatures from Basohli, a tiny state in the Himalayan 

foothills north of Lahore (42, 43). In one example (1660-1670 C.E.), smalt 

was used to paint areas of the sky and Krishna's skin; the pigment contained 

74 


500 

450 

As 

Ph 

As 

400 

350 

Fe 

Co 

300 

Ni 

250 

200 

150 

Bi 

Ph 

Bi 

100 

50 

0 

0 2 4 

6 

8 

10 

12 

14 16 18 20 

KeV 

Figure 5. EDXRF spectrum of smalt with a high arsenic content and traces if bismuth and lead. 

Krishna and Girls, 1730-1735, Basohli. Spectrum prepared by David Ford, Science Group, Victoria 

& Albert Museum (I.M.87-1930). 

cobalt, iron, nickel, and a trace of arsenic. On another of the same date and 

on two later miniatures (1730-1735 C.E.), a very different type of smalt was 

fo und in the sky and Krishna's skin (Plate 14). EDXRF spectra showed a 

high arsenic content with iron, cobalt, nickel, bismuth, and lead (Fig. 5). No 

orpiment (arsenic) was present on these paintings. 

By coincidence, the author had been analyzing a number of enameled Venetian 

glass vessels (44). Blue enamel, opacified with tin oxide, occurs on 

many of these objects. In a group of five, all late-fifteenth to early-sixteenth 

century, the blue enamel appeared to be opacified with arsenic; bismuth and 

lead were also present (Fig. 6) . Lead oxyarsenate, 3Pb3(As04)2'PbO, is said to 

have first been used as an opacifier in European heavy lead glass in the eighteenth 

century (45). Neither the Venetian enamel nor the Basohli smalt seem 

500 

450 

400 

Fe 

As & Ph 

350 Co 

300 

250 Bi 

Bi 

200 

150 

Ca 

Sr 

100 

50 

2 4 6 8 10 12 14 16 18 20 

Figure 6. EDXRF spectrum of cobalt blue enamel with a high arsenic content and traces of bismuth 

and lead. Venetian eflamelled glass bowl, 1521-1523. Spectrum prepared by David Ford, Science 

Group, Victoria & Albert Museum (C. 170-1936). 

KeV 

Darrah 75

to contain sufficient lead to fulfill the requirements of the above formula or 

for a heavy lead glass. 

In Venice, there is blue enamel opacified with arsenic 250 years before it is 

known to have been used to opacifY glass. In Basohli two centuries later, a 

pigment previously unknown in India was used and in three cases; it is very 

similar to the Venetian enamel. The enamels are of exactly the same period 

as when smalt first appears in European paintings, but in no example yet 

identified does it contain arsenic and bismuth. Do we have two completely 

independent discoveries, or were both importing enamel fr it from some Middle 

Eastern source that was using arsenic as an opacifier earlier than the 

Venetians? The author has been unable to find any analyses which may hold 

the answer to the sources of Persian or Middle Eastern enamels of the relevant 

period. 

As a footnote, smalt contammg a trace of bismuth was found on a south 

German oil painting (ca. 1400-1500 C.E.) in the collection of the Courtauld 

Institute of Art (46). In this case, it seems likely that the cobalt oxide colorant 

was obtained from the residue of the smelting of bismuth (as described previously), 

and that it is an accidental constituent. 

Conclusion 

Trade, industry, and art are ancient, ubiquitous, and international. This paper 

has endeavored to show that "new" pigments can still be found in unexpected 

places. The connections and coincidences that become apparent to the scientist 

through laboratory testing indicate historical trade and industrial developments, 

and may sometimes ultimately depend on the craftsperson who 

experiments with new materials. Our ability to understand these connections 

and coincidences requires knowledge of metalworking, ceramics, glass, and 

enamels to explain these connections and coincidences both in time and 

place. 

Notes 

1. Watson, 0. 1994. Personal communication. Ceramics and Glass Collection, Victoria 

& Albert Museum, London SW7 2RL. 

2. Turner, W E. S., and H. P. Rooksby, 1959. A study of the opalising agents in 

ancient opal glasses throughout three thousand four hundred years. Glastechnische 

Berichte 32K (VIII): 17-28. 

3. Neri, A. 1612. L'Arte Vetraria. Florence. 

4. Turner, op. cit. 

5. Darrah, J. A. 1992. Venetian and Facon de Venise glass. Unpublished analyses. 

London: Victoria & Albert Museum. 

6. Merrifield, M. P. 1849. Original treatises Jrom the XIIth to the XVIIlth centuries on 

the arts oj painting. John Murray, London, 200-5. 

7. Merrifield, op. cit., 704-5. 

8. Merrifield, op. cit., cliii. 

9. Harley, R. 1982. Artist's pigments c. 160(}-1835. 2nd ed., Butterworth Scientific, 

172-73. 

10. Harley, op. cit. 

11. Ibid. 

12. Ibid. 

13. Tingry, 1830. The Painter's and Colourman's Complete Guide. 3rd ed. London: 

Sherwood, Gilbert and Piper, 45. 

14. Linton, W 1852. Ancient and Modem Colours Jrom the Earliest Periods to the Present 

Time. Longman, Green, Brown and Longman. London, 55. 

15. Field, G. 1890. The Rudiments oj Colours and Colouring. Strahan and Co. London, 

63. 

16. Mayer, R. 1951. The Artists Handbook oj Materials and Techniques. London: Faber, 

69. 

17. Isacco, E., and J. A. Darrah. 1993. The ultraviolet and infrared method of analysis. 

A scientific approach to the study of Indian miniatures. Artibus Asiae, LIII (3/ 

4):470-91. 

18. Merrifield, op. cit., 650-5 1, 700. 

76 


19. Merrifield, op. cit., 745-6. 

20. Merrifield, op. cit., ccxxi-ccxxiii. 

21. Merrifield, op. cit. 

22. Linton, op cit., 64. 

23. Toch, M. 1911. Materials fo r Permanent Painting. Constable & Co., 169. 

24. The paint box (W65-1920) also contained Krapplack, Pariser blau, verdigris, 

Neapelgelb, Prussian blue, beirnschwarz, lichter ocker, smalt blau, kernschwarz, 

Cremserweiss, chrome yellow, Indian red, and two red ochres and a yellow ochre 

(with no labels). 

25. Grissom, C. A. 1986. Green earth. In Artists' Pigments. Ed. R. L. Feller. Washington: 

National Gallery of Art, 141-67. 

26. Cove, S. 1991. Constable's oil painting materials and techniques. In Constable. L. 

Parris and 1. Fleming-Williams. London: Tate Gallery, 493-5 18. 

27. I am grateful to Sarah Cove (Constable Research Project) for drawing my attention 

to four fu rther references to burnt green earth and Verona brown in the 

latter half of the nineteenth century, unfortunately too late to include them in 

this paper. 

28. Darrah,]. A., Victoria & Albert Museum Conservation Department Reports:]. 

F. Millet, The Wood Sawyers, ca. 1850 (CAl 47);]. F. Millet, The Well at Cruchy, 

1854 (CAl 49);]. F. Millet, Landscape with Terminal Figure, ca. 1864 (CAl 172); 

and P. de Wint, A Woody Landscape, 1812-1816 (261-1872). 

29. Merrifield, op. cit., ccxxi-ccxxiii. 

30. Singh, R. N. Ancient Indian Class. Delhi: Parimal Publications, 53, 169-71. 

31. Vose, R. H. 1980. Class. Collins Archaeology, 30. 

32. Muhlethaler, B., and ]. Thissen. 1969. Smalt. Studies in Conservation (14) :47-61. 

33. Gettens, R.]., and G. L. Stout. 1958. A Monument of Byzantine wall painting: 

The method of construction. Studies in Conservation (3):107-18. 

34. Frank, S. 1982. Class and Archaeology. Academic Press, 23. 

35. Medley, M. 1976. The Chinese Potter: A Practical History of Chinese Ceramics. Phaidon 

Press, 176-78. 

36. Lazzarini, L. 1982. The discovery of Egyptian blue in a Roman fresco of the 

medieval period (A.D. ninth century) . Studies in Conservation (27):84-86. 

37. Merrifield, op. cit., ccvii. 

38. Grissom, op. cit. 

39. Gettens, R.]., and G. L. Stout. 1966. Painting Materials. New York: Dover, 157- 

59. 

40. Kuhn, H. 1973. Terminal dates for paintings derived from pigment analysis. In 

Application of Science in the Examination of Works of Art. Seminar proceedings. 

Boston, 199-205. 

41. Frank, op. cit. 

42. I should like to thank V. Sharma, visiting Nehru Fellow of the Shirnla Museum, 

India, for bringing these paintings to my attention. 

43. Archer, W G. 1973. Indian Paintings of the Punjab Hills. Vol. 1. London, 16, 21. 

44. Darrah, 1992, op. cit. 

45. Turner, op. cit. 

46. I am grateful to A. Burnstock, Conservation and Technology Department, Courtauld 

Institute of Art, London University, for allowing me to use the results of 

her analysis of St. John of Patmos (CIA 302). 

Darrah 77

Abstract 

A comprehensive examination was 

conducted on a set of twenty-three 

seventeenth-century Tibetan thangkas 

owned by the Museum of Fine Arts, 

Boston (BMFA). The examination 

was undertaken because of the 

dearth of technical information on 

Tibetan thangkas in Western literature. 

Infrared reflectography was used 

to document color notations drawn 

on the ground layer by the artist(s) 

as a guide fo r the artist(s) and apprentices. 

Samples were taken from 

areas displaying such notations and 

multiple analytical techniques were 

utilized to identify the pigments. 

The existence of more than one color-code 

system became evident 

when the color notations and the 

identified pigments from the BMFA 

set were compared to those found 

on other Tibetan paintings. Nine 

comparative paintings were examined, 

fo ur from the BMFA and five 

from the Los Angeles County Museum 

of Art. 

Figure 1. Detail IR-riiflectogram from The 

Buddha Shakyamuni Preaching at Dhanyakataka, 

showing color codes throughout an 

off ering bowl if jewels. Denman Waldo Ross 

Collection, Museum if Fine Arts, Boston 

(06.333). 

An Investigation of Palette and Color Notations 

U sed to Create a Set of Tibetan Thangkas 

Kate I. Duffy* 


Los Angeles County Museum of Art 

5905 Wilshire Boulevard 

Los Angeles, California 90036 

USA 

Jacki A. Elgar 

Asiatic Department 

Museum of Fine Arts 

465 Huntington Avenue 

Boston, Massachusetts 021 15 

USA 

Introduction 

Tibetan thangkas are scroll paintings that incorporate Buddhist iconography. 

The painting is done on cloth that is stitched into a framework of silk borders. 

Along the top edge is a wooden stave from which the painting is hung. Along 

the bottom edge is a wooden dowel around which the thangka is easily rolled 

for storage and transport. In 1906 the Museum of Fine Arts, Boston acquired 

a set of twenty-three thangkas (1). The paintings in this set had lost their 

original thangka format and were mounted on panels upon their arrival. 

Originally thirty-three or thirty-four paintings comprised the set, depicting 

the thirty-two Kings of Shambhala, the Buddha Shakyamuni, and possibly a 

Kalachakra Mandala. The existing set, referred to as the "Shambhala paintings" 

throughout this paper, contains only the Shakyamuni and twenty-two 

images of the Kings (Plates 15, 16). Over the years the set has been assigned 

various dates and places of origin. Today it is generally accepted to originate 

from late seventeenth-century Tibet. 

Initially, each Shambhala painting was surveyed by infrared reflectography 

(IRR), a nondestructive technique that enabled color codes on the ground 

layer to be viewed (2). The color notations are handwritten in Tibetan 

dbu. med script. In areas with complex juxta positioning of numerous colors 

(such as offering bowls containing multicolored jewels, or garments with intricate 

folds), notations were observed in abundance (Fig. 1). Once the infrared 

data were compiled, X-ray fluorescence (XRF) analyses of twelve to 

fifteen areas were carried out. After examining these results, approximately 

ten areas were sampled. Multiple analytical techniques were utilized, including 

polarized light microscopy (PLM), Fourier transform infrared spectroscopy 

(FTIR), electronprobe microanalysis (EPMA), X-ray diffraction (XRD), highperformance 

liquid chromatography (HPLC), ultraviolet/visible absorption 

spectrometry (UV /vis), and fluorescence spectrophotometry (FS). Black pigments, 

presumably carbon-based, were not studied. Decorative gold was also 

not examined except for initial XRF analyses. 

Preparing the support 

The Shambhala thangkas are all painted on cotton cloth supports. According 

to Jackson and Jackson, who documented the practices of living thangka 

painters, the painting support is typically first made taut by stitching it to four 

pliable sticks (3) . It is then laced into a larger wooden frame, leaving a space 

of several inches between outer and inner frames. This space is crucial for 

adjusting tension. The support is stiffened by sizing both sides with a gelatin 

solution. A mixture of finely ground white pigment and size solution is then 


78 


Table 1. Tibetan color names, translations, and identified pigments. 

COLOR 

white 

blue 

Iliaht blue 

OOTA1lQII ENGUSH SPEWNG TIBETAN COLOR NAME ENGLISH TRANSLATION 

ka dkar 0 white 

Iha mthina azurite 

n oo Isnoo Inoht blue 

PIGMENTS IDENTIFIED 

kaolin 246 

azurite (1 2 4,6\ 

azurite kaolin 2 3 4 6 

Iliehter blue 

blue/Qreen 

blue/oreen 1 

Iliaht blue/eraen t=. 

dark blue/ reen 

ellow 

ellow 

ellow " 

briaht red 

dark red 

ink 

ink 

oranae '" 

n kva snno skva IJinht blue + white 

loa IsoanoDO malachite 

azurite kaolin 2 3 4 6 

malachite brochantite (2.4) 

'= \ 'ana sb ar ra nn "comnounded nreen" 

ornimentlreal ar indiao kaolin 1 2 3 4 6 

to. , , 'ann sbvar Fana "comnnunded nreen" ornimentlreal ar indiQo kaolin (1 2 3 4 6) 

'2.. 'ana kva 

fana skva "comoounded reen" + white samole not taken 

'= \ IXQ") 'ana naa 

fann na "comnnunded teen" + black sam Ie not taken 

;:, 

.:) 

, 

'" 

fAA. 

"- 

0( "

Figure 2. Detail IR-refiectogram from The 

Twenty-ninth King of Shambhala, showing 

the color code ngo kya on the leg of a 

goat. Denman Waldo Ross Collection, Museum 

if Fine Arts, Boston (06. 335). 

The notation ngo (sngo), found on twenty-two Shambhala paintings, means 

light blue or sky blue (9). Ngo was used for sashes, linings of garments, jewels, 

nimbuses, rock crags, and lotus petals. Samples of this color taken from four 

paintings were identified as a mixture of azurite and kaolin. Further whitening 

of this pigment is indicated by the notation ngo kya (sngo skya) and is used 

to depict the color for water, a goat, the sky, and the skin color of a king (10, 

11, 12). This notation is shown in Figure 2. The additional character kya 

indicates that more than 50 percent of the color mixture is white. An actual 

quantitative measurement is difficult to carry out; however, one XRD pattern 

of a ngo kya sample indicated a higher kaolin:azurite ratio than an XRD 

pattern taken from a ngo sample. 

Greens. Basic green is denoted by pa (spang), an abbreviation fo r the Tibetan 

word meaning malachite (13). Pa, observed on all twenty-three Shambhala 

paintings, is used for rock crags, land masses, garments, jewels, and foliage. 

The green pigment was identified by FTIR and XRD in four samples as a 

mixture of malachite and brochantite. Since brochantite is associated with 

malachite deposits, the combination is probably a natural one (14). The mined 

source for malachite in Tibet was probably the same as previously mentioned 

for azurite (15). 

A mixed green was found on twenty-one Shambhala paintings. Jackson and 

Jackson describe a "compounded green" (sbyar ljang) derived from a mixture 

of orpiment and indigo (16). The notation uncovered is jang, occasionally 

written 'jang, and is used for nimbuses, mountains, lotus centers, and leaves. 

The color varies from deep blue to aqua to yellowish green. Five samples 

taken from five paintings were examined by FTIR, XRD, EPMA, and PLM. 

Examination of two of the five dispersed pigment slides revealed a mixture 

of realgar and clay. Two other slides contained a mixture of realgar, orpiment, 

and clay. The fifth contained a mixture of only orpiment and clay. Indigo, 

tentatively identified in three dispersed pigment slides, was positively identified 

by FTIR in one of the samples. UV Ivis spectrophotometry analysis of 

these samples is planned fo r the fu ture in hopes of definitively identifYing 

indigo. The chromatic differences found in the samples may be due to the 

fugitive nature of the indigo or to the discoloration of either realgar or orpimento 

Indigo may fade when applied thinly, especially when exposed to 

sunlight (17). Although each sample appears to be slightly different, it is 

thought that the original ingredients were the same: namely, clay, orpimentl 

realgar, and indigo. The notation jang kya mang skya), meaning light green, 

was revealed on two Shambhala paintings for decorative elements of a king's 

throne, clouds, and land masses (18). The color notation jang nag jang nag), 

meaning dark green, is documented on four Shambhala paintings and is used 

for leaves and land masses (19). The paint covering this notation as well as 

the jang kya has yet to be sampled. 

Yellow. Color notations fo r yellow (ser po) were documented on nineteen 

Shambhala paintings. The notations are actually written as one of the following 

three: se, ser, or sare. Se is the most abbreviated and sare is a misspelling, 

perhaps to facilitate writing. The pigment mixture for these notations was 

consistently fo und to contain orpiment mixed with kaolin and a small amount 

of red lead. Five samples from four paintings were taken from areas with sare. 

FTIR analyses indicate a kaolin-type clay, similar to the spectrum described 

for ka. Examination of four samples by EPMA identified silicon, aluminum, 

magnesium, and lead, as well as minor amounts of calcium, iron, and arsenic. 

PLM revealed clay particles mixed with spherical aggregates of red lead and 

fine to medium particles of orpiment. On one slide several large particles of 

realgar were identified. Realgar (arsenic disulfide) is often fo und in natural 

deposits with orpiment (arsenic trisulfide). Large deposits of orpiment exist 

near Chamdo in eastern Tibet and in the Yunnan Province of China (20, 

21). 

One of the samples taken from the area marked se was nearly colorless. Surface 

elemental analysis by XRF identified arsenic as a major element, yet neither 

80 


J 

Figure 3. Three-dimensional spectrum of na sample from The Twentieth King of Shambhala. 

Denman Waldo Ross Collection, Museum of Fine Arts, Boston (06.334). 

orpiment nor realgar was fo und by PLM. The FTIR spectrum showed kaolin. 

The remainder of the sample was analyzed by XRD. The majority of the 

peaks correlate with those of metahalloysite. The remaining peaks were assigned 

to kaolinite and arsenolite (As203), a white mineral. It is unlikely that 

arsenolite is original to the painting. The yellow orpiment probably faded 

and became a whitish arsenic trioxide. The impermanence of arsenic trisulfide 

has been documented elsewhere (22). 

Reds. Two shades of red are observed on the Shambhala paintings. One is a 

bright red, the other is muted. The bright red uses the notation ga (rgya mtshal) 

for Chinese or Indian vermilion (23). Ga is found on fourteen paintings and 

depicts a color used for garments, lotus petals, and nimbuses. PLM, XRD, and 

FTIR identified the color as a mixture of vermilion and kaolin. Under the 

microscope, the vermilion particles appear as finely ground spherical particles 

with a deep orange-red color. Several historical references are given by Jackson 

and Jackson which state that Tibetan painters had access to both synthetic 

mercuric sulfide and to the natural mineral cinnabar from China, India, and 

Tibet (24). 

The darker red is denoted by ma (dmar po), meaning red color (25). Ma is 

found on fourteen paintings and describes a color used for garments and 

sashes. The paint consists of a dual layer with a dark red organic coating over 

a red pigment layer consisting of vermilion and kaolin. Further analyses are 

pending and no conclusive identification has been made. 

Pink. The color pink is used mainly fo r clouds but also for jewels, buckles, 

makaras, and garudas (26). Pink is denoted by na or, occasionally, na kar, and 

represents lac dye (na rosy mixed with a white pigment (27, 28). Na is fo und 

on all twenty-three Shambhala paintings. Areas were sampled on six of the 

paintings and examination of dispersed pigment slides revealed a clay base 

mixed with a red dyestuff. FTIR analyses identified the clay as kaolin, as 

described earlier. Identification of the dyestuff was achieved through absorption 

spectrometry. The identification was fu rther confirmed by fluorescence 

.spectrophotometry, utilizing a method described elsewhere (29). The pink 

was identified as a lac dye. Figure 3 shows a three-dimensional plot of the 

result, which is characteristic of lac. Jackson and Jackson state that much of 

the dyestuff was traditionally gathered and prepared in Tibet (30). The dye 

is extracted from sticks encrusted with a resinous secretion produced by the 

lac insect, Kerria laeea Kerr.; the resin is still found today in the eastern Himalayas 

where the warmer climate is more conducive to its fo rmation. 

Orange. The color orange is represented by la, an abbreviation for the Tibetan 

word fo r minium i khri) (31). The notation was uncovered on garments, 

nimbuses, finials, crowns, flames, belts, jewelry, roof tiles, wheels, and vases. La 

was found on fifteen Shambhala paintings. Three sampled areas were examined 

by XRD, EPMA, FTIR, and PLM. The results identified a mixture of 

Duffy and Elgar 81

ed lead and calcium carbonate. Historical references given by Jackson and 

Jackson state that Tibetan painters did not use the natural mineral minium 

(32). Instead, the synthetic lead tetroxide was imported from China, Nepal, 

and India. Calcium carbonate was available in Tibet, particularly in Rinpung, 

an area north of Lhasa and the seat of government in the sixteenth century 

(33). In certain parts of Tibet, the cost of calcium carbonate was prohibitive 

and less expensive white pigments were often used. It is unknown why calcium 

carbonate was used to lighten the red lead. Other pigments were mixed 

with kaolin. The choice may be due to the purity and color of the calcium 

based pigment. 

Summary of palette and color notations for the Shambhala set 

Pigment identification determined that the notations on the ground layer 

directly correlate to the pigments used, supporting the belief that these notations 

were a guide for the artist(s) and apprentices. Pigments used for the 

respective color notations are similar throughout, which is not surprising given 

that the Shambhala paintings belong to one set. 

The artist's palette for the Shambhala paintings from the Museum of Fine 

Arts, Boston (BMFA) consists of pigments derived from minerals or synthetic 

mineral analogues with the exception of two organic dyes, indigo and lac 

(Table 1). The pigments identified in this investigation deviate little from the 

modern-day painter's palette, as documented by Jackson and Jackson. The 

pure blue and green colors were painted with coarsely ground, unadulterated 

mineral pigments. All other colors were mixed with a white pigment either 

to lighten the color or to achieve translucency. The orange color is unique 

in that it was created by mixing calcium carbonate with red lead. In all other 

cases kaolin clay was used, perhaps due to its availability and low cost. 

All color notations are handwritten in Tibetan dbu. med script on the ground 

layer. Some derivative colors are indicated by additional dbu. med characters 

to the root notation. Figure 2 is an infrared reflectogram depicting such a 

color notation. In this case, kya is added to the root notation ngo. The entire 

notation, ngo kya (sngo skya), indicates a light blue color. The second character, 

kya, is used to indicate whitening of the existing color, ngo. Another notation 

found indicates the darkening of an existing color. The notation has the 

additional characters na and ga, pronounced nag. These additions represent 

the Tibetan word for black (nag po). In the case of jang nag jang nag), nag 

indicates darkening of the existing green color, jang. 

Comparative paintings 

Comparative paintings from the Museum of Fine Arts, Boston and the Los 

Angeles County Museum of Art (LACMA) were examined using infrared 

reflectography (34, 35). Only a few pigment samples, however, were taken. 

The LACMA nineteenth-century commemoration thangka for the bhimaratha 

rite from Tashi Lhunpo monastery in central Tibet did not reveal any color 

codes. 

Figure 4. Detail IR-riiflectogram from A 

Mahasiddha and Taklungpa Lamas, showing 

the color code pkya on a landmass. Los 

Angeles County Museum of Art, Los Angeles 

(M. 81.206. 12). 

Another LACMA painting, A Mahasiddha and Taklungpa Lamas (ca. late 1700s) 

from Taklung monastery in central Tibet, revealed color codes written in 

Tibetan dbu. med script. Many of the notations were similar to those documented 

on the BMFA Shambhala paintings. One distinct difference, however, 

is in the manner of differentiating derivative colors. For a light green color, 

fo r example, the notation used was pkya (spang skya). A subjoined kya is added 

to the root character pa to indicate the addition of white (Fig. 4). Three other 

paintings examined using IRR revealed color codes in dbu.med script. These 

were Shakyamuni with Disciples and Dharmatala, two paintings from a set 

of five at the BMFA accepted as sixteenth-century eastern Tibetan; and 

LACMA's Portrait oj the Fifth Karmapa, accepted as originating in eighteenthcentury 

Kham, a region of eastern Tibet. 

82 


Figure 5. Detail IR-reflectogram from Shakyamuni 

and the Eighteen Arhats, showing 

a color code written in Chinese on an Arhat's 

back. Los Angeles County Museum of Art, 

Los Angeles (M. 83. 105. 18}. 

Shakyamuni and the Eighteen Arhats, owned by LACMA and thought to be 

an eighteenth-century work from the Kham region, has color codes on the 

ground layer written in Chinese. Unlike the brilliant white magnesite ground 

found in the Shambhala paintings, this ground is buff-colored and composed 

of hydrocerussite and kaolinite. Two color-code systems are employed. The 

first is numerical: numbers represent specific colors. The second uses Chinese 

color names or idioms. Under the dark blue pigment of a begging bowl, IRR 

revealed the Chinese character for the number seven. The pigment was identified 

by PLM and FTIR as azurite. On an Arhat's back, IRR revealed the 

Chinese character for hulled rice (Fig. 5). Examination of a dispersed pigment 

slide revealed orpiment mixed with a small amount of red lead. A character 

not yet translated was uncovered on several deep red colored areas. Examination 

of a dispersed pigment slide revealed vermilion mixed with a small 

amount of red lead. 

Another LACMA painting with Chinese color codes is Palden Remati and 

Her Retinue, accepted as originating from the Gelukpa monastery in Central 

Tibet, 1800-1850 C.E. This painting employs the two color-code systems 

described earlier with similar notations. The Chinese numbers for three and 

seven were used to denote a light blue color. This color has yet to be sampled. 

Presumably "seven" represents azurite and "three" represents the second addition 

of a white pigment. 

Two BMFA paintings displaying Chinese color codes come from a set of five 

entitled Stories from the Life of Buddha. The set is thought to be eighteenthcentury 

Tibetan. Again, evidence of the two color-code systems described 

earlier as well as notations fo r derivative colors were found on these paintings. 

The Chinese numbers for two and six are used to denote a light green. This 

color has yet to be sampled. Presumably "six" represents malachite and "two" 

the first addition of a white pigment. 

Conclusion 

Since thangka painting is a tradition passed on from master to apprentice, 

determining palette and deciphering color-code systems may prove to be 

helpful in the identification of specific workshops or painting lineages. The 

color-code system used on the BMFA Shambhala paintings has several distinctive 

traits. These include an additional character kya (skya) for whitish 

tints (Fig. 3) and the additional characters na ga (nag) for darker tints. The 

artist of the painting A Mahasiddha and Taklungpa Lamas indicated whitish 

tints with a kya subjoined to the root notation (Fig. 4). This system for 

distinguishing derivative colors is similar to one still practiced by some modern 

Tibetan thangka painters. The paintings with color codes in Chinese 

script pose many questions which are beyond the scope of this paper. In terms 

of the ethnic background of the creators of these works, one can only speculate. 

Perhaps the painters were Chinese, since a native Tibetan speaker would 

be unlikely to write such "private" communications in a fo reign script. On 

the other hand, a bilingual Tibetan might use Chinese characters to communicate 

with Chinese apprentices or coworkers. Likewise, one can only 

speculate on whether such paintings were produced in Tibet or China. In 

order to answer these questions, more infrared data and pigment analyses need 

to be compiled from thangkas of known Tibetan as well as Chinese origin. 


The authors are indebted to the following people at the Museum of Fine Arts, 

Boston: Rhona MacBeth, John Robbe, Darielle Mason, Anne Morse, Wu Tung, Arthur 

Beale, and John Elwood. The authors wish to thank the staff at the Los Angeles 

County Museum of Art for their generous support, particularly Victoria Blyth-Hill 

and John Twilley. For his analytical contributions, the authors are grateful to Arie 

Wallert of the Getty Conservation Institute. The authors wish to thank the following 

Lamas and thangka painters for their help in deciphering color codes: Archung Lama 

and Gega Lama of Kathmandu, Nepal; Megmar, Alexander Kocharov, and Chating 

Jamyang Lama of Dharamsala, India. For his enthusiasm, support and analytical con- 

Duffy and Elgar 83

tributions, the authors would like to acknowledge Richard Newman at the Museum 

of Fine Arts, Boston. 

Notes 

1. Ross, D. W 1904. The purchase of twenty-six Tibetan paintings from Mme. 

Tangeveil, Paris. Unpublished typed letter to M. S. Prichard. Boston: Museum of 

Fine Arts. 

2. An infrared TV camera equipped with a Hamamatsu No. 2606-06 vidicon was 

used at the BMFA. 

3. Jackson, D., and J. Jackson. 1984. Tibetan Thangka Painting: Methods & Materials. 

London: Serindia Publications, 15-23. 

4. Palache, c., H. Berman, and C. Frondel. 1951. Dana's System of Mineralogy. Vol. 

II. New York: John Wiley and Sons, Inc., 165. 

5. Shaftel, A. 1986. Notes on the technique of Tibetan thangkas. Joumal of the 

American Institute for Conservation 25 (2):98. 

6. Jackson, op. cit., 174. 

7. Ibid., 175. 

8. Ibid., 75. 

9. Kocharov, A., and ChatingJamyang Lama. 1993. Personal communication. Dharamsala, 

India: Library of Tibetan Works & Archives. 

10. Ibid. 

11. Archumg Lama and Gega Lama. 1993. Personal conmumication. Kathmandu, 

Nepal. 


13. Ibid. 

14. Palache, op. cit., 543. 


16. Ibid., 176. 

17. Gettens, R. J., and G. L. Stout. 1966. Painting Materials: A Short Encyclopaedia. 

New York: Dover Publications, Inc., 120. 

18. Kocharov, op. cit. 

19. Megmar. 1993. Personal communication. Dharamsala, India: Library of Tibetan 

Works & Archives. 


21. Palache et aI., op. cit., 268. 

22. FitzHugh, E. W n.d. Orpiment and realgar. In Artist's Pigments: A Handbook of 

their History and Characteristics, Vol. III. Forthcoming. 


24. Ibid., 80. 

25. Ibid., 174 

26. Makaras are mythological sea creatures. Garudas are guardians of the sky (part 

bird and part human). 


28. Kocharov, op. cit. 

29. Wallert, A. 1986. Fluorescent assay of quinone, lichen, and redwood dyestuffs. 

Studies in Conservation 31 (4):145-55. 


31. Ibid., 93. 

32. Ibid., 81. 

33. Ibid., 82. 

34. Paintings at LACMA were surveyed with an infrared TV camera, Quantex QVC 

2500, equipped with a bandpass interference filter (1.6 micron wavelength) to 

eliminate chromatic aberrations of the lenses. 

35. Kossolapov, A. J. 1993. An improved vidicon TV camera for IR-reflectography. 

ICOM Committee fo r Conservation Preprints, 25-31. 

84 

Histon:cal Painting Techniques, Materials, and Studio Practice

Abstract 

Icons preserved in the Nile valley 

are the least known of Egypt's antiquities. 

In the gap between the early 

Coptic icons from the fourth to 

sixth centuries and those of the second 

half of the eighteenth century, a 

recently discovered group of medieval 

icons that are difficult to date 

and attribute has assumed prominence. 

Two beam icons, painted on 

similarly made panels of sycamore 

timber, were selected to describe in 

detail. Even preliminary research on 

their technology has resulted in new 

material for the study of icon painting 

and the continuity of historical 

Egyptian techniques, materials, and 

mythology into Christian times. 

Figure 1. Detail of the reverse side if beam 

B. Photograph by Z. Skalova. 

New Evidence for the Medieval Production of Icons 

in the Nile Valley 

Zuzana Skalova 

Foundation for the Preservation of Icons in the Middle East 

Conservation of Coptic Art Project 

12 Tolombat Street, Apt. 17 

Garden City, Cairo 

Egypt 

Introduction 

Even the earliest preserved Coptic icons demonstrate that they were created 

for monastic circles. They repeatedly portray local saints (often monks) or 

they express the doctrines of Coptic religious thought. This monastic patronage 

remained constant. 

Many Coptic icons were until recently in a poor state of preservation. Today, 

only a small number of icons in the Nile valley are restored to their full 

advantage (1). Consequently, few art historians have looked for old icons or 

recognized them (2) . However, important medieval icons hanging in the 

church of St. Mercurius Abu's-Saifain in Old Cairo were admired and quite 

correctly assessed by Alfred Butler during the 1870s. Although they were 

"dim with age and indistinguishable," he concluded that "the icons generally 

speaking are ancient and well executed." Butler's comments on the disfIgurement 

of the pictures caused by remarkably careless technology, when compared 

with Italian panels, can serve as the point of departure fo r this study 

(3) . 

During the last five years, the author has fo und, studied, and in some cases 

restored, some twenty medieval icons, which can be preliminarily dated between 

the thirteenth and the fifteenth centuries. They are nearly all large 

icons, clearly made for public veneration in Eastern churches. Some of these 

reflect distinct Coptic patronage through their iconography (4). 

In assembling and attributing the group, the author relied primarily on technical 

aspects emphasizing the icons' kinship, which clearly shows that they 

emerged from the same local tradition of workmanship. Common features 

include the use of indigenous wood (usually sycamore), the awkward construction 

of the panels, a ground layer containing anhydrite, a limited range 

of pigments, the use oflow-quality azurite blue, an unburnished golden background, 

and a thin gray layer of varnish (5). 

Stylistic and iconographic aspects are more hybrid, but also reflect a rather 

peripheral and culturally mixed background. This suggests that local craftsmen 

and foreign painters were working together in Egypt. Thus, though these 

paintings are varied in style, the panels have such consistent parallels in the 

choice of wood and peculiar carpentry, that they can be classified together. 

They are manufactured from narrow, roughly assembled planks held together 

with huge traverses and narrow boards, and nailed with big iron nails driven 

in from the front. The rifts between the planks are filled on both sides with 

plaster and covered with palm bark fiber, and sometimes textiles, to smoothen 

the surface (Fig.1). When compared with panels attributed to the Greek icon 

workshops, they appear clumsy and extremely heavy to transport, a fact that 

would additionally testifY to local provenance. The choice of omnipresent 

sycamore wood might have been dictated by scarcity and the costs of more 

suitable material in the Nile valley (6) . 

The finish of the back sides of these paintings is also characteristic for the 

group. They were invariably covered with a thick layer of plaster, decorated 

with alternating lines of wavy pink-brown and gray-blue brush strokes. This 

is a rare feature in icon painting, and it may suggest the formula of a work- 

Skalova 85

shop. Dissemination of such formulae may even have caused decoration of 

this type to occur on the back sides of icons in the collection of St. Catherine's 

Monastery at Sinai (7). However, these Sinaitic panels are skillfully 

made from imported wood; and they are analogous to many Byzantine icons 

preserved in Europe, rather than to the Coptic icons (8). 

The icon supports in the Nile valley must have been made according to a 

different studio tradition. According to their technology, they clearly fit into 

the context of native (presumably Coptic) practice. Furthermore, the majority 

of the previously mentioned technical characteristics can be traced back to 

the industries and materials of Graeco-Roman and Pharaonic Egypt (9). 

Two medieval beam icons 

Figure 2a, b. One of two cuts, top, showing 

medalliollS of all arChat1gel alld an Old Testament 

prophet; m'ld Blessing Christ, depicted 

ill the central medal/ioll, below. Photographs 

by Z. Skalova. 

Two unique beam icons in the church of St. Mercurius Abu's-Saifain may be 

singled out for discussion. This church was the seat of the Coptic patriarchy 

in the Middle Ages and it is likely that both sacred pictures were made for 

it (10). 

The first icon, The Virgin with Child Enthroned between Archangels, Nine Church 

Fathers and Nine Coptic Monks, measures 44.5 X 246.5 cm, and is shown in 

Plate 17 and Figures 3 (left) and 4a, b. The second icon, Six Equestrian Saints 

(originally ten saints), measures 45 X 207.5 cm and is shown in Plates 18a, 

b and in Figures 2a, b; 3 (right); and 4c-e. The two icons will be referred to 

here as beam A and beam B, respectively. 

Thanks to Butler's description and drawing, the conservation history of these 

two icons can be traced back more than a century (Fig. 3). Both icons were 

repeatedly restored. Beam A is preserved in its original form, which has 

helped in reconstructing the structurally altered beam B. Butler counted only 

four horsemen in beam B (Fig. 3, right). Today, remarkably, beam B consists 

of six horsemen. Some time in the past, this longitudinal icon was cut into 

pieces. Meanwhile, two more horsemen from the original beam were added 

to Butler's fragment (Figs. 2a, 4c). 

The equestrian saints are depicted in sculptured and gilded arches that carry 

eleven alternating medallions with Old Testament prophets, archangels, and 

the blessing Christ. Each horseman is identified by inscriptions in Coptic and 

Arabic. 

Additionally, the symmetry of beam B is marred by the fact that the six holy 

horsemen are riding in conflicting directions. It is obvious that their initial 

order and number have been changed. Two cuts, visible in the restored panel, 

confirm this observation (Fig. 4c). Such an intervention must have been the 

work of a person who failed to understand the symbolism of this picture. It 

seems unlikely that these saints were arranged to gallop away from the Savior 

instead of toward him, as is proper in a hieratic composition. Christ, depicted 

en buste in one of the medallions, can be seen on Butler's fragment (Fig. 2b). 

If the placement of Christ's medallion is accepted as being in the middle of 

the beam icon, the number of horsemen comes to ten. When reconstructed, 

beam B should be about 350 cm long (Fig. 4e). Clearly, both beams were 

conceived together to form part of broader didactic program. 

A peculiar aspect of both paintings is their carpentry. They are assembled 

from irregularly cut horizontal pieces, three across the width and nine across 

the length (Figures 4b and 4d). Additional sculptured arches are nailed on the 

front. On the reverse, vertically placed traverses hold the planks together. Both 

\lnnooonnl llnoOOII 

FiJ!urc 3. Bea1l1 A (lift); bealll B (riJ!ht). Drawillg by A!fred BIIIler, 1884. 

86 


I )--------------------------1 I i 

(b) 

------ --- 

------- ------ 

---- 

-------------------- 

------- 

----.-- 

--- 

--- 

----- 

(c) 

t 

t ! 

1- - - 

--- 

--- 

1--- 

- - - --- 

(d) 

018 

(e) 

i 

FIgure 4a. Beam A, front. Drawing by Sameh Adly. 

Figure 4b. Beam A, reverse. Drawing by Sameh Ad/y. 

Figure 4c. Beam B, front. Drawing by Sameh Adly. 

Figure 4d. Beam B, reverse as preserved today. Drawing by Sameh Ad/y. 

Figure 4e. Beam B, recol1structed. Drawil1g by Sameh Ad/y. 

beam icons in the church of St. Mercurius Abu's-Saifain are not only constructed 

from planks of the same irregular dimensions, but the wood looks 

so similar that it could come from one tree. The wood was identified as 

sycamore (Ficus sycomorus L.): the traverses are made from cypress (Cupressus 

sempervirens var. horizontalis Gord.) (1 1). 

The Egyptian sycamore tree 

The peculiar shape and dimensions of the support planks of the beam icons 

A and B might have been dictated by the diameter of the tree trunk that 

provided the timber. It does not seem probable that the patron of such significant 

icons would economize on the timber. The choice of painter and 

quality of materials usually testifY to the resources and intentions of the client. 

For the art historian, the interest of this research lies not only in the date and 

provenance of the wood (provide by dendrochronological tests), but mainly 

in the reason why sycamore timber was preferred (12). The use of sycamore 

for these icons might have derived from the holy attributes of the tree, which 

would have been of great importance to the patron. 

Skalova 87

Figure 5. One of many scenes from Pharaonic tombs: an image of the tree goddess pouring blessings 

011 the deceased. Under the sycamore's shadow are the souls of these deceased ill the form of Ba birds. 

After N. de Garis Davis's Seven Private To mbs at Kurnah. Mond Excavations at Thebes [I. 

1948. Ed. A. H. Gardiner. London, plate XXXIV. 

The symbolic importance of the sycamore tree for Copts originates in Luke 

19:4, where the sycamore tree was said to have been climbed by Zachaeus 

in his eagerness to see Christ. In Coptic folklore, the sycamore apparently 

symbolizes the Coptic people (13). This suggests strong continuity with the 

traditional worship of the sycamore. 

In ancient Egypt, the sycamore was so common that one of the names of 

Egypt was "Land of the Sycamore" (14). It was considered the most holy 

tree, thanks to the deep shadow its protective crown offered in this sunny 

country. Hathor, Nut, and Isis-the three ancient goddesses-were believed 

to dwell in the sycamore and were often depicted nestling in its crown, mostly 

as a personification of the tree itself (Fig. 5) (15). 

The Pharaonic sycamore cult seems to have survived into local Christian 

mythology (16). Even today, people do not like to cut old sycamore. The tree 

grows in village cemeteries to provide protective shadow. Associations with 

the Virgin Mary resting in Egypt under the sycamore also remain alive to 

this day. At the well at Matariyya in Heliopolis, today a suburb of Cairo, a 

centuries-old sycamore still grows that is believed to have been visited by the 

Virgin with the infant Jesus. This holy place is abundantly described by many 

pilgrims as having contained, through the Middle Ages, an enchanting orchard 

of balsam and other exotic trees, such as cypresses (17). 

Christian Ethiopia, which is closely connected with Coptic Egypt, still believes 

that in each sycamore one Maria lives; they call the sycamore Marianet. 

The link between the ancient Egyptian goddess Hathor, Lady of the Sycamore, 

and Christian Virgin Mary can be surmised (18). 

Conclusion 

In Egypt, image veneration and sycamore tree veneration have been practiced 

since antiquity, and great importance was attached to the use of special wood 

for sacred images. This tradition was so strong that its survival into Coptic 

times is not surprising. Thus the Coptic icons are the repository of an earlier 

heritage. 

88 


The understanding of an icon and its appropriate restoration according to its 

role and spiritual meaning in the church requires not only modern conservation 

training and scientific apparatus, but also a certain degree of theological 

sensitivity and awareness. The symbolism with respect to the actual process 

of icon painting used to have a fu ndamental significance. What was common 

knowledge to contemporaries must be reinterpreted after centuries of oblivion 

(19, 20). The icon cannot be understood apart from the wider cultural 

and theological context to which it belongs, as local traditions differ. 

With these typically Egyptian traditions in mind, the author wonders but 

cannot scientifically prove, if the twin beam icons might have been made 

from wood of an ancient sycamore from some sacred Coptic place, such as 

Matariyya. This would perhaps explain why the quality of timber was deemed 

unimportant. 

The construction of panels and the use of sycamore, as well as the distinctive 

Coptic iconography, all strengthen the attribution of these sacred pictures to 

the Nile valley. Future comparative study of the ancient Egyptian beliefs and 

technologies, and their survival through icon painting in the Nile valley, may 

enrich our knowledge about one of the oldest and most conservative pictorial 

traditions in the world. 


The training program, "Conservation of Coptic Icons," in Egypt was made possible 

by the generous support of the Netherlands Directorate General of the Ministry of 

Foreign Affairs and the Egyptian Antiquities Organisation. The author, an art historian 

specializing in the conservation and restoration of icons, has been working 

since 1989 as the Field Director of the Coptic Museum in Cairo. 

Notes 

1. Skalova, Z. 1990. Conservation problems in Egypt: icons, preliminary classification 

and some case studies. In ICOM Committee fo r COl/servation Preprints, 777- 

82. 

2. Skalova, Z. 1991. A little noticed thirteenth-century Byzantine icon in the 

church of St. Barbara in Old Cairo, "The Virgin with Child Enthroned." Bulletin 

de la Societe d'Archeologie Copte, xxx. Cairo, 93-103. 

3. Butler, A. 1884. The Anciel/t Coptic Churches of Egypt, Vol I. Reprint 1970. Oxford, 

104. 

4. These and other medieval icons in Egypt will be studied and included in my 

Ph.D. dissertation, Mediaellal lcol/s ill Egypt, at the University of Leiden. 

5. I anL very indebted to Paolo and Laura Mora, who, when in Egypt as directors 

of the Conservation team of the Nefertari Wall Painting Conservation Project 

1986-1992, kindly shared their experience and advised me in this matter. 

6. Rutschowscaya, M-H. 1986. Catalogue des bois de l'Egypte copte. Paris. See also 

Rutschowscaya, M-H. 1991. Coptic Woodwork. Coptic EI/cyclopedia, Vol. 7, 

2325-47. To my knowledge, the research on Coptic wood and carpentry is 

limited and until recently there has been little scientific investigation on medieval 

icon panels extant in the Nile valley. 

7. Mouriki, D. 1990. Icons from the twelfth to the fifteenth century. In Sinai: Treasures 

of the MOl/astery of St. Catherine. Ed. K. A. Manafis. Athens, 385, note 25. 

8. This conclusion is based on my observations of the beam icons hanging in the 

church of the Transfiguration in St. Catherine Monastery at Sinai and various 

medieval icons in the European collections. 

9. Lucas, A., and J. R. Harris. 1962. Allciel/t Egyptiar/ Materials and Illdustries. 4th ed. 

London. 

10. Sawirus Ibn al-Mukaffa. 1970. History of Patriarchs of the Egyptian Church. Translated 

by A. Kather and 0. H. E. KHS-Burmester. Cairo. Vol. III, Part II. 

11. [ am very grateful to Prof. dr. P Baas, Onderzoekinstituut Rijksherbariuml 

Hortus Botanicus, Rijksuniversiteit Leiden, fo r identifying these species. 

12. Rutschowscaya, M-H. 1990. Le bois dans l'Egypte chrt:tienne. In Artistes, artisal/s 

et productiol/ artistiqlle au Moyel/ Age, [I. Colloque international. 2-6 Mai 1983. 

Ed. Xavier Barral I Altet. Paris: Universite de Rennes, 214-1 5. See the discussion 

following her article about difficulties encountered when analysing wooden objects 

from Egypt at the Louvre by the carbon-14 method and dendrochronology. 

Skalova 89

13. MacCoull, L. S. B. 1993. The Apa Apollos Monastery ofPharaon and its papyrus 

archive. In Museon. Tome 106, note 11. MacCoull cites Les Sycamores, a book 

about the accomplishments of 1930s Coptic cultural figures that was published 

in 1978 in Cairo. 

14. Baum, N. 1988. Arbres et Arbrusles d'Egypte ancienne. Orientalia Lovainiensis Analecta. 

Leuven, 17-75. 

15. Moftah, R. 1959. Die Heiligen Baume im Altem Agypten. Ph.D. diss. Gottingen: 

Georg-August-University. 

16. My thanks to Egyptologists Ramses Moftah and Michael Jones in Cairo fo r 

encouraging this line of interdisciplinary inquiry. OlafE. Kaper provided suitable 

illustration. 

17. Zanetti, U. 1993. Matarieh, La sainte famille et les baumiers. In Analecta Bollandiana 

(1 11):21-68. 

18. Moftah, op. cit., 23. 

19. These are concepts that belong to religious tradition and are transm.itted in a 

way that cannot always be scientifically investigated, which should not dim.inish 

their sign.ificance. I would like to thank the priests of the church of St. Mercurius 

Abu's-Saifain for their kind support in the project, as well as my senior trainees 

Magdi Mansour Badawy, Hanan Nairouz, and Mona Hussein, for their assistance. 

20. Abu-el-Makarim, a Coptic priest from the church Harat Zuweilah in Cairo, left 

a manuscript dated 1206 that describes the monasteries and churches of the Nile 

Delta. He mentions that in the church of St. George at Netel near Port Said, 

one icon of St. George is painted on a piece of the very wood used to torture 

the martyr saint himself. See Father Samuel. 1990. Icons et iconographie en 

Egypte au XIIe siecle d'apres Ie manuscrit d'Abu-el-Makarim, publie en Arabe 

en Caire en 1984. Le Monde Copte, No. 18. Limoges. 

90 


Abstract 

As part of a five-year systematic survey 

of the techniques of English 

medieval wall paintings, the highly 

important late Romanesque and early 

Gothic paintings in the Holy 

Sepulchre Chapel in Winchester are 

being examined. The use of red lake 

in the Romanesque scheme is the 

earliest recorded use of this pigment 

in English wall painting; the identification 

of vivianite, now partially altered 

to a yellow form, is the first in 

any English medieval wall painting. 

All aspects of the technique of both 

schemes, including their complex 

laying out, pentimenti, pigments, 

media, and gilding are discussed in 

the context of contemporary European 

painting. 

Te chniques of the Romanesque and Gothic 

Wall Paintings in the Holy Sepulchre Chapel, 

Winchester Cathedral 

Helen C. Howard 

Leverhulme Research Fellow 

Conservation of Wall Painting Department 

Courtauld Institute of Art, Somerset House 

Strand, London WC2R ORN 


Introduction 

The exquisite Romanesque and Gothic paintings in the Holy Sepulchre 

Chapel constitute what is arguably the finest medieval painted interior in 

England. Dating from circa 1175 and circa 1220, respectively, their iconography 

and style have been discussed exhaustively by Park (1). From the technical 

point of view, they provide fascinating insight into the execution of two 

schemes of exceptionally high quality, separated in date by scarcely forty years, 

but painted in very different techniques. Although the technique of the earlier 

decoration is rooted in the tradition of painting a fresco, an additional proteinaceous 

binding medium and also lead pigments have been identified, 

which, together with the sweeping compositional changes made at an advanced 

stage of the painting process, suggest that significant portions were 

completed a secco. The Gothic painting is perhaps more typical of English 

medieval wall painting; though carbonation of lime is still the principal mechanism 

of binding, the use of large plaster patches, inclusion of additional 

organic binding media, and incorporation of pigments unsuitable for application 

in alkaline conditions preclude describing the technique as fresco. 

Investigation of the painting technique was undertaken as part of a comprehensive 

five-year study, funded by the Leverhulme Trust, of English medieval 

wall painting techniques and in conjunction with a conservation campaign 

in the chapel undertaken by the Courtauld Institute of Art and sponsored by 

the Skaggs Foundation (2) . 

The paintings 

The most striking area of late twelfth-century painting is on the east wall of 

the Chapel, comprising a Deposition in the upper register and, below the 

dividing geometric border, an Entombment with the Maries at the Sepulchre and 

the Harrowing of Hell. Uncovered when the overlying thirteenth-century 

scheme was detached in the 1960s, the painting has escaped the ravages of 

multifarious conservation treatments to which it would almost certainly have 

been subjected had it been exposed earlier. The surface was severely keyed 

in preparation for the thirteenth-century plaster, but despite these numerous 

damages the superb quality of the painting is evident (Fig. 1 and Plate 19). 

Traces of the Romanesque scheme survive elsewhere, as on the south wall 

above the western recess where a Resurrection of the Dead with an angel blowing 

the last trumpet is positioned on either side. Originally situated in the 

corresponding position above the eastern recess, but now transferred to the 

north wall of the chapel, is a striking mitered head and a sinopia for three 

scenes set under architectural canopies. To date, this sinopia is a unique find 

in the context of English medieval wall painting, and provides important 

evidence fo r the technique of the Romanesque scheme. 

Extensive remodeling of the chapel in the early thirteenth century, including 

the insertion of a rib vault, destroyed large portions of the Romanesque 

scheme and necessitated a complete redecoration. The resulting scheme is 

also of exceptionally high quality, but has been marred by a series of invasive 

conservation treatments. The painting originally decorating the east wall (and 

Howard 91

Figure ,' , Holy Sepulchre Chapel, Winchester Cathedrar General view of the twe!Jih-celltllry paillfill.R Oil the east wan Photograph courtesy if the 

COllservation if Wall Paintillg DepartlUCllt, Courtauld Institflle of Art, LOHdon, 

92 Historical Painting Techniques, Materials, and Studio Practice

later transferred to an artificial support at the west end of the hapel) mirrored 

the iconography of the Romanesque painting beneath with a few minor 

adjustments. The Magdalene in the Deposition was replaced by a centurion 

holding a scroll, and the Harrowing of Hell was moved to the recess of the 

adjacent south wall and paired with the Noli me Tangere. Above this recess is 

the Entry into Jerusalem and a newly identified scene of the Washing of Christ's 

Feet (3). Within the western recess of the south wall are scenes of the Martyrdom 

of St. Catherine of Alexandria. A striking image of Christ as Pantocreator, 

surrounded by evangelist symbols, fills the eastern segment of the vault, 

while the remaining segments are painted with Infancy scenes (Annunciation, 

Nativity, and Annunciation to the Shepherds) , foliage, and busts in roundels. 

Other thirteenth-century painting survives on the northern arches of the 

chapel. 

Conservation history 

Until the conservation program in the 1960s, very little of the twelfth-century 

painting was visible, as it lay beneath the later scheme. The thirteenth-century 

painting was itself in extremely poor condition, due to the combination of 

humidity and the wax coating and hair nets applied by Professor Tristram in 

the 1920s to consolidate the surface (4) . The conservation campaign in the 

1960s involved the following: transfer of the thirteenth-century painting on 

the east wall to an artificial support at the west end; detachment and replacement 

in situ of other areas of thirteenth-century painting; transfer of the 

mitered head and sinopia from the twelfth-century scheme onto the north 

wall; partial removal of the wax coating; and consolidation with skimmed 

milk and lime water. This complex conservation history has far-reaching implications 

for the conclusions that can be drawn for the original technique 

of the thirteenth-century paintings. 

Technique of the twelfth-century painting 

Plaster. Information on the construction mortar, depth, and stratigraphy of 

the Romanesque plaster is largely concealed by edging repairs and fills. There 

are, however, two areas which supply valuable evidence for the original technique. 

The ashlar support is visible at dado level where the plaster has been 

lost, and it is clear that a single plaster layer approximately 0.5 cm in thickness 

was applied to the stone. Elsewhere there are two distinct plaster layers: the 

uppermost layer, the intonaco, is approximately 0.5 cm in thickness, while 

below that a slightly coarser and more yellow plaster appears to be about 1.0 

cm thick. It seems reasonable to assume that a layer of plaster was applied to 

level the masonry where necessary, and that the intonaco was then either 

applied over this, or directly on to the stone. 

Unfortunately, it was not possible to take a sufficiently large sample for full 

analysis of the original plaster, but from visual examination of the polished 

cross sections it was estimated that a fmely graded inert aggregate, principally 

subangular quartz, was combined with lime in a ratio of approximately 2: 1 

to create the intonaco. The conspicuous addition of chopped straw and fibers 

to the plaster suggests that they were included to increase the mechanical 

strength of the render. This organic material may have had the additional 

function of acting as a mechanical buffer by absorbing moisture during the 

setting of the plaster (5). In some samples, it is also clear that charcoal particles 

have been incorporated, with the greatest concentration beneath areas that 

are predominantly blue (Plate 20, samples 3 and 5). It seems likely that the 

charcoal was incorporated to reduce the light scatter from the white substrate, 

thereby increasing the covering power of the blue mineral pigments. 

Although the surface texture itself is fairly smooth, the overall topography of 

the painted surface is remarkably uneven . The application of the plaster is 

rather crude, with deep undulations readily visible in raking light (6) . 

There are two principal horizontal zones of plaster corresponding to the 

narrative registers, and another for the ornamental border that divides these 

Howard 93

egisters (Fig. 2). Additionally, within each of the principal plaster zones, there 

are further divisions that roughly conform to the various individual figures 

or groups of figures. These secondary plaster joins, many of which are rather 

indistinct (particularly in the lower register) make an understanding of the 

plastering sequence problematic. It is evident however, that the narrow central 

plaster patch for the geometric border was applied before the plastering of 

either narrative register. In the upper register, the roughly executed plaster 

joins, applied wet-over-dry, clearly indicate that the patch for the central cross 

was applied first, followed by that fo r the main figure group, and finally those 

for the flanking figures. By contrast, the plaster joins in the lower register are 

indistinct, and the application appears to have been wet-on-wet, allowing the 

edges to merge. Careful examination of the surface in raking light does, however, 

indicate that the central portion may have been applied first. 

The plastering of the east wall of the chapel differs significantly from that 

typically found elsewhere in England, where accumulated evidence at Kempley, 

Witley, and Canterbury suggests that application in broad horizontal 

bands is characteristic of the Romanesque period (7). The plaster patches at 

Winchester do, however, find parallels both in England and on the Continent: 

in the scheme (ca. 1100) in the parish church at Hardham (Sussex), where 

separate plaster patches were applied for each scene as well as for the borders; 

and in the Romanesque paintings of Vicq (France), in which the plaster was 

applied in a grid of large rectangular patches corresponding to the narrative 

and ornamental divisions (8, 9). 

Preparatory techniques. The detachment of a portion of the thirteenth-century 

scheme from the south wall in the 1960s provided particularly significant 

information concerning the laying out of the Romanesque painting. A small 

but exquisite mitered head was uncovered here; underneath was found a 

sinopia for three scenes set beneath architectural canopies. 

The existence of a sinopia on the east wall, traces of which are visible where 

tiny losses in the into naco exist, was noted by Park (10). Evidence from the 

stratigraphy of the plaster and from the examination of cross sections, such as 

that of Sample 29 showing a trace of red pigment (red earth with some 

cinnabar) beneath approximately 0.4 cm of plaster, suggests that the sinopia 

was applied either directly on the ashlar support or leveling plaster. 

Examination of the surface of the Romanesque plaster in raking light reveals 

the use of snapped lines marking the position of the central ornamental border. 

The geometric elements within the border were set out by incision into 

the wet plaster with the aid of a compass. When this border was moved to 

a slightly lower position, new incisions were made, but in this case a slightly 

different quality of line is evident since the plaster was clearly no longer as 

fresh. Incision into the wet plaster is evident in other distinct areas, as in 

Christ's arm and the flagon of holy oil in the Entombment. In addition, a 

preparatory drawing in yellow iron oxide, clearly visible wherever the paint 

layers have been lost, was used to place the main pictorial elements within 

the visual field. 

Pentimenti. The Romanesque paintings are particularly fascinating from the 

point of view of the changes made by the painter at an advanced stage of 

the painting process. This is particularly evident in the Entombment, in which 

the remains of another head and a broad-brimmed hat can just be seen to 

the left of the Virgin's head, indicating the original position of the figure 

anointing Christ's body with oil (in the final version placed to the right of 

the Virgin) . 

Many other alterations are visible. For instance, the figure of St. John in the 

Deposition was finally painted with his hand gesturing toward Christ, but the 

yellow preliminary drawing indicates that he was originally conceived with 

his hand held to his face in grief. 

Palette. The present research has established that the palette of the Romanesque 

paintings included: gold leaf, Au; natural ultramarine, 

94 



Holy Sepulchre Chapel 

o 20 60 80 100 em 

East Wall 

, 

§ 

i 

/ 

.f 

J 

.j 

...... . . ......... ....... ....... ............. .. -. ... _, . ....... , .... , .. ... -........... --.... _ . .,_ .... . 

\, 

\ 

t. , " 

Incisions 

D Snapped Lines 

Plaster Joins 

F(rzure 2. Diagral1l of the east lIlall scholle, showillg the positioll of plaster joills, illcisioll, allli S/lapped cord Illarks. Diagrmll collrtesy of the COllserPatioll 

if Wall Paimillg Department, Courtauld Illstitute of Art, Londoll. 

Howard 95

3Na20·3Al203·6Si02·2Na2S; vivianite Fe3 + 2(P04)2·SH20; cinnab:l.r, HgS; 

red lake; red lead, Pb304; hematite, Fe203; green earth, 

K[(Al,Fe"')·(Fe",Mg)]· (AlSi3,Si4)O,o (OH)2; yellow iron oxide, Fe203·H20; 

and lime white, CaC03. 

One of the most interesting findings is that red lake was used for the unusual 

pink color of Nicodemus's robe. This is the earliest identification of this 

pigment in English wall painting, though at Miistair (Switzerland) a red lake 

pigment thought to be madder has been identified in the Carolinian scheme 

of circa SOO (11). 

The inclusion of vivianite in the palette is particularly surprising, not only 

since this pigment has not previously been identified in English medieval wall 

painting, but also because it was clearly selected fo r its distinctive coloristic 

qualities rather than as an economic alternative to other mineral blues (12). 

The characteristic deep indigo blue of vivianite, set against a pale blue of 

natural ultramarine combined with lime white, was employed for the central 

details on the vair (bluish gray and white squirrel fur) lining of Nicodemus's 

cloak in the Deposition. Although the iron phosphate mineral has now altered 

and appears green, an examination of the Morgan leaf from the Winchester 

Bible (ca. 11 70-1 1 SO) shows that vair linings were represented by pale blue 

elements with a dark blue center, and it is clear that this was also the intention 

in the wall paintings. 

Binding media. Analysis undertaken initially by microchemical tests on thin 

sections and cross sections was followed in some cases by Fourier transform 

infrared microspectroscopy. Results confirmed that although the carbonation 

of lime is the principal binding mechanism, a proteinaceous component is 

also present in some samples. A proteinaceous component was identified, for 

example, in a paint layer consisting of red lead, lead white, and calcium carbonate 

in Sample 27, taken from the impasto decoration of the geometric 

border. Likewise, a proteinaceous component was identified in a layer of calcium 

carbonate applied beneath a resinous mordant for gold leaf in Sample 

30, from the halo of the angel in the Entombment. 

Instrumental analysis of the media of comparable wall paintings is rare, but it 

is significant that, where available, it indicates similar findings. For instance, 

in the paintings (ca. 1130) of Idensen (Lower Saxony), the presence of a 

proteinaceous binding medium is associated with blue and green pigments, 

while in the scheme (ca. 1130) in St. Gabriel's Chapel, Canterbury Cathedral, 

both linseed oil and protein have recently been identified as part of the 

original technique (13, 14). 

Application. The presence of a sinopia, and the application of the into naco in 

overlapping patches, indicate that the primary binding mechanism for the 

pigments was the carbonation of calcium hydroxide from the lime plaster; 

that is, at least the preliminary drawing and initial pigment layers were applied 

a fresco. However, the stratigraphy of the final painting is remarkably complex 

and varies substantially across the pictorial surface, from thin single layers 

applied directly on the lime plaster substrate to paint applied in considerable 

impasto, often over colored grounds (Table 1). In Sample 24, a single layer, 

just 30 f.Lm thick, of yellow iron oxide combined with umber and a few 

charcoal black particles has been applied directly to the white plaster substrate. 

Layers such as this are likely to have been applied a fresco. For areas of flesh 

painting, multiple layers have been applied to produce complex effects of 

modeling. In Sample 7, the mid-dark flesh tone has been produced by the 

application of five different layers of earth pigments-green, red, and yellowapplied 

either singly or in combination, often with the addition of lime white 

and charcoal black. Here, carbonation of the lime white pigment provides 

additional binding capacity within the complex layer structure. By contrast, 

red lead combined with an additional proteinaceous binding medium was 

applied in a layer some 200 f.Lm thick for the decorative motifs on the central 

border, over a thin layer of carbon black applied directly to the lime plaster. 

96 


Gilding. Gold leaf was applied for a small number of particular features such 

as halos and the decorative borders of drapery. Two samples were taken to 

establish the gilding technique; in both cases, analysis by FTIR indicated that 

a resinous mordant was used to adhere the gold leaf. The supporting layer 

consists of calcium carbonate combined with a protein, probably glue. At 5 

jJ.m, the gold leaf is exceptionally thick, more than twice that found at Idensen 

(1-2 jJ.m) where it was applied over a "bole" of lead white and carbon 

black bound with prepolymerized linseed oil (15). 

Fading oj the lake pigment. Although the susceptibility of red lakes to fa ding 

was known in the Middle Ages, they were nonetheless often used in wall 

paintings (16). The key factor in the fading of lake pigments is exposure to 

ultraviolet radiation, and, indeed, a darker pink is apparent where fresh losses 

have occurred in linear details overlaying Nicodemus's drapery. Lake pigment 

mixed with white, applied over a white ground and unprotected by a glaze 

(i.e., the Holy Sepulchre Chapel paintings), is particularly vulnerable to internal 

reflection (17). 

Alteration oj vivianite. Initial identification of the pigment by polarized light 

microscopy, in which the pigment is typified by blue-yellow pleochroism 

every 90°, was confirmed by X-ray diffraction (18, 19). In recent years, vivianite 

has been identified in medieval painting in Germany, as on the Romanesque 

lectern at Freudenstadter (ca. 1150) where it was applied over a 

gray ground of lead white combined with carbon black (20). In the context 

of English medieval polychromy, a preliminary identification of the mineral 

has been made on an Anglo-Saxon stone sculpture from York (21). 

Vivianite occurs naturally in two discrete environments. It is fo und in the 

oxidized upper layers of some metalliferous ore deposits, as at St. Agnes in 

Cornwall, where it generally appears as dark indigo, blue-black, or green 

crystals (22). It is also fo und in organic, phosphate-rich environments, and is 

frequently associated with bones, decaying wood, and other organic remains. 

Vivianite is generally stable and dark blue or green in color, though the 

mineral may be colorless when initially exposed (23). 

It seems likely that in the medieval period vivianite was used only where the 

mineral was locally available; thus, mineral deposits are well known in Germany. 

Good crystalline deposits of the mineral have been found at Whale 

Chine on the Isle of Wight, and in its earthy form at Fordingbridge in Hampshire, 

both close to Winchester (24). Current work on samples from the 

chapel includes an analysis of trace elements and examination of the crystalline 

structure to determine whether a mined or peaty alluvial deposit was 

the source of the mineral. 

Vivianite is known to be generally stable in its blue fo rm but at Winchester 

some of the particles have altered to a yellow color, giving an overall green 

effect. The mineral's color change from colorless to blue on initial exposure 

is due to increased ferric ion concentrations, and it has been established that 

mechanical grinding of the colorless crystals, heating in air, storage in a vacuum, 

or chemical treatment of samples can produce a more rapid conversion 

from ferrous to ferric ions, and so to a blue color (25, 26). It is therefore 

interesting to speculate whether grinding of the blue mineral to produce a 

particle size suitable fo r use as a pigment may have contributed to an additional 

increase in ferric ion concentration and ultimately to a further color 

alteration from blue to yellow. 

Technique of the thirteenth-century paintings 

Invasive conservation interventions-including waxing, facing with glue for 

detachment, thinning of the original plaster support, consolidation, and cleaning-have 

compromised the results of the technical examination of these 

paintings. Nevertheless, certain conclusions can be drawn. 

Plaster. Following the keying of the twelfth-century painting, a single layer 

of plaster approximately 5-8 mm thick was applied to the surface. This 

Howard 97

Fi,RlIre 3. This thirteellth-cel1tury pailltil1,R, ori,Ril1aily 011 the east wall, has beel1 transferred to al/. artificial slIpport at the west el1d of the chapel. 

Photo,Rraph collrtesy of the COllServatioll of Wall PaintilJ,R Department, Courtauld [llStitllte of Art, Londoll. 



Holy Sepulchre Chapel 

Jo 

tOO em 

West Wall 

I2J P 

laster Joins 

Figure 4. Diagralll of the thirteenth-century east schenle (now tral1sferred to the west el/d), showil/g the positiol/ of plaster joil/s. Diagralll col/rtesy 

of the COIIServatiol1 of Wall Painting Departlllent, Courtauld Institl/te

plaster consists of lime combined with an aggregate of subangular quartz in 

a ratio of 1:3. It is generally more yellow and less well prepared than the 

plaster for the earlier decoration, with large lumps of unmixed calcium carbonate 

clearly visible. 

It is now difficult to establish the sequence of application of the plaster, since 

surface features have been completely flattened by the detachment process 

(Figs. 3, 4) . Examination under raking light of the surface of the east wall 

paintings (now transferred to the west end) does, however, indicate that plaster 

joins may originally have run along the upper edge of the central border and 

the lower horizontal of the cross in the upper register. This indicates that 

generally the plaster may have been applied in broad horizontal bands, roughly 

corresponding to scaffold lifts. However in the east recess of the south wall 

there appears to be a vertical join between the Harrowing of Hell and Noli me 

Tangere, suggesting that in this area at least a fu rther division was considered 

necessary. 

Preparatory techniques. A preparatory sketch in red iron earth was used to set 

out the main features of the composition. Unlike the twelfth-century painting 

wherein both a sinopia and then fu rther preparatory drawing on the final 

plaster layer exist, in the thirteenth-century scheme the preparatory drawing 

is confined to the final plaster layer. 

Direct incisions into the plaster are also clearly visible, such as the fine, sharp 

incisions that outline the curls in the hair of the figure at far right in the 

Entombment and the hair of St. John in the Deposition. 

Pentimenti. An interesting alteration discovered during the detachment process-and 

presumably visible on the reverse of the thinned plaster layer-was 

that above the horizontal arm of the cross in the Deposition was a sketch fo r 

the sun and moon, which are symbols more usually associated with the Crucifixion 

(27). 

Pigments. The present research has established that the original palette included: 

natural ultramarine (3Na20·3Al203·6Si02·2Na2S); vivianite (Fe3 + 

2[P04L'8H20); copper chloride green; cinnabar (HgS); hematite (Fe203); yellow 

iron oxide (Fe203'H20); lead white (2PbC03·Pb[OHl2); lime white 

(CaC03); charcoal black (C). 

Binding media. The inclusion of lead pigments in the palette makes highly 

likely the incorporation of organic binding media as part of the original 

technique but, owing to the previous conservation interventions, it was not 

possible to confirm their presence. 

Application. Preliminary drawing in red ochre was followed by blocking in of 

the basic background colors. Oakeshott considered that the powerful black 

outlines, which are such a major feature of the scheme, were also painted at 

this stage and were thus more firmly bound by the carbonation of the plaster 

(28). The stratigraphy is remarkably simple, with one or at most two paint 

layers applied directly to the plaster (Tables 2, 3, 4, 5). In some cases the final 

paint layer was applied in thick impasto, and it seems likely that here an 

additional binding medium was used. 

Pigment alteration. Although the background colors of the scheme are now 

red and green, it seemed much more likely on the basis of comparisons with 

other paintings of the period that red and blue would originally have been 

used. Analysis indicated that this was indeed the case: the present green color 

is due to the alteration of some of the original blue pigment (vivianite) to 

form a yellow alteration product. 

The alteration of a lead pigment is evident in the fo rm of black spots on the 

paint surface throughout the scheme, and analysis confirmed the presence of 

a dark red/brown lead-based material. In wall paintings elsewhere where this 

type of alteration occurs, the alteration product has been found to be plattnerite 

(lead dioxide) (29). In the Holy Sepulchre Chapel however, analysis by 

100 


scanning electron microscopy I energy dispersive spectroscopy (SEM/EDX) 

analysis indicated a low oxygen peak, which may indicate that a mixture of 

lead dioxide and lead sulfide is present here. In addition, cinnabar has altered 

to its dark form, metacinnabar, as shown in Sample 37, taken from the darkened 

cheek patch of a prophet on the vault. 


I am grateful to the Leverhulme Trust for their generosity in funding the five-year 

study, and to the Skaggs Foundation for their support of the conservation program 

at Winchester. Particular thanks are also due to G. Cressey and P. Tandy of the Natural 

History Museum, and to David Park, Sharon Cather, and Robyn Pender of the 

Conservation of Wall Painting Department, Courtauld Institute of Art. 

Table 1. 

Samples Jrom the east wall oj the Holy Sepulchre Chapel, Winchester Cathedral 

Sample No: 

3 

Location and Description 

E. wall, upper tier, S. side, olive green 

background to left of Nicodemus' foot. 

Original Polychromy 

• vivianite 20llm 

• natural ultramarine 30llm 

Indications of binding medialPigment 

alterations and other additions 

• vivianite has partially altered from blue to yellow 

________ 

-- 

-- 

-- 

-- 

--. I ;m e p I t er w ; il i c ha r o o al b la ck ;n cl us ;o n ; s =. 40 O m 

 

 

-- 

----------__________ 

- E. wall, upper tier, S. side, red background 

5 

6 

7 

8 

9 

10 

21 

22 

adjacent to cross, red over blue. 

E. wall, upper tier, N. side, uppermost edge of 

Joseph of Aramathea's hat, grey surface over red 

pigment on blue. 

E. wall, upper tier, N. side, Joseph of 

Aramathea's red drapery, translucent dark red 

shadow just below beard. 

E. wall, upper tier, Christ's right fOOl, mid-dark 

flesh tone. 

E. wall, upper tier, Nicodemus' pink drapery, at 

hip level, pale pink with highlight. 

E. wall, lower tier, N. side, Joseph of 

Aramathea's red robe, darkest red tone at uppermost 

edge. 

E. wall, lower tier, N. side, Christ's proper hand, 

flesh painting with pink drop of blood on 

surface. 

E. wall, lower tier, Harrowing a/Hell, deep 

purple shadow in drapery. 

E. wall, lower tier, Harrowing of Hell, green 

outer drapery (query modem retouching). 

• cinnabar in red lead matrix with a f e w particles of natural ultramarine • calcium carbonate crust on surface 

220J,lm 

• natural ultramarine 30llm 

- cinnabar IOllm 

• natural ultramarine and charcoal black in a lime white matrix 10llm 

• lime plaster with charcoal black inclusions 250m 

• haematite 20m 

• yellow iron oxide 2m 

• trace calcium carbonate 

• cinnabar 551lm 

• lime plaster! OOm 

• red, yeUow and green iron oxides and charcoal black 10m 

• green iron oxide 45 m 

- lime white combined with yellow iron oxide IOOllm 

• yellow iron oxide 25).101 

• red iron earth 50).1m 

• trace I ime plaster 

• red lake with a trace of ultramarine in a calcium carbonate matrix 

100).1m. The trace of calcium sulphate present in the layer may 

represent the substrate of the lake pigment, now faded. 

• plaster substrate with charcoal black 80).1m 

• cinnabar 30l1m 

• yellow iron oxide IOl1m 

• plaster substrate with a few charcoal black inclusions 200l1m 

• cinnabar 20l1m 

• trace yellow iron oxide 

• green iron oxide 4511m 

• lime white with yellow iron oxide particles 100l1m 

• lime white, yellow iron oxide and charcoal black 60l1m 

• natural ultramarine, charcoal black and lime white 90l1m 

• lime white 40l1m 

• plaster substrate 200Jlm 

• red lake pigment has faded 

• cobalt green (CoO'nZnO) 

23 

24 

25 

26 

E. wall, lower tier, lower edge of sarcophagus, 

deep purple over yellow. 

E. wall, lower tier, Christ's loin cloth, yellow 

shadow over deep yellow. 

E. wall, upper tier, Magdalene's drapery, hem at 

calf level, yellow & white bole over red/purple 

and black paint layers. 

E. wall, upper tier, Nicodemus' cloak, blue and 

white vair lining, green area over blue. 

27 E. wall, central decorative border, 4th circular 

motif from north side, impasto motif in orange 

28 

29 

30 

31 

32 

and white. 

E. wall, upper tier, Nicodemus' right proper leg , 

brilliant rich red highlight over red lead and 

yellow earth. 

E. wall, upper tier, wavy ground between 

Nicodemus' legs, deep red brown over pink. 

E. wall, lower tier, censing angel, halo. trace 

gilding on bole over brown paint layer and red 

lead. 

E. wall, lowertier, N. side, angel's drapery, 

green over yellow near hem. 

E. wall, upper tier, N. side, grey/green 

background between legs of Joseph of 

Aramathea. 

• charcoal black, redlbrown iron oxides with a few particles of cinnabar 

80/1m 

• yeJlow iron oxide combined with lime white 90l1m 

• plaster substrate l20/1m 

• yellow iron oxide and umber with a few charcoal black particles 30/1m 

• plaster substrate 100)1m 

• resinous layer 50)1m 

• white layer (calcium carbonate + protein) 30)1m 

• resinous layer lOO).lm 

• white layer (calcium carbonate + protein) 20/1m 

• resinous layer 100llm 

- vivianite in calcium carbonate matrix 90l1m 

• natural ultramarine in lime matrix 100J.lm 

• plaster substrate 300).lm 

• red lead combined with lead white and calcium carbonate with a trace 

of proteinaceous binding medium present 200).lm 

• carbon black 10/1m 

• cinnabar combined with red lead 50).lm 

• red lead IOO/1m 

• red iron oxide 15J.lm 

• cinnabar 10).lm 

• plaster substrate/ground 400/1m 

• underdrawing, red iron oxide and cinnabar IOOllm 

• gold leaf5/1m 

• resinous mordant 20J.lm 

• white layer (calcium carbonate + protein) 10/1m 

• dark resinous layer 15/1m 

• white layer (calcium carbonate + protein) 20m 

• dark resinous layer 25).lm 

• green iron oxide combined with some charcoal black particles 50/1m 

• yellow iron oxide 75m 

• yellow iron oxide combined with lime white 180/1m 

• plaster substrate with charcoal black particles 120/1m 

• charcoal black with a few red and yellow iron oxide particles 80).lm 

• plaster substrate 200llm 

• a proteinaceous binding medium is present in the white layers 


• translucent crust of calcium carbonate combined with calcium sulphate on 

surface 

• a proteinaceous binding medium is present in the red paint layer 

• white crust on surface 

• white crust and trace pigment on surface 

• a proteinaceous binding medium is present in the white layers 

• trace white crust combined with a few charcoal black particles 

The table gives the location where each sample was taken, the sample number. and the stratigraphy of the various layers from the top down, 

with the thickness of each layer in microns (pm). 

Howard 101

Table 2. 

Samples from the vest wall if the Holy Sepulchre Chapel, Winchester Cathedral 

Sample No: 

34 

35 

36 


W. wall, upper tier, N. side, green background 

colour below white scroll. 

W. wall. upper tieT, St. John, drapery below 

knee, black over pink. 

W. wall, upper tier, S. side, (left as viewed) leg 

of Joseph of Aramalhea, dark green linear detail 

on back of calf. 


• viviani!e IOOllm 

• charcoal black 80llm 

• lime plaster substrate 60llm 

• darkened lead pigment l5IJm 

• lime plaster substrate 250 11m 

• vivianite 150JJm 

• charcoal black and lime white 40llm 

• I ime plaster substrate IOOJJm 

• trace of red pigment from 12th century scheme 20JJm 

Coatings! Indications of binding media! 

Pigment alterations 

• beeswax coating containing calcium carbonate and calcium sulphate 20llm 

• proteinaceous coaling 30llm 


• beeswax (trace) 

• proteinaceous coating 

• lead pigment has altered to fonn a dark product which appears to be a mixtur 

of lead dioxide (plattnerite) and lead sulphide. 

• beeswax (trace) 



Table 3. 

Samples from the north wall of the Holy Sepulchre Chapel, Winchester Cathedral 

Sample No: 

11 

12 

13 

14 


N. wall, E arch, soffit, E. side, upper figure, left 

proper arm, pale pink drapery. 

N. wall, E arch, chamfer on S. side, E end, 6th 

ashlar block from apex, yellow over green on 

plaster. 

wall, E arch , S face, E side, C13th vine 

scroll in red on plaster. 

N. wall, E arch, soffit, E side, lower figure, 

thick beige coating over green paint of drapery. 


• white, lead-containing pigment with charcoal black, calcium carbonate, 

calcium sulphate & red pigment particles 

• yellow, lead-containing layer which appears white towards perimeter. A 

few yellow iron oxide particles are present 

• lime ground - calcium carbonate with calcium sulphate and a trace of 

wax present 

• yellow earth pigment in lead white matrix 

• vivianite in lead white matrix 

• lime white ground 

• red earth 

• lime ground 

• vivianite in lead white matrix 




• proteinaceous material is indicated in both paint layers 


• beeswax 


• beeswax 


• thick lead white layer 


The tables give the location where each sample was taken, the sample number, and the stratigraphy of the various layers from the top down, 

with the thickness of each layer in microns (Jim). 

Table 4. 

Samples from the south wall of the Holy Sepulchre Chapel, Winchester Cathedral 

Sample No: 

16 

17 

18 

19 

20 

41 


S. wall, E. bay, above recess, blue background 

behind donkey's left ear, thick dark impasto 

over blue background. 

S. wall, W bay, above recess, E. side, old 

cleaning test area, coating from in key mark 

S. wall, W. bay, above recess, E. side, old 

cleaning test area, coating over red! brown paint 

layer. 

S. wall, W. bay, above recess, W. end, CI2th 

angel, area fluoresces in uv, coating over black 

and orange paint. 

S. wall. W. bay, recess, lower zone, central area 

of dark red 'drapery', coating over redlblack 

paint. 

S. wall, E. bay, above recess, Washing offeer, 

bright green from 'border' above Christ's head. 


• darkened lead pigment 100llm 

• natural ultramarine (2Sllm) 

• organic analysis only 

• carbon black pigment 5J1m 

• red earth in lime matrix 10JJm 

• lime ground 40).lm 

• lime plaster 200llm 

• black and red particles in a lead white matiix 351lm 

• lime ground l40llm 

- lime plaster substrate 400JJm 

• red iron oxide 2SIlm 

• charcoal black 40llm 

• blue/green copper chloride green layer 50llm 

• paler blue/green copper chloride layer 50llm 

• calcium carbonate with a little calcium sulphate present 20llm 



• proteinaceous material on surface 

• lead pigment has altered to fonn a dark product which appears to be a mixture 

of lead dioxide (plattnerite) and lead sulphide 

• layers of beeswax 

• beeswax with a small amount of proteinaceous material 

• extremely thin coating present which is almost completely combined with the 

upper pigment layer. As yet unidentified 

• beeswax 

• trace coating on surface 

• copper chloride green may be the product of an alteration prOcess 

The table gives the location where each sample was taken, the sample number, and the stratigraphy of the various layers from the top down, 

with the thickness of each layer in microns (/-un). 

Table 5. 

Samples from the vault if the Holy Sepulchre Chapel, WiNchester Cathedral 

Sample No: 

15 

37 

38 

39 

40 

42 

43 

44 

45 

46 


Vault, E. bay, E. segment, under left proper 

hand of Christ, 'pea' green background of 

mandorla. 

Vault, E. bay, S. segment, W. side, prophet in 

roundel, darkened cheek patch. 


roundel, yellow flesh painting of hand. 


roundel, hand, dark flesh area. 

Vault, E. bay, W. side, S. segment, prophet in 

roundel, left side, dark band around edge of 

roundel. 


roundel, red linear detail of robe over now black 

layer. 

Vault, E. bay, E. segment, mandorla, S. side, 

green border with black on surface. 

Vault, E. bay. E. segment, S. side, green 

background below Christ's hand, white 

decorative spot over green. 

Vault, E. bay, E. segment, mandorla, S. side, 

green over red. 

Vault, E. bay, E. segment, mandorla, N. side, 

brilliant green of border. 


• vivianite 75JJm 

• black (darkened cinnabar) particles in lead white matrix ISllm 


• yellow iron oxide 15JJm 

• lime white ground 50l1m 

• plaster substrate 200llrn 

• charcoal black and red iron oxide 40l1m 

• lime white ground 100llm 

• plaster substrate 

• dark red/brown lead-based layer 651lm 

• lime white ground 751lm 

• red earth 1251lm 

• dark redlbrown lead-based material 551lm 

• white ground 751lm 


• dark red/brown lead-based material 75J.1m 

• lime white ground 180llm 

• lead white 90l1m 

• vivianile 55J.1m 

• white layer (calcium carbonate combined with calcium sulphate) 50llm 

• trace red pigment 

• pale blue/green copper-based layer 351lm 

• white ground 751101 




• beeswax 



• proteinaceous coating with additional components of calcium carbonate, 

calcium sulphate and a trace of wax 151lm 

• cinnabar has altered to its black meta fonn 

• proteinaceous coating 10llm 

• proteinaceous coating with additional components of calcium carbonate and 

calcium sulphate 30JJm 

• proteinaceous material has penetrated paint layers 

• proteinaceous coating 20llm 

• lead pigment has altered to fonn a dark material which appears to be a mixture 


• proteinaceous coating 2SI1m 

• lead pigment has altered to fonn a dark material which appears to be a mixture 

oflead dioxide (plattnerite) and lead sulphide 

• proteinaceous coating IOllm 

• lead pigment has altered to fonn a dark material which appears to be a mixtur 


• proteinaceous coating 51lm 

• beeswax coating 

• proteinaceous material is incorporated in wax coating 351lm 

• proteinaceous coating 51lm 

The table gives the location where each sample was taken, the sample number, and the stratigraphy of the various layers from the top down, 

with the thickness of each layer in microns (Jim). 


Notes 

1. Park, D. 1983. The wall paintings of the Holy Sepulchre Chapel. In Mediellal Art 

and Architecture at Winchester Cathedral. British Archaeological Association Conference 

Transactions 1980, VI, 38-62. 

2. I am indebted to Stephen Rickerby and the students of the Conservation of 

Wall Painting Course, Courtauld Institute, for their observations during the work 

in the chapel, and for providing the drawings used in Figures 4 and 6. Previous 

investigations of aspects of the paintings include: Hluvko, S. 1991. Red piglTlents 

in English medieval wall painting; and Howard, H. 1988. Blue pigments in English 

medieval wall painting. Diploma dissertations. London: Conservation of 

Wall Painting Department, Courtauld Institute of Art, University of London. 

3. This scene, previously described as the Raising of Lazarus, was identified by 

Christoph Tinzl. 

4. Lee, M. 1975. Paintings in the Holy Sepulchre Chapel, Winchester. Unpublished 

M.A. report, University of London. 

5. For the role of organic additives in plaster, see Sickels, L. 1981. Organics v synthetics: 

their use as additives in mortars. In Mortars, Cements and Grouts Used in 

the Conservation of Historic Buildings . Rome: International Centre fo r the Study 

of the Preservation and Restoration of Cultural Property, 25-49. 

6. Deeply undulating surfaces occur in other Romanesque paintings, such as the 

scheme of ca. 1130 in St. Gabriel's Chapel, Canterbury Cathedral. See Cather, 

S., and H. Howard. 1994. Romanesque wall paintings in the apse of St. Gabriel's 

Chapel, Canterbury Cathedral: their technique, condition and environment reassessed. 

Arbeitslufte zur Denkmalpfiege in Niedersachsen (1 1):143. 

7. See Rickerby, S. 1990. Kempley: A technical examination of the Romanesque 

wall paintings. In Early Medieval Wall Painting and Painted Sculpture in England. 

Ed. S. Cather, et al. Oxford: BAR British Series 219, 249-261; Howard, H. 1992. 

All Saints' Church, Witley, Surrey: scientific examination of the Romanesque 

wall paintings. Unpublished report. London: Conservation of Wall Painting Department, 

Courtauld Institute of Art, University of London; Cather and Howard, 

op. cit. (note 6). 

8. These patches were mapped in the 1980s by the Canterbury Cathedral Wallpainting 

Workshop. 

9. Kupfer, M. 1986. Les fresques romanes de Vicq: E tude technique. Bulletin Monumental 

144:98-132. 

10. Park, op. cit., 40. 

11. Mairinger, F., and M. Schreiner. 1986. Deterioration and preservation of Carolingian 

and medieval wall paintings in the Mlistair Convent. Part II: Materials 

and rendering of the Carolingian wall paintings. In Case Studies in the Conserl!ation 

of Stone and Wall Paintings. Ed. N. S. Bronmlelle, et al. London: International 

Institute fo r the Conservation of Historic and Artistic Works, 195-96. 

12. Ultramarine was the mineral blue normally employed in Romanesque wall 

paintings, though azurite has been identified at Kempley (Rickerby, op. cit., 256) 

and Marslet, Denmark (Graebe, H., K. Trampedach, and M. Jensen. 1986. Kalkmalerierne 

i Marslet Kirke. National museete Arbejdsmark 23:164-82). 

13. Matteini, M., and A. Moles. 1994. Mural paintings in Wunstorf-Idensen: Chemical 

investigations on paintings, materials and techniques. Arbeitschifte zur Denkmalpfiege 

in Niedersachsen 11:85-86. 

14. Cather and Howard, op. cit., 144-46. 

15. Matteini and Moles, op. cit., 85. 

16. Cennino Cennini. II Libro Dell'Arte: The Craftsman's Handbook. 1933. Translated 

by D. V. Thompson. New Haven, 26. 

17. The acceleration of the rate of fading over time, due to an increased proportion 

of white, has been shown experimentally fo r a lake mixed with white, and for 

a lake used as a glaze over a white ground. See Johnston-Feller, R., and C. W 

Bailie. 1982. An analysis of the optics of paint glazes: fa ding. In Science and Technology 

in the Service of Conserl/ation. Ed. N. S. Brommelle, et al. London: International 

Institute for Conservation of Historic and Artistic Works, 180-85. 

18. Faye, G. H., P. G. Manning, and E. H. Nickel. 1968. The polarized optical absorption 

spectra of tourmaline, cordierite, chloritoid and vivianite: ferrous-ferric 

electronic interaction as a source of pleochroism. The American Mineralogist (53): 

1174-201. 

19. I am grateful to Dr. G. Cressey of the Natural History Museum for carrying out 

analysis by X-ray diffraction using a Debye-Scherrer camera. The possible ad- 

Howard 103

ditional presence of metavivianite (a polymorph of vivianite) was indicated by 

this analysis. 

20. Richter, E. L. 1988. Seltene pigmente im Mittelalter. Zietschrift Jur Kunsttechnologie 

und Konservierung 2 (1):171-77. 

21. J. Darrah. 1987. Personal communication. Victoria & Albert Museum, London. 

22. Mineral deposits are fo und in England, France, Germany, Italy, Portugal, Serbia, 

and the Ukraine, as well as in the United States and elsewhere. I am grateful to 

Mr. Peter Tandy of the Natural History Museum, London for providing this 

information and for guiding me through the extensive range of samples in the 

collection. I am also indebted to Dr. Brian Young of the British Archaeological 

Survey for his enthusiastic interest and informed opinion on the provenance of 

the mineral. 

23. Watson, T. L. 1918. The color change in vivianite and its effect on the optical 

properties. The American Mineralogist 3 (8): 159-61. 

24. P. Tandy. 1994. Personal corrununication. Natural History Museum, London. 

25. Following initial exposure of the colourless mineral, oxidation of one of the 

paired ferrous ions may occur at the surface and along crystal cleavages. The 

resulting intervalency charge transfer between the paired ions (now Fe2+ and 

Fe3+) produces a blue color. Cressey, G. 1994. Personal communication. Natural 

History Museum, London. 

26. Hanzel, D., W Meisel, 0. Hanzel, and P. Glitlich. 1990. Mossbauer effect study 

of the oxidation of vivianite. Solid State Communications 76 (3):307-10. 

27. This information was kindly provided by Mr. David Perry and published by 

Park, op. cit. 

28. Oakeshott, W. 1981. The paintings of the Holy Sepulchre Chapel. Winchester 

Cathedral Record 50, plate 98a. 

29. Welford, P. 1991. Investigation of the phenomenon of dark flesh areas in English 

medieval wall painting. Diploma diss. London: Conservation of Wall Painting 

Department, Courtauld Institute of Art, University of London. 

104 


Abstract 

It is rare for ancient external paint 

to survive the English climate, particularly 

in the original location of 

the paint. It is vital that evidence of 

color be documented, and studied 

current recognition of this need has 

created opportunities to survey 

the fabric of important historic 

buildings for evidence of paint. 

Analysis of paint fragments fo und on 

the richly sculpted west front of Exeter 

Cathedral has revealed evidence 

of a magnificent polychromy on 

both architecture and sculpture, giving 

vital information on fourteenthand 

fifteenth-century materials and 

techniques. A similar investigation on 

the west front of Salisbury Cathedral 

has just commenced. An initial look 

at surviving evidence of polychromy 

enables some comparisons to be 

made, in addition to other relevant 

examples. 

The Polychromy of Exeter and Salisbury Cathedrals: 

A Preliminary Comparison 

Eddie Sinclair 

10 Park Street 

Crediton 

Devon EX17 3EQ 


Introduction 

There has been in recent years an increasing awareness of the historic importance 

and vital role of color in medieval architecture, as there is at last 

recognition that our ancient buildings, and not just the artifacts within them, 

were painted as an integral part of their overall design. 

The Reformation in England in the sixteenth century resulted in the destruction 

or obliteration of much polychromy. Where any evidence of color still 

exists it is frequently only fragmentary, but those fragments retain much valuable 

information. 

Conservation work on the west-front image screen of Exeter Cathedral carried 

out from 1979 to 1984 revealed much evidence of a rich polychromy 

in the form of paint fragments surviving in the most sheltered corners of 

both architecture and sculpture. Although the study of this polychromy is 

discussed in detail elsewhere, the establishment of a large archive of paint 

samples has created an invaluable resource that can be used with other emerging 

fragmentary evidence (1, 2). As well as providing useful reference material, 

these samples contain much information not yet explored that future similar 

proj ects may yet discover (3) . 

The work on the Exeter polychromy, while not the first English cathedral to 

receive such attention, was on an unprecedented scale (4). As conservation 

work is carried out on more cathedrals and other important buildings in 

England, investigations into the color become a vital element of the work 

undertaken. Similar investigations in other European countries have provided 

a wealth of detailed information, summarized by Rossi-Manaresi and more 

recently by Brodrick (5, 6). Each cathedral adds its own invaluable evidence 

to the complex picture of materials, techniques, and workshop practice of 

medieval times. 

With conservation work due to commence on the west front of Salisbury 

Cathedral in winter 1994, an inspection fo r polychromy was requested. A 

preliminary investigation with access to only part of the facade has shown 

evidence of color. Some analysis has been carried out to date, with further 

work anticipated in 1995. 

Exeter Cathedral west front 

Figure 1. Exeter Cathedral, west front, 

showing image screen. Photograph courtesy of 

the Cathedral of St. Peter in Exeter. 

The west-front image screen of Exeter Cathedral dates from the fourteenthand 

fifteenth-centuries, although much of the crenellated parapet and some 

portions of the architectural elements, along with six sculptures and the heads 

of four others, have been replaced over the years (Fig. 1). Most of the fabric 

is built from Beer stone, a local compact, close-grained limestone. 

Much vital information on the materials and techniques of the medieval 

period are provided in the Exeter Cathedral fabric accounts for the period 

1279-1353 (7, 8). Although the fabric accounts are missing for most of the 

period during which the west front was being constructed and decorated, 

they provide a wealth of information relating to the polychromed bosses of 

the high vault which, combined with an examination of the fabric discussed, 

Sinclair 105

enhances our picture of the working process (9). This, too, can be of immense 

value for comparative purposes with the west front. 

Much of the west front of Salisbury Cathedral (Fig. 2) was restored in the 

nineteenth century, when most of the empty niches were filled with new 

sculpture. Only portions of eight sculptures derive from the original scheme 

of 1245-1260, which may never have been completed. The majority of architectural 

and all of the ornamental and sculptural stonework on Salisbury's 

west front is constructed of the local Chilmark stone, a sandy limestone. With 

much of the stone weathered and covered in lichen, it is not always immediately 

apparent whether the fabric is original or replacement. Unfortunately, 

while there is documentation for much of the restoration work, there appears 

to be none recording the original fabric. 

Figure 2. Salisbury Cathedral, west front. 

Photograph courtesy rif the Courtauld Institute 

rif Art. 

A preliminary inspection for paint on the facade revealed evidence only on 

the moldings above the lower register figures on the north face of the north 

turret, the most protected part of the west front (Fig. 3). More evidence of 

paint may become apparent as cleaning commences on the west front in 1995. 

The inspection for paint extended into the sheltered central porch where, 

again, much of the fabric is from the nineteenth century. On the tympanum, 

however, the medieval surfaces that have remained are in areas still thickly 

painted, with ample evidence of a magnificent polychromy (Fig. 4) . The only 

original carving is that of four heads at the apex of the tympanum; they are 

particularly well protected and retain paint on their hair, beards, and eyes. 

An examination of the roof bosses in the west walk of Salisbury cloister, 

which date from 1263-1270, also reveals evidence of color though here the 

paint is extensive, with some bosses retaining almost all their color. While 

their polychromy cannot be seen as part of the same scheme as that of the 

west front and central porch it serves as valuable additional reference material, 

as well as an important surviving fact in its own right. 

Figure 3. Salisbury Cathedral, north face rif 

north turret, lower register. Evidence rif paint 

was found 011 the moldings above the figures. 

Photograph courtesy of Dean and Chapter, 

Salisbury Cathedral. 

Figure 4. Salisbury Cathedral central porch, 

ly,upauul11. Photograph courtesy rif Dean and 

Chapter, Salisbury Cathedral. 

Paint samples were taken from all three locations at Salisbury to analyze the 

type of pigments and binding media used by the medieval painters. These 

analyses are still in progress and much remains to be done. Further evidence 

may yet come to light that will alter current perception of the preliminary 

results. 

Salisbury polychrorny 

Although the scarcity of evidence of paint on the west-front facade makes it 

rather premature to talk of differences or similarities, a preliminary examination 

of the paint from all three locations immediately reveals a difference 

in technique (Fig. 5). 

Those west-front samples examined thus far show the existence of a thick 

white ground with a single, colored layer on top (Fig. Sa), while those from 

the cloister bosses show a thin, translucent white ground with one or two 

colored layers on top and extensive use of gilding, which may also bear painted 

decoration (Fig. Sb). By comparison those from the central porch display 

a more complex structure (Fig. Sc), with some samples exhibiting up to fifteen 

layers and several repaintings, often resulting in a change of color, with the 

use of a red earth primer, liberal applications of white lead and a broad range 

of colors (Plate 21). Presumably the paint was reapplied whenever it started 

to look shabby; there are three layers of gilding on the beard of one of the 

carved heads. Samples examined thus far from the cloister bosses show no 

evidence of repainting. 

Although more paint may yet be discovered on Salisbury west front, current 

results indicate several possibilities. If the sculptural scheme was left unfinished, 

perhaps only isolated, completed areas were painted. Alternatively, the scheme 

may have been completed, but left largely unpainted due to lack of funds. 

Weathering and human intervention may have caused loss of most of the 

106 


Sa. Location : Salisbury west front, 

north face of north turret. 

Sb. Location : Salisbury, cloister boss. 

2.Smm 

I. White ground 

2. Iron-oxide red 

,, 

I 

lmm 

1. White ground 

2. Yellow ochre, gold size? 

3. Gold leaf 

4. Black decoration 

Sc. Location : Salisbury central porch, 

tym panum . 

Sd. Location : Exeter west front, 

drapery of king. 

2.Smm 

. 

I 

l.Smm 

Multi-layered sample, with 

1. Red earth primer, 

2. White lead undercoat, 

3. Vermilion top layer and several 

repaintings of primer, white lead 

and vermilion. 

1. Pi nk primer 

2. White lead 

3. Vermilion 

4. Black cement layer 

Figure 5. Paint cross sections: a prelimillary view. 

paint, but protected corners usually retain odd fragments, as was fo und on 

the north return. 

At this point in the investigation, the discovery of the existence of paint in 

just one area on Salisbury west front could be the result of a poor technique. 

The careful preparation of the stone with appropriate sealant, primer, and 

ground played a major role in the durability of the paint layers above. The 

initial examination of the north turret samples at Salisbury suggests that the 

white ground is of a chalk and gesso type, and not very tough, though it is 

thick and would have provided a smooth surface for the paint. Its softness 

suggests a loss of medium, though the lack of evidence so far for paint elsewhere 

may be due to a poor choice of medium. 

Visual analysis, largely through the study of cross sections at this stage, reveals 

a typical medieval palette, with a liberal use of costly exotic pigments, particularly 

in the central porch. Pigments here include vermilion, verdigris and 

copper resinate greens, black (probably lamp black), red and white lead, a dark 

blue that appears to be indigo, red and yellow ochre, and gold leaf. The 

cloister pigments include red and yellow iron oxide, black, gold leaf, a cool 

blue-green and a calcium carbonate white. No lead pigments have yet been 

identified here. Only two pigments, red and yellow ochre (with a chalky 

white layer below), have been identified on the west-front facade. 

Exeter polychromy 

That any paint survives at Exeter, in spite of several major cleaning programs, 

must be due to a careful selection of superior quality materials (Fig. Sd) with 

abundant use of durable red and white lead. Most samples are well bound, 

though migrating salts and some loss of medium cause some paint to delaminate. 

A typical sample from Exeter west front has-on top of an invisible sealanta 

red earth primer followed by a pale pink primer consisting of iron-oxide 

red, chalk, white and red lead, all in varying proportions (Plate 22). On top 

Sinclair 107

of this, a lead white undercoat provides a tough resistant layer, often also used 

to enhance the colored layer above, which may in turn be covered with a 

glaze color. 

Techniques: Salisbury and Exeter 

Although the structure of the Exeter west-front samples in no way resembles 

that of the samples from Salisbury west front, there is a greater similarity to 

those samples from the central porch, where a similar range of pigments are 

combined with a complex layer structure. The presence of white lead in the 

porch at Salisbury will have played some part in the survival of the paint 

there. 

The use of three different techniques, apparent even at this early stage of the 

Salisbury investigation, with the indication of two different techniques 

operating on the west front, is in keeping with evidence found on other 

European cathedrals of a considerable variation in technique existing on 

large-scale schemes (10, 11). 

The richest colors and effects at Salisbury may have been reserved for the 

main doorway, the grand ceremonial entry into the cathedral (12). If less 

expensive preparations were used elsewhere, as is suggested by the north turret 

evidence, this could explain their disappearance. At Exeter, the evidence of a 

rich repainting around the main entry also suggests the important role played 

by color in the liturgy of a great church, where the main entry was not 

considered complete without its polychromy and was renewed more frequently 

than elsewhere (13). 

The medium of weathered exterior paint has naturally deteriorated, making 

reliable analysis extremely problematic. Both linseed oil and egg tempera appear 

to be present in the samples from Exeter west front, either as an emulsion 

or to temper different layers. While no media analysis has yet been carried 

out on the Salisbury west-front samples, those from the porch appear to be 

in an egg tempera medium and those from the cloister in linseed oil. At 

Salisbury, however, the usual problems of media analysis are fu rther complicated 

by the presence of lichens, algae, bacteria, and also the identification of 

wax. A protective coating applied in a previous restoration has left a brownyellow 

wrinkled skin over much of the stonework, fu rther complicating anal 

YSlS. 

With most surviving paint at Exeter found in deep crevices, much of the 

detail of the finely carved Beer stone must have been lost, as layers of paint 

fill corners of crockets, nostrils, or finely carved hair. The skill of the medieval 

painter was to use his craft not merely to paint the carved form, but to 

embellish and further enhance it. At times this must have entailed repainting 

obscured detail by redefining highlights and shadows to carved hair. 

Similarly, the moldings around the tympanum and the carved heads at the 

apex in the central porch at Salisbury must have needed picking out in fresh 

color, as the corners lost their clarity over the years, with a build-up of paint 

particularly thick where different planes meet and overlapping of color occurs. 

Where the stone is sound on the west aisle cloister bosses, the carving remains 

crisp, as there is no build-up of paint layers to clog up the detail. 

The painter did not rely on the play of light alone to give form to the 

sculptures, but added depth to hollows with shading or highlights to high 

points where paint rarely survives on an exterior. Traces of black in the inner 

moldings of the orders around the tympanum at Salisbury show how hollows 

were further emphasized by the choice of dark colors. Carvings are thrown 

into relief by the background color from which they emerge. Thus, in the 

porches at Exeter, the foliage around the doorways, now mostly bare stone, 

is defined by the thick traces of indigo still surviving in the corners of the 

background. 

The painter would add details to carving-even if they would never be seen 

by the ordinary viewer-such as the delicate red outline defining the most 

108 


sheltered eye on the southernmost carved head in the porch at Salisbury, 

remarkably similar in style to that fo und on one of the polychromed heads 

on the cloister bosses. 

No metal leaf was found on the facade at Exeter, although the fabric accounts 

list the purchase at Christmas 1341 of " gold, silver, and various colors for 

painting the image of the blessed Peter" high up in the gable of the west 

front (14). Instead there were several occurrences of orpiment where gold 

might have been expected, such as on crowns and hats; presumably this was 

used elsewhere for reasons of economy. 

A reference to "the painting of the bishop in the gable," makes it clear that 

the final costly colors and foils were applied in situ (15). As for primings and 

earlier preparations, these may have been carried out in the shelter of the 

workshop, as was the case with the interior bosses in the crossing, where red 

lead priming can be seen disappearing into the masonry joints under the 

medieval mortar (16). However, recent developments have shown that the 

quire bosses must have been totally painted in situ (17). Both practices, therefore, 

were possible and were operating at Exeter. Workshop practice at Salisbury 

is yet to be explored, though the presence of dirt beneath the lowest 

paint layer on the cloister samples suggests that they were left unpainted for 

some time. 

Conclusion 

While it is too soon to draw conclusions about surviving external color at 

Salisbury, the discovery of its very existence is important, placed in the context 

of a tradition that was soon to disappear from English ecclesiastical buildings, 

although it lingered a little longer in a secular context (18). The search for 

further evidence of polychromy will continue at Salisbury, along with a more 

detailed study of identified paint traces. 

It is hoped that evidence of polychromy from Exeter, and now Salisbury, 

however fragmentary, can be seen as part of a European Gothic tradition. It 

would be interesting to see if a more detailed study reveals any differences in 

style to distinguish English architectural polychromy from that of its European 

counterparts. 


I would like to thank the deans and chapters of Salisbury and Exeter Cathedrals for 

supporting this research, and all individuals nam.ed in the notes. Recent media analysis 

for Salisbury Cathedral was carried out by N. Khandekar of the Hamilton Kerr 

Institute. The research is being carried out with the technical support of the Earth 

Resources Centre, Exeter University. 

Notes 

1. Sinclair, E. 1991. The west front polychromy. In Medieval Art and Architecture at 

Exeter Cathedral. Ed. F. Kelly. British Archaeological Association, 116-33. 

2. Sinclair, E. 1992. Exeter Cathedral: Exterior polychromy. In The Conservator as 

Art Historian. Ed. A. Hulbert, et al. London: United Kingdom Institute for Conservation, 

7-14. 

3. The author, a practicing conservator, carried out this research, with invaluable 

help and tuition from Drs. Ashok Roy and Raymond White of The National 

Gallery, Josephine Darrah of the Victoria & Albert Museum, Peter Mactaggart, 

and Anna Hulbert. However, access to sophisticated analytical equipment and 

techniques was limited, and most results are based on the study of cross sections 

and dispersions. 

4. Conservation work on Wells Cathedral west front, carried out from 1974-1 986, 

uncovered extensive evidence of polychromy that will be reported on in Sampson, 

J. Wells Cathedral: West front archaeology and conservation. Forthcoming. 

5. Rossi-Manaresi, R. 1987. Considerazioni tecniche sulla scultura monumentale 

policromata, romanica e gotica. Bollettino d' Arte (41): 173-86. 

6. Brodrick, A. 1993. Painting techniques of early medieval sculpture. In Romanesque 

Stone Sculpture from Medieval England. 18-27. 

Sinclair 109

7. The Accounts of the Fabric of Exeter Cathedral, 1279- 1353. 1981. Part I, 1279- 

1326. Ed. A. Erskine. Devon and Cornwall Record Society, 24. 

8. The Accounts of the Fabric of Exeter Cathedral, 1279- 1353. 1983. Part II, 1328- 

1353. Ed. A. Erskine. Devon and Cornwall Record Society, 26. 

9. Hulbert, A. 1991. An examination of the polychromy of Exeter Cathedral roof 

bosses, and its documentation. In Medieval Art and Architecture at Exeter Cathedral. 

Ed. F. Kelly. British Archaeological Association, 188-98. 

10. Rossi-Manaresi, op. cit. 

11. Sampson, J. Personal communication. At Wells, three different techniques were 

identified by Dr. Roy (National Gallery, London) with the use of either white 

lead or gypsum (?) as a ground. The niche in which The Coronation if the Virgin 

is situated, above the main west door, retains evidence of the more exotic pigments 

on a lead white ground. 

12. In a search fo r polychromy during recent work on the west front of Bath Abbey, 

Jerry Sampson found that, apart from one isolated fragment elsewhere, the only 

paint was in the west door. This may be the result of a poor technique resulting 

in almost total disappearance of color, but it suggests again the importance of 

the main doorway. See Sampson, J. 1992. Bath Abbey West Front: The History of 

the Restorations. Private report. Bath City Council and English Heritage. 

13. Rossi-Manaresi, R., and A. Tucci. 1984. The polychromy of the portals of the 

gothic cathedral of Bourges. [COM Committee fo r COl1seYilatiol'l preprints, 7th Triennial 

Meeting, Copenhagen. During construction of the west front, the south 

portal was used as the main entry. Here, twelfth-century carvings, unpainted until 

incorporated into the south portal in 1225, were then polychromed in situ, 

forming a temporary main entrance. Bourges Cathedral is contemporary with 

Salisbury Cathedral. 

14. Erskine, 1983. Op. cit., 269. 

15. Ibid., 270. 

16. Hulbert, op. cit., 192. 

17. Hulbert, A. 1994. Personal cOl11l11unication. 

18. Sinclair, E. 1990. Appendix: Sampling and analysis of paint traces from the exterior 

elevations of the front block. In Exeter Guildhall. S. R. Blaylock. Exeter: 

Devon Archaeological Society Proceedings (48): 174-6. Research on the exterior 

of Exeter Guildhall revealed a rich polychromy that included azurite, vernlilion, 

and gilding, with the earliest layers dating from 1593 to 1594. 

110 


Abstract 

Extensive examination of Andrea 

Mantegna's Adoration of the Magi at 

the J. Paul Getty Museum reveals 

Mantegna's unusual technique of 

painting. Although this painting was 

generally described as painted in 

tempera or oil, analysis has revealed 

the medium to be distemper (animal 

glue). The rationale for this technique 

is explained here in terms of 

aesthetic and environmental constraints 

of the fifteenth century as 

well as the art-historical context of 

the period. 

Andrea Mantegna's Adoration of the Magi 

Andrea Rothe 

Department of Paintings Conservation 

The J. Paul Getty Museum. 


Malibu, California 

USA 

Introduction 

In 1985 the J. Paul Getty Museum acquired the Adoration oj the Magi by 

Andrea Mantegna (1431-1 506) (Plate 23) . Nothing was known about the 

painting until the early nineteenth century, when it was thought to have been 

brought to England by Alexander Baring, first Lord Ashburton (1). It was 

first shown publicly in London at the Royal Academy in 1871, and then at 

New Gallery in 1894; but with few other exceptions, the painting was not 

readily accessible. This might explain why it was sometimes confused with a 

nineteenth-century copy, now in the Johnson collection in Philadelphia (2) . 

The Adoration was shown again in London in 1981 in the exhibition Splendours 

of the Gonzaga, and has since been widely accepted as a late work by 

Mantegna, dated circa 1495-1505. The composition of this popular subject 

must have been much admired, for at least seven copies by other artists survive 

(3). 

The horizontal composition with five half-length figures placed tightly 

around the Christ Child bears resemblance to other works by Mantegna, such 

as The Virgin and Child with Saints in the Galleria Sabauda in Turin and The 

Presentation in the Temple in the Staatliche Museen zu Berlin. Mantegna seems 

to have been the initiator of this type of composition, which inspired Giovanni 

Bellini, Cima da Conegliano, and Vincenzo Catena, among others. Remarkable 

are the descriptive features and the distinctive skin color of each 

individual figure. Noteworthy is Mantegna's foreshortening of the Christ 

Child's right leg, which is reminiscent of the Lamentation in the Brera in 

Milan. Most of the execution adheres to Mantegna's original composition, 

with the exception of the faces of the Virgin and the Christ Child, in which 

the X-radiograph shows Mantegna's original reworking or pentimenti 

(Fig. 1). 

Analysis of painting technique used in the Adoration 

In order to properly restore the Adoration, extensive research on the unusual 

technique of this painting was necessary. Although the Getty Adoration was 

generally described as painted in tempera or oil, recent analysis carried out 

by the Getty Conservation Institute has identified the medium as distemper 

(animal glue) (4) . Pioneering research done by John Mills of the National 

Gallery in London and other experts in the 1970s drew attention to the fact 

that this medium was widely used during the Gothic and Renaissance periods 

(5). Within the next decade more accurate methods of analysis were developed 

(6). The principal reason fo r the lack of more information has been the 

difficulty in analytically distinguishing egg from glue, as both are complex 

proteins. The analytical results can also be unreliable because infusions of glues 

from later relinings and consolidants make it difficult to determine whether 

they are part of the original medium. 

Throughout the history of painting, the use of glue as a medium has been 

quite common, yet very often paintings in this medium are classified as tempera 

paintings, such as the mummy portraits of the late Fayum period (third 

century C.E.), which were no longer painted in encaustic, but probably with 

a glue medium (7). In Germany and particularly in the Netherlands of the 

fifteenth and sixteenth century, thousands of distemper paintings were pro- 

Rothe 111

Figure 1. X-radiograph of Andrea Mantegna's The Adoration of the Magi. The]. Paul Getty 

Museum, Malibu. 

FIgure 2. Dierie Bouts, The Annunciation. 

Distemper on linen, 90 X 74.5 em. The]. 

Paul Getty Museum, Malibu (85.PA.24). 

duced, including examples by great artists such as Albrecht Durer and Pieter 

Brueghel (8) . Some of these that have survived relatively untouched show us 

the beauty and brilliance of this medium. Well-preserved examples of this 

technique include three paintings by Dieric Bouts: (a) the Annunciation in the 

collection of the J. Paul Getty Museum, Malibu (Fig. 2); (b) the Resurrection 

at the Norton Simon Museum, Pasadena; and (c) the Deposition in the National 

Gallery, London (9). Many of these paintings, however, have either been 

destroyed or severely damaged and altered. The glue renders the paint film 

brittle and readily damaged by water; because of the glue's hygroscopic nature, 

it attracts dust and candle soot. Consequently, most of these paintings suffered 

the fate of being varnished in order to "liven up" the colors, thus becoming 

oil or varnish paintings by absorption. Most of them were destroyed because 

they were so fragile, and less than 100 of the above-mentioned Netherlandish 

paintings have survived (10). 

In Italy, however, Andrea Mantegna was the only artist to make extensive use 

of this technique. Approximately thirty of his distemper paintings have survived. 

Mantegna, with some of his fellow painters north of the Alps, must have been 

intrigued with the effect of light, or the lack of it, on his paintings. In order 

to visualize the effect of different media in the same light, we need only look 

at an illuminated manuscript in the subdued light of a modern showcase to 

marvel at the glowing colors (11). By contrast, it would be practically impossible 

to view an oil painting under the same light conditions because of 

the deep saturation of the pigments and the different refractive index. Mantegna 

probably faced the same dilemma of wanting to create paintings with 

incredible detail and luminosity that would not lose their visual power due 

to lack of light and the distracting reflections caused by a varnished surface. 

Because we live today in a technically advanced age in which we can regulate 

the light we need, it is difficult to imagine relying on candles and torches. 

Yet, until the wide distribution of electricity less than one hundred years ago, 

light in a closed environment was scarce. Until the Renaissance, many palaces 

and buildings in Italy had only a few small windows; many of Mantegna's 

paintings were probably destined to be hung in a private bedroom or chapel 

in the fortresslike Ducal Palace in Mantua, thus a technique that would not 

112 


create surface gloss and could be viewed in subdued light was ideal. Of all 

the techniques, only fresco and distemper have these characteristics and only 

distemper, which is more brilliant, gives the artist a limitless time to work 

(12). Distemper paintings are also light and portable. A letter by Mantegna 

to Ludovico Gonzaga, Marquis of Mantua, suggested painting portraits on 

fme linen in this technique, as they can be rolled up on a dowel and easily 

shipped (13). These advantages and characteristics might explain why Mantegna, 

the absolute perfectionist, had such a predilection for distemper. 

The animal glue for distemper painting was made from either skins or parchment, 

preferably sheep or goat (14). Glue made out of fish bones was also 

used; it has been identified in the Cult of Cybele by Mantegna at the National 

Gallery in London (15). The glues were prepared and dried into tablets in 

the winter to keep them from growing mold. When the glue was needed, it 

was first soaked in water and then dissolved in a hot water bath. 

Historical methods and recipes 

A rather succinct description of the practice of this distemper technique is 

given in the so-called Eraclius manuscript. In recipe XXVI of De Coloribus 

et Artibus Romanorum, the text states (16): 

If you wish to paint on linen cloth, and lay gold upon it, prepare it thus: 

Take parchment, or clippings if parchment, and put them in a jar with 

water, which must be placed over the f ire and made to boil as before directed; 

then dip a cloth into it, take it out immediately, and stretch it out on a 

wet panel and let it dry. Then burnish or polish it all over with a glass 

muller, and stretch it out, fastening it on to a wooden frame with the 

thread. You may then paint upon it with colours distempered with size, 

or egg, or gum. 

The recipe describes how the cloth was treated to make it less absorbent, so 

that the colors would not be soaked into the cloth and spread out. Only such 

a pretreated cloth could be painted on with some detail. It is this same property 

that is alluded to in Jehan Ie Begue's 1431 recipe compilation Experimenta 

de Coloribus (17): 

In England the painters work with these waters upon closely woven cloths 

[sindone?], wetted with gum water, made with gum arabic and then dried, 

and afterwards stretched out on the floor if the soler ... and the painters 

... paint upon them figures stories and other things. And because these 

cloths lie stretched out on a flat suiface, the coloured waters do not flow or 

spread in painting upon them . . . the touches of the paint brush made 

with these waters do not spread, because the gum with which, as already 

mentioned, the cloth is wetted, prevents their spreading . ... 

Glue gels at room temperature, however, making it difficult to use. The difficulties 

of keeping the medium in a workable consistency are described in 

the De Coloribus Diversis Modis Tractatur in Sequentibus, written in 1398 by 

Johannes Alcherius. In a recipe given to Alcherius by the Flemish painter 

Jacob Coene, instructions are provided for painting and "laying gold on 

parchment, paper, linen cloth, sindone [very fme linen], and on primed wooden 

panels. The recipe describes the making of glue out of parchment or cutting 

of fine leather: "Lastly, let the size, or sized water be warm; I say warm, lest 

it may be conglutinated, ... when it is cold it will be congealed like jelly ..." 

(18). 

Alcherius continues, "Moreover when using a paintbrush, the colour may be 

held in the hand, which by its warmth or heat will not allow it to congeal 

... And in painting with a pen, as well as with a paint brush, it is a good 

thing to keep the colour over a slow fire of charcoal at such a warmth, that 

it may not congeal, but may remain liquid" (19). 

The recipe stresses the fact that-in contrast to painting on panel-when 

painting on cloth or Tiichlein, the artist should apply paint in several layers 

(20): 

Rothe 113

On cloth and sindone it is more necessary that this colour should be layd 

on twice, wile tempered with size, bi f ore it is put on Jo r the last coat 

tempered with white oj egg. And this is because sindone and cloth, owing 

to their porosity, are too absorbent, flowing, flexible, and unstable, and 

theri f ore soak up the colour, so that there does not remain a good and firm 

substance oj colour upon the cloth or sindone, unless, as usi f ul experience 

tells us, it is laid on several times. 

After the first layer was applied, the surface of the canvas was burnished to 

receive the second layer and finally a last layer and gilding (21): 

Let those things dry which you have drawn and painted, and when they 

are dry burnish them, that is, polish or smooth them gently with a tooth 

oj a horse or a boar, or with a polished hard stone fitted Jo r this purpose, 

in order that all the roughness may be softened down, ... and again paint 

over and draw upon those same places, with this colour, as bi f ore, and 

afterwards let it dry, and then polish and burnish it as bi f ore. Afterwards 

go over and repaint those places which you did bifore, with the same 

mordant or colour, but let this third and last coat oj colour be tempered 

with white oj egg .... 

Sometimes painters tried to prepare the medium in such a way that it retained 

a workable consistency without being heated. This is why some recipes, including 

the following two recipes, survive for keeping glue fluid during preparation 

without heating it: "If you have the time, allow the mordant to get 

stale, for several days or weeks, for it will be better putrid than fresh" and "if 

you want to keep it liquid, put in more plain water, and let it stand; and after 

a few days it will stay liquid without heating. It may smell bad, but it will be 

very good" (22, 23). 

Others describe adding more water and vinegar with honey to the dissolved 

glue (24). The next day the vinegar is poured off, and clear water is added 

and heated. When it cools, the glue is stored and eventually becomes liquid. 

The small amount of vinegar soaked up by the glue also seems to help preserve 

it. The De Arte Illuminandi says, "And know that it is a very good plan 

to soften parchment or stag's horn glue with the best vinegar; and when it 

is softened, and the vinegar poured off, add plain water and melt it, and 

proceed as has been said" (25). Pigments were mixed with this medium in 

much the same way as with egg tempera, presumably by first grinding them 

in water and then adding the medium. 

Painted on fine linen, the Adoration measures 54.6 em X 69.2 em. Judging 

by the age and type of strainer, it is reasonable to presume that the painting 

was removed from its original support, which could have been a panel or 

another strainer as described in the Alcherius manuscript. This would presumably 

have been carried out in the early nineteenth century for the purpose 

of mending two horizontal tears. Presumably to repair this damage, a 

lining canvas was applied to the back with an animal glue or starch paste. In 

what is preserved of the original canvas, no tack holes or pronounced stretch 

marks (called "scalloping") are visible. This might indicate that the canvas was 

glued to its support, either a panel or a strainer (26). Two paintings by Mantegna 

with their original supports still exist. One is the Ecce Homo at the 

Musee Jacquemart Andre in Paris, which is glued to a panel, and the other 

is the Presentation at the Staatliche Museen in Berlin, which has its original 

strainer (Fig. 3) (27). A slightly raised edge on the top might indicate that 

the Adoration had an engaged frame, such as the Berlin Presentation, which 

has a series of splinters adjacent to the painted frame with the color continuing 

up along these splinters. An example of an engaged frame survives 

(Metropolitan Museum of Art, Lehman Collection) that has a remnant of the 

original canvas by an unknown artist sandwiched between the strainer and 

the frame (28). 

114 


Figure 3. Andrea Mantegl1a, The Presentation in the Te mple. Origil1al strailler with panel inserts 

seen from front, 68. 9 X 86.3 CII. Staatliche Museen zu Berlin. 

Gold highlighting 

The intricate details and extensive gold highlighting of the Getty Adoration 

are laid out over preparatory base colors such as those visible on the borders 

of the Ecce Homo. In the latter, Mantegna probably finished the painting after 

it was placed into a frame (which no longer exists), as suggested by the fine 

details painted up to a one-half-inch border all around. The feathered-out 

colors on two edges of the Getty Adoration give an indication of such preparatory 

colors (Plates 24, 25). The figures themselves are placed against a 

uniform black background. The halos of St. Joseph, the Virgin, and the Christ 

Child, painted in shell gold (fmely ground gold in gum arabic), are executed 

with fine, concentric brush marks that gradually become longer and thicker 

toward the outer rim (Plate 24). The simulated Cufic letters on the Virgin's 

dress, the coins in Balthazar's cup, the censer and red cap of Melchior, and 

the earrings and brooch of Balthazar are all highlighted with fm e gold lines. 

This all was done in a technique that is consistent with Alcherius' descriptions. 

Concluding observations 

The recent Mantegna exhibit at the Royal Academy in London and the 

Metropolitan Museum in New York provided us with a better understanding 

of Mantegna's use of distemper and gave us an opportunity to study the 

varnished and unvarnished paintings next to each other (29). 

Many distemper paintings that have been varnished have been considered 

irreversibly damaged. The recent restoration of the Adoration was a successful 

attempt to improve the aesthetic quality of the painting and reverse the glossy 

surface to a matte one, a surface that is more consistent with the original 

technique. 

Notes 

1. Important Old Master Pictures (Part 1). 1985. London: Christie, Manson, and Woods, 

Ltd. 18 April 1985:39-44. 

Rothe 115

2. Elam, C. 1982. The Splendours oj the Gonzaga. Catalog. London: Victoria & Albert 

Museum, 122, entry no. 32. 

3. Lightbown, R. W 1986. Mantegna. Oxford, Berkeley, and Los Angeles, 446. 

4. I am indebted to Dusan Stulik, Michele Derrick, Michael Schilling, and Eric 

Doehne from the Getty Conservation Institute fo r their extensive media analysis. 

5. Mills,]., and R. White. 1978. National Gallery Technical Bulletin (2):71-76. 

6. White, R. 1984. The characterization of proteinaceous binders in art objects. 

National Gallery Technical Bulletin (8):5-14. 

7. Thompson, D. L. 1982. Mummy Portraits in the J. Paul Getty Museum. Malibu, 

California: The ]. Paul Getty Museum, 7. 

8. Wolf thai, D. 1989. The Beginnings oj Netherlandish Canvas Painting: 1400-1530. 

Cambridge: Cambridge University Press. 

9. Leonard, M., F. Preusser, A. Rothe, and M. Schilling. 1988. Dieric Bouts's Annunciation. 

The Burlington Magazine Guly): 517-522. 

10. Wolf thaI, op. cit., 6. See also 38-87 in catalogue. 

11. Illuminated manuscript artists used media, such as gum arabic, that are similar to 

watercolors and have very similar optical properties to distemper paint. 

12. Mantegna was known to have been an exasperatingly slow painter, as revealed 

in the correspondence with his benefactor, Ludovico Gonzaga, the Marquis of 

Mantua. 

13. Kristeller, P 1902. Andrea Mantegna. London, 534, doc. 69. 

14. Cennini, C. 1960. The Crciftsman's Handbook. Translated by D. V Thompson. New 

York: Dover Publications, 87. 

15. White, op. cit., 5-14. 

16. Merrifield, M. P 1849. Original Treatises on the Arts oj Painting. London: John 

Murray, 232. 

17. Ibid., 89-90. 

18. Ibid., 255. 

19. Ibid., 262. 

20. Ibid. 

21. Ibid., 262-64. 

22. Ibid., 282. 

23. De Arte Iliuminandi. 1933. Translated from the Latin (Naples MS XII.E.27) by 

D. V Thompson and G. H. Hamilton. New Haven: Yale University Press, 12- 

13. 

24. Eastlake, C. L. 1960. Methods and Materials oj Painting oj the Great Schools and 

Masters. Vol. 1. New York: Dover Publications, 106-7. 

25. De Arte Iliuminandi, op. cit., 13. 

26. Strainers differ from stretchers in that they have no keys. Stretchers became 

widely used after the eighteenth century. 

27. The Ecce Homo has the original panel, confirmed by carbon-14 dating, on which 

the original canvas is glued only by its edges to the back of the panel. The 

Presentation still has its original strainer with two original backing boards inserted 

into the strainer, as if it were on panel. The Lamentation in the Brera, which no 

longer has its original support, quite probably had one similar to the Presentation. 

28. Newberry, T., G. Bisacca, and L. Kanter. 1990. Italian Renaissance Frames. New 

York: The Metropolitan Museum of Art, no. 63 (diagram erroneously under no. 

64). 

29. So far, not a single painting of Mantegna's has been identified as painted in 

oil, even though there are records of his having ordered walnut oil. 

116 


Abstract 

In September 1992 the J Paul Getty 

Museum acquired Titian's VertUS and 

Adonis, which is considered to be 

one of the few of Titian's autograph 

paintings of the subject. The inorganic 

and organic pigments, oil 

medium of the paint layers, and proteinaceous 

medium of the ground 

were examined using various analytical 

methods. In addition, the authors 

studied Titian's other, much later 

version of Venus and Adonis located 

in the National Gallery in Washington, 

D.c. The comparison revealed 

distinct differences in style and technique. 

This multidisciplinary study 

has shown to be very useful in gaining 

a better understanding of Titian's 

working methods. The examination 

also revealed useful information 

about various aspects of early Italian 

oil painting methods in general and 

helps place the technique of this 

painting by Titian in context with 

other paintings of the period. 

Te chnical Examinations of Titian's Venus and Adonis: 

A Note on Early Italian Oil Painting Technique 

Ulrich Birkmaier 


National Gallery of Art 

4th Street and Constitution Ave. NW 

Washington, D.c. 20565 

USA 

Arie Wallert* 


Museum Services Laboratory 

The J Paul Getty Museum 



USA 

Andrea Rothe 

Department of Paintings Conservation 

The J Paul Getty Museum 



USA 

Introduction 

The story of Venus and Adonis, from Book X of Ovid's Metamorphoses, ends 

with the name of a fragile flower: anemone. As the tale is told, Adonis was a 

young man of unequaled beauty, who became the lover of Venus (1). Venus 

was so much in love with him "that she even stayed away from heaven, 

preferring Adonis to the sky. She used to hold him in her arms, and became 

his constant companion." She warned her lover not to hunt dangerous animals: 

"Your youth and beauty, and the charms which make Venus love you, 

have no effect upon lions or bristling boars, or the eyes and minds of other 

wild beasts. The fierce boar deals a blow with his fangs, as swift as a lightning 

flash ..." 

Of course, Adonis did go out hunting. His dogs fo und a fresh trail, followed 

it, and roused a wild boar. Adonis tried to kill it, but with the help of its 

crooked snout the boar dislodged the spear. The boar pursued Adonis and 

"it sank its teeth deep in his groin, bringing him down, mortally wounded 

on the yellow sand." Venus, on her way to Cyprus, driving through the air 

in her chariot, heard her lover's groans. She went down, and with the dying 

Adonis in her arms, said that an everlasting token of her grief would remain 

there. His blood dripping to the ground would change into a flower, the 

anemone. But, just like their love, the enjoyment of this flower is brief "for 

it is so fragile, its petals so lightly attached, that it quickly falls, shaken from 

its stem by those same winds that give it its name, anemone." 

This story is represented in a series of poesie paintings of Classical subjects, by 

Titian. Such paintings were very popular among wealthy patrons, as their 

mythological subject matter usually provided an excuse for the depiction of 

overt sensuality. King Philip II of Spain commissioned a series of poesies from 

Titian. The Venus and Adonis now located in the Prado Museum in Madrid 

is considered the first of the series to reach its patron in 1554. The last one 

was The Rape oj Europa of 1562, now in the Isabella Steward Gardener Museum 

in Boston. 

Another version of Venus and Adonis by Titian recently came into the collection 

of the J. Paul Getty Museum. This work, showing Venus attempting 


Birkmaier, Wallert, and Rothe 117

to prevent Adonis from leaving for the hunt, is considered one of the few of 

Titian's autograph paintings on the subject (Plate 26). In addition to this work 

and the Venus and Adonis in the Prado, another of Titian's paintings by the 

same name is located at the National Gallery of Art in Washington, D.c. 

(Plate 27). 

Conservation work on the Getty Museum's recent acquisition of Venus and 

Adonis gave the authors the opportunity to perform a technical examination 

of the painting. It was examined with the use of the following: microscopic 

methods, including polarized light microscopy (PLM) and scanning electron 

microscopy (SEM); spectrometric methods, including ultraviolet visible spectroscopy 

(UV Ivis), Fourier transform infrared (FTIR) spectroscopy, fluorescence 

(FS) spectroscopy, and energy dispersive X-ray fluorescence (EDXRF) 

spectroscopy; and chromatographic methods, including gas chromatography 

(GC) and thin-layer chromatography (TLC). X-ray diffraction (XRD) and a 

number of staining tests and microchemical tests were also used (2). The 

cooperation between conservators and scientists proved instrumental in gaining 

a better understanding of Titian's painting technique. This understanding 

was greatly improved through the cooperation of David Bull of the Washington 

National Gallery who gave us the opportunity to examine a later 

version of the Venus and Adonis in the gallery's collection using a stereomicroscope. 

This examination made it possible to compare Titian's execution 

of the same theme at different stages in his career. Our study was greatly 

facilitated by Joyce Plesters's pioneering publications on the examination of 

Titian pieces in the London National Gallery. Her work served as an example 

for our own research and as a continuing source for comparison, and helped 

in the interpretation of the analytical results. It also helped the authors place 

the technique of this particular painting in relation to that of other paintings 

of the same period, as well as those of Titian's later and earlier periods. 

Support 

The painting measures 160 X 196.5 cm. The original canvas is a plain weave 

linen, having 16 threads per centimeter in the warp direction and 18 threads 

per centimeter in the weft direction. It is made of two strips of canvas, joined 

by a vertical seam. The right strip measures 101 cm; the left-hand strip is 

slightly smaller and measures 95.5 cm. The painting has been cut on all sides; 

it is possible that the left side was cut by 5-9 cm. PI esters has found that the 

loom width of sixteenth-century Venetian canvases tends to range between 

1.06 and 1.10 cm (3). 

The painting was first documented in the collection of the Queen Christina 

of Sweden and then in that of the Duc d'Orleans. The French eighteenthcentury 

engraver Delignon made prints of objects in the duke's collection 

(4). One of these prints shows the Getty Venus and Adonis, in which it can 

be seen that the original composition may have extended out a bit more on 

the left side and that it was not originally cut off as close to Venus's fo ot as 

it is now. Since the scalloping of the original canvas on the left side is clearly 

still visible, the losses cannot have been too extensive. 

In the Prado version, which measures 186 X 207 cm (as compared to the 

National Gallery version, which measures 107 X 136 cm), the painting's 

surface extends a bit further to the left. Comparison of the loss in the Getty 

painting with corresponding areas in the print and in the Prado version leads 

to the conclusion that the width of the loss may be accounted for by the 

difference between the present 101 cm strip and the 106-1 1 0 cm width of 

the average sixteenth-century Venetian loom. At the right of the Getty painting, 

a relatively large section of the dog's tail, still complete in the Delignon 

print, is cut off. The amount of the loss is difficult to estimate, but judging 

from comparison with the print (the Prado version has suffered an even 

greater loss), it is probably not more than 4-5 cm. 

The loss at the upper side, again based on comparison with the Delignon 

print, and the London and Prado versions, is more extensive. The shaft of the 

118 


arrow in Adonis's right hand ends at the edge of the painting. The loss may 

be estimated at 5-9 cm. On the basis of a comparison with the Delignon 

print, the original size of the painting may have been 172 X 213 cm. 

Ground 

The painting has a conventional gesso ground. The ground was chemically 

identified as calcium sulfate. XRD has shown that it consisted of a dihydrate 

gypsum. This composition is common fo r Venetian painting at that time. 

Gypsum of the dihydrate form occurs when it is used in its natural state. The 

use of unburned gesso-as opposed to burned gesso, where crystallization 

water is driven off to an anhydrite form-seems to occur primarily on Venetian 

paintings. PLM examination, however, has shown that a considerable 

amount of anhydrite particles were also present. Paintings from Florence and 

Sienna more often show the presence of anhydrite or hemihydrite grounds 

(5). Gesso grounds of this type were the same as those used for ground layers 

on panel paintings, and its use stems from this tradition. In panel paintings, 

however, several layers of gesso-gesso grosso and gesso sottile--were applied. 

The brittleness of the gypsum layers, which form a good first ground on 

panels, would cause it to crack off too easily from a canvas. Therefore, artists 

began using thinner gesso layers on canvas; eventually, in the eighteenth century, 

an oil priming was used instead. 

The ground on the Getty Titian is rather thinly applied so that it only fills 

the spaces between the warps and wefts of the canvas. Cross sections show 

that it is an unpigmented ground. In some areas, an oily layer containing a 

few charcoal-black particles was found between the white ground and the 

first paint layer. It was only later, in the paintings by Tintoretto, for instance, 

that a colored ground became more common. A passage in the Volpato 

manuscript refers to Titian's use of white gesso grounds as opposed to the 

increasing use of colored grounds by other, more modern Venetian painters. 

Staining of microscopic cross sections with specific reagents gave strong indication 

that the binding medium of the ground contains proteinaceous material; 

this was confirmed by the presence of specific FTIR absorption bands. 

Infrared mapping allowed us to locate the presence of specific bands in the 

cross sections. 

Underdrawing 

Infrared reflectography revealed some evidence of underdrawing. The interpretation 

of the infrared reflectograms is difficult, however, as some features 

that showed up as broad, dark lines in the reflectogram could be carbon black 

used in the paint, rather than the underdrawing. It is not immediately obvious, 

for instance, whether the dog's curled tail in the original underdrawing was 

intended to be straight. The broad, dark form in the reflectogram may represent 

a dark, carbon-black pigment in the painted tail, which was actually 

intended to be curled in the underdrawing. The reflectogram shows sketchy 

lines in the trees in the background. The sleeping Cupid seems also to have 

been rapidly sketched before painting. In all, no significant deviations from 

the preparatory drawing appear to have been made. 

Paint layers: the medium 

Staining of cross sections indicated that the actual painting was executed in 

an oil medium. This was confirmed by the presence of characteristic absorption 

bands in the FTIR spectrum (Fig. 1). Three samples were selected for 

examination with gas chromatography. One sample consisted Inainly of paint 

fo r the golden vase in the painting's lower left corner. The other samples 

contained blue particles for the sky. Contemporary sources often indicate the 

use of walnut oil, which was generally considered to yellow less with age, for 

the making of blue paints. Earlier studies of Titian's paintings have revealed 

that the artist used both types of oil on different occasions. Chromatography 

showed an azelaic:palmitic acid ratio in all three different samples, indicating 


60 

55 

"! ,... '" 

.. ..; ,.: 

50 

N 0 

 

'" 

:;!; 

, , , , , 

IT '000 3600 3200 2800 

'00 

Wavenumber (cm-1) 

2000 raoo 1600 "00 1200 1000 800 600 

Figure 1. FTIR spectrum of paint sample (no. 17) takenfrol11 blue if the sky area. Hydrocarbon 

stretch bands at 2923 CI11-1 and 2851 cm-I• as well as the hydrocarbon bend at 1404 cm-I, indicate 

the presence of an oi/ medium. The band at 2342 cm -I is characteristic for natural ultramarine. 

the use of a drying oil. The palmitic:stearic acids ratios of the samples suggest 

that the sky was executed in walnut oil or in a mixture of walnut and linseed 

oils in contrast to other areas of the painting, in which the faster drying, but 

more yellowing, linseed oil was used. 

Painting 

The thinness and relatively simple structure of the paint layers are in accordance 

with what seems to have been Titian's practice around the 1550s. 

Examination of the cross sections shows that most of the paint was fairly 

directly and thinly applied, as opposed to the technique he used in his later 

paintings. In the foreword to Marco Boschini's Ricche Minere della Pittura Veneziana 

(1664), an authoritative and contemporary description is given by 

Palma Giovane of the manner in which Titian gave form to the paintings he 

made after the 1550s (6): 

He used to sketch in his pictures with a great mass if colours, which served 

as a base fo r the compositions he then had to construct. [The compositions 

wereJformed with bold strokes made with brushes laden with colours, sometimes 

with a pure red earth, which he used for a middle tone, and at other 

times of white lead; and with the same brush tinted with red, black and 

yellow he formed an accent; and thus he made the promise if a figure 

appear in fo ur strokes . ... Having constructed these precious fo undations 

he used to turn his pictures to the wall and leave them without looking at 

them, sometimes for several months. When he wanted to apply his brush 

again . .. he would treat his picture like a good surgeon would his patient, 

reducing if necessary some swelling or excess of flesh, straightening an arm 

if the bone structure was not exactly right . ... After he had done this, 

while the picture was drying, he would turn his hand to another and work 

on it the same way. Thus he gradually covered those quintessential forms 

with living flesh, bringing them by many stages to a state in which they 

lacked only the breath of life . ... In the last stages he painted more with 

his fingers than with his brushes. 

This process could involve not just two or three weeks but, with several steps 

and with long interruptions, could continue for months, even years, resulting 

in a painting with several paint layers. 

The process, as described by Palma Giovane, is distinctly different from the 

earlier manner in which Titian painted. This difference was already noticed 

by Vasari in 1566: 

120 


For the early works are executed with a certain f inesse and an incredible 

diligence, so that they can be seen from close as well as from a distance 

while these last pictures are executed with broad and bold strokes and 

smudges, so that from nearby nothing can be seen whereas from a distance 

they seem peifect. 

In the case of the Venus and Adonis at the Getty, the earlier technique was 

used. The paint layers are few and relatively thinly applied. Where visible in 

infrared reflectography, the underdrawing seems to have been closely followed. 

No significant changes from the carefully planned composition have been 

made. Some later additions seem to have been made, however, in areas such 

as Adonis's arms and Venus's legs and back. Some areas in the landscape also 

appear to be built up with several paint layers. The pigments seem to be 

thoroughly ground directly into the binding medium. This contrasts strongly 

with the pigments in the National Gallery's Venus and Adonis. Observed 

under a stereo microscope, many lumps of pigment are visible, suggesting that 

in making the paint, the pigments were not really ground into, but simply 

stirred into the oil. Some differences can be noted between the version in 

the Getty and the one in the Prado. Judging from the development as described 

by the two contemporary comments, the Getty painting is more 

loosely executed than the Prado version, and must have been made at a later 

date, perhaps around 1560. The National Gallery version can definitely be 

considered to have been painted after the Getty and Prado versions. 

Paint layers: the pigments 

Yellow pigments. The highlights of Adonis's belt were executed in a bright 

yellow with occasional small orange-red dots placed on the highlights. When 

examined with PLM, samples taken from the area showed that the yellow 

consisted of highly birefringent mica-like particles (n>1 .66). The yellow pigment 

had all the optical characteristics of orpiment, including a laminated 

form and a waxy luster of the fairly large crystalline particles. The identification 

of orpiment was confirmed by finding larger amounts of arsenic in 

the sample. A small microchemical test showed that the sample also contained 

sulfides. 

The red dots on Adonis's belt and the orange-red of Adonis's shoe were also 

executed in an arsenic-containing pigment realgar. These pigments were 

identified by the usual means of microscopic and microchemical analysis. In 

addition, the findings were confirmed by XRD and XRF (Fig. 2). 

The use of orpiment and realgar in European easel paintings is comparatively 

rare, but it is not unusual to find them in Venetian painting; the use of 

orpiment occurred more often in manuscript illuminations. This lack of popularity 

may be due to the fact that both pigments are highly poisonous, and 

good alternatives, such as lead-tin yellow or yellow lakes fo r orpiment or 

vermilion for realgar, were readily available. Because of their sulfidic nature, 

both orpiment and realgar are not very compatible with many other pigments. 

The Theophilus manuscript even states that orpiment cannot be mixed with 

any other color because it would destroy them. While realgar was sometimes 

used to prevent putrefaction of binding media by bacteria, fungi, and microorganisms, 

historical recipes describing the use of realgar as an actual pigment 

are extremely rare (7). 

Considering their drawbacks, it is surprising to see how frequently these 

pigments were used in sixteenth-century Venetian painting. They have been 

identified in several paintings by Titian and Giorgione and in many paintings 

by Tintoretto and Bassano. The reasons for this preference are not immediately 

evident. Both mineral substances could not be fo und in the immediate 

vicinity and had to be imported. They are conspicuously absent in paintings 

made in the area around Naples where the minerals actually do occur in the 

fumaroles near Mount Vesuvius. Their frequent use in sixteenth-century Venice 

may relate to the increasing use of oil as the paint medium. The pigment 


!:; 

;Dr -ID'LdDtDw 

at' ______________________________-' __ r-__ ,, __ '-'-, 

Titian YlIOIl low .pot on 

baIt Adon i. 

80/51" tcrg8t. 

240 50kV. 3. 3m". 

acac. 11. I. 1993, AW 

A. 

t- 

.. 

7L-____________________________________________________________ 

KloV 10 11 12 13 

Figure 2. XRF spectrum of yellow area on belt of Adonis. High peaks fo r arsenic and sulphur suggest 

the presence qf orpil11ent. 

particles, being enclosed in the oily medium, would no longer be incompatible 

with other pigments, as they were in the previously used tempera techniques. 

The radiant brilliance of their color in this new medium may have 

contributed to their popularity. 

Ochres were used to paint the landscape, the brownish color of the dogs, and 

the golden vase near Venus's seat. Yellow ochres (hydrous iron oxides) were 

identified under the microscope by their optical properties and by the presence 

of high peaks fo r iron in the XRF spectrum. 

Blue pigments. The major blue pigment used on this painting was natural 

ultramarine (Fig. 1). Samples taken from the deep blue of the mountain range 

appear under the microscope as pale blue, splintery particles with a low refractive 

index (n < 1.66). In the cross section, one can see a single layer of 

densely packed blue particles. 

XRF of several blue areas showed high peaks for cobalt, potassium, and arsenic, 

strongly suggesting the presence of smalt, an artificial pigment made 

from potassium-rich glass deeply pigmented with cobalt oxide and ground 

to a powder. Gettens and Stout suggest that the earliest occurrence of cobaltcolored 

glass in Europe may have been in the early fifteenth-century Venetian 

glass industry (8) . As the colorant for the glass, a substance called zafran is 

mentioned. A recipe in a fifteenth-century treatise already mentions the preparation 

of smalt as a smalto cilestro (9). 

The source of the smalt in the Getty painting may have been Saxony or 

Bohemia, where sixteenth-century glassmakers used the locally mined cobaltite 

(CoAsS) and smaltite (CoAs2) minerals, which contain large amounts 

of arsenic, to make smalt. The high peaks of arsenic measured by XRF, in 

combination with those of cobalt, seem to indicate that cobaltite or smaltite 

minerals were the source for the blue pigment. It is unclear whether the 

zafran, or zaffer of Italian descriptions, has the same composition as the northern 

European cobaltite. In other areas of the painting's sky, the original ultramarine 

was scumbled over with a pale, milky blue. Examination of cross 

sections of those areas shows that the pale blue consists of a layer of almost 

completely discolored glassy particles. 

Red pigments. Vermilion appears in Adonis's red sleeve. Its presence was established 

by discovering high mercury peaks with XRF and confirmed by 

122 


J 

Figure 3. Fluorescence spectrul1'l of red lake pigment on shirt of Adonis. Quantum yields at A em 445 

nl1'l to A ex 397 nm and A em 433 nl1'l to A ex 355 rim indicate preSe/1ce of a cochineal dyestuff. 

microchemical tests and PLM of a sample taken from the area. The other 

reds in Adonis's tunic and the clothing draped over Venus's seat were of an 

organic nature. In a sample taken from the seat, an organic colorant was fo und 

that also showed the presence of a few textile fibers. This led us to believe 

that the red lake may very well have been a so-called lacca de cimatura de grana. 

These cimatura lakes were made by an early recycling process. The dyestuffs 

were precipitated into a pigment lake extracted from red textile clippings or 

shearings (cimature) from a tailor's workshop. The making of such lakes was a 

fairly common practice, and its description can be fo und in numerous fifteenth- 

and fourteenth-century recipe texts (10). The procedure could be 

performed with most of the available red textiles. The dyestuffs most likely 

to be found in these lakes are anthraquinone-type dye, such as kermes, cochineal, 

lac dye, and madder. The analyses by FS and TLC showed the presence 

of more than one organic colorant; the red appeared to be a mixture. A sample 

taken from Adonis's garment showed a good match with a library scan of the 

cochineal standard (Fig. 3). The results were confirmed by chromatography, 

which also indicated the presence of purpurin, a colorant present in madder 

type dyestuffs. No spot for alizarin, the major coloring component of common 

madder (Rubia tinctorum), could be found. Several possible conclusions 

are indicated: (a) wild madder (Rubia peregrina) was used, (b) a specific technique 

was used for the dyeing of the original textile, or (c) the alizarin was 

not fully extracted in the cimatura process. This finding supports the suggestion 

that the lake was a cimatura de grana-lake, and the painters who extracted 

the dyestuffs from textile shearings had no control over the actual dyestuff 

composition. Since they are similar in color, and not readily distinguishable, 

any madder-dyed textiles could easily have been included in a larger batch 

cochineal-dyed clippings. 

Lake pigments are known to dry poorly. It was common practice, therefore, 

to add ground glass to lakes. Sixteenth-century glass was typically very rich 

in lead; the lead in the glass would act as a siccative, thus promoting the 

drying of the paint. An interesting passage in the Paduan manuscript describes 

the process: "To make lake indigo and lamp-black, dry quickly. Grind them 

with oil, then take glass ground to a very fine powder, and incorporate with 

the colors by grinding them together again; and thus, in the space of 24 

hours, they will dry" (11). 

Examination of a microsample of the red lake of Adonis's shirt did, indeed, 

show the presence of small glass particles. The purplish red of Venus's seat, 

on the other hand, was made of a mixture of an organic red lake and smalt 

(Fig. 4). In this case, the smalt had a double function. It gave the carmine lake 

a more purple color and at the same time acted as a siccative. 

White pigments. The white pigment found on the painting was lead white. 

XRF revealed the presence of lead in the sky and in the flesh tints. Polarized 


Titian r&d 91 aZI!;! on aQot 

12.1.199:3. 80/5,.. target. 240 SOkV, ::I. grru' 

eee. AW 

F. 

Pb 

A. 

Pb 

Figure 4. XRF spectrum if purple area on seat of Venus. High peaks fo r cobalt and arsenic suggest 

the presence if smalt. 

light microscopy of pigment samples taken from those areas showed them to 

contain birefringent particles with a high refractive index (n> 1.66) that 

matched lead-white laboratory standard. This conclusion was confirmed by 

staining cross sections with Lugol stain. X-radiographs of the painting show 

how Titian used the lead white, particularly to enhance the contrasts in his 

composition. It helped to make Adonis's silhouette stand out against lighter 

areas in the sky, and emphasized the reddish blush on the face of Venus against 

the bright crimson shirt of Adonis (Fig. 5). 

Figure 5. X-radiograph if Venus and 

Adonis, showing Titian's use of lead-whitebased 

highlights. 

A sample taken from the sky area showed that there was not very much 

pigment present. The sky appeared to be executed in thin glazes over an 

anhydrite gesso ground, with the white of the ground showing through. The 

finding of the lathlike anhydrite particles came as a surprise since anhydrite 

is generally considered to have been used in more southern areas of Italy. In 

the Venice region, the dihydrate fo rm was usual. We do not yet have a satisfactory 

explanation for this. In samples taken from the sky area, a relatively 

large amount of splintery, isotropic particles with low refractive index 

(n

This property was deliberately used in the making of copper resinates. These 

are complex compositions of copper salts with various resinous acids, such as 

abietic and succinic acids, achieved by heating verdigris with Venetian turpentine 

or pine resins. The resulting substance shows no particulate matter, 

and a strong, bright green color. Copper resinate became the pigment of 

choice whenever bright green glazes were desired. An early account fo r the 

deliberate making of such a compound can be found in a fifteenth-century 

manuscript in the Biblioteca Casanatense (13). Unfortunately, the copper resinates 

show the same tendency to discolor, giving the meadow and the trees 

in the background of the Getty Titian their present brown color in place of 

the originally deep green color. In various areas on the painting, however, 

the copper resinate has retained some of its original color. In particular, the 

green leaves of grass near Adonis's foot stand out because of their bright, 

strong color. Examination of a cross section taken from that area showed that 

the copper resinate was not laid over the underpaint but rather mixed with 

lead white. This mixture may possibly have helped in preserving much of its 

original color. The green of the trees in the background was produced with 

copper resinate. 

The green that was used to paint the meadow near the vase at the lower-left 

corner consisted of a mixture of copper resinate and yellow ochre. The cross 

section showed that there were only two different paint layers on the gesso 

ground: one layer of yellow ochre, possibly applied in two coatings, and an 

upper layer of copper resinate. The leaves of the plants appear to be highlighted 

with white, and glazed with copper resinate. 

No other green pigments could be identified. This finding tallies with the 

use of greens in Titian's later Tarquin and Lucrezia (14). We were surprised 

that malachite, fo und in Titian's almost contemporary Bacchus and Ariadne or 

in Tintoretto's paintings in the National Gallery, could not be detected in the 

samples we took from the Getty painting (15, 16). 

Conclusion 

Our examinations show that the execution of the J. Paul Getty Museum's 

Venus and Adonis represents a stage between the Prado version and the National 

Gallery version. It stands technically in the middle between two extremes 

in Titian's stylistic development. The earliest stage is the Prado version 

in which an already conceived image is carefully designed and then executed 

and filled in accordingly. In Titian's later style, however, the painting grows 

out of an interaction between matter and concept. As every touch of the 

brush has its impact on all previous touches, there is a shift in the appearance 

of the final painting. Rather than resulting from a fixed plan, this way of 

creating a painting with the total problem of the picture in mind, is apt to 

be a continually developing and self-revising one. 

The Getty piece already represents a concept of painting in which form does 

not merely follow function, but rather grows out of a continuous interaction 

between the demands of the material and the artistic idea: "Obwohl das Werk 

erst im Vollzug des Schaffens wirklich wird und so in seiner Wirklichkeit 

von diesem abhangt, wird das Wesen des Schaffens vom Wesen des Werkes 

bestimmt" (17). 

Notes 

1. Ovid. 1986. Metamorphoses. Translated and introduced by M. M. 1. Middlesex, 

Penguin Books, Ltd., 239, 244-45. 

2. van Asperen de Boer. J. R. J. 1975. An introduction to the scientific examination 

of paintings. Netherlands Yearbook Jo r History if Art (26): 1-40. 

3. Plesters, J. 1980. Tintoretto's Paintings in the National Gallery, National Callery 

Technical Bulletin (4):37. 

4. Wethey, H. E. 1975. The Paintings oj Titian, Vol. III: The Mythological and Historical 

Paintings. London: Phaidon, plate 189. 


5. Gettens, R.]., and M. E. Mrose. 1954. Calcium sulphate minerals in the grounds 

of Italian paintings. Studies in Conservation (4):182-83. 

6. Valcanover, F. 1990. An introduction to Titian. In Titian, Prince oj Painters. Venice, 

3-28. 

7. The only documentary evidence for it can be fo und in the fifteenth-century 

manuscript in Modena Biblioteca Estense (MS a.T.7.3), fol. 3r. Also see Wallert, 

A. 1984. Orpiment and realgar, some pigment characteristics. MaltechniklRestauro 

(90):45-58. 

8. Gettens, R.]., and G. L. Stout. 1966. Painting Materials: A Short Encyclopedia. New 

York: Dover Publications. 

9. Rome, Biblioteca Casanatense MS 2265, fol. 91r, describes the making of the 

blue glass. On fol. 88r, the manuscript describes the actual processing of the glass 

into a pigment for painting: "A macinare ismalto azuro che poi pingere in suo 

colori. Pilia il smalto et rompelo sotile poi macinelo sopra il porfido cun sorro 

de ovo cun uno pocho di melle. Masinalo sutile como vopi et se andasse secondo 

azonze uno pocho de aqua. Et quando e masinato lava fora il mele et 10 ovo 

cun l'aqua e remanera cossa bellisima." 

10. Wallert, A. 1991. "Cimatura de grana"; Identification of natural organic colorants 

and binding media in mediaeval manuscript illumination. Zeitschrijt ju r Kunsttechnologie 

und Konservierung (5):74-83. 

11. Merrifield, M. P. 1849. Original Treatises on the Arts if Painting, Vol. II. London: 

John Murray, 667. 

12. Kiihn, H. 1964. Griinspan und seine Verwendung in der Malerei. Farbe und Lack, 

(70):703-1 1. 

13. Rome, Biblioteca Casanatense MS 2265, fol. 99r : "A fa r stagnolo verde. Piglia 

vernice liquida e olio de linosa equal misura e meteli denmtro verderame bene 

pulverizato. Poi fa bolire tanto che habia il colore a tuo modo. Poi 10 adopeera." 

The result of this recipe is a complex copper resinate/oleate compound. This 

recipe shows that copper resinates were not always made of Venice turpentine 

as is generally assumed on the basis of a description in De Mayerne's treatise. 

14. Jaffe, M., and K. Groen. 1987. Titian's Tarquin and Lucrezia in the Fitzwilliam. 

The Burlington Magazine, (CXXIX):162-71. 

15. Lucas, A., and J. Plesters. 1978. Titian's Bacchus and Ariadne. National Callery 

Technical Bulletin (2) :25-48, especially 4 1. 

16. Plesters, op. cit., 32-48. 

17. Heidegger, M. 1970. Der Ursprung des Kunstwerkes. Stuttgart: Reclam Verlag, 66. 

126 


Abstract 

The St. Anne Altarpiece (National 

Gallery of Art, Washington, Widener 

Collection) is the largest work attributed 

to the Bruges artist Gerard 

David (ca. 1460-1523). Several panels 

of the altarpiece were examined 

to identify the materials and methods 

used. The results show that the 

center panel was painted using a different 

technique from that for the 

side and predella panels. There are 

distinct groupings of underdrawings 

among the panels. The main panel 

was laid out more specifically at the 

underdrawing stage, with three separate 

drawing campaigns and a correction 

on top of a paint layer. In 

the center panel only, all the major 

fo rms were laid out in the first paint 

stage with a lighter value of the final 

paint color, with multiple layers of 

progressively more saturated paint 

built up to create the final appearance. 

In contrast, the predella and 

side panels were painted directly, using 

fewer layers of paint. The program 

of painting would have allowed 

for workshop participation, especially 

in the main panel. 

Gerard David's St. Anne Altarpiece: Evidence for 

Workshop Participation 

Catherine A. Metzger and Barbara H. Berrie* 



USA 

Introduction 

Gerard David (ca. 1460-1523) is known to have been a member of the painters' 

guild in Bruges by January 1484. It is generally accepted that he operated 

a workshop, evidenced by the use of pricked drawings and his documented 

dispute with Ambrosius Benson over the alleged theft of such drawings (1 , 

2). The recent restoration of the St. Anne Altarpiece provided an opportunity 

fo r its reexamination and technical analysis (3). This paper addresses the evidence 

for workshop participation in the altarpiece. The full results of the 

technical investigation will be published in the future. 

The altarpiece 

The St. Anne Altarpiece depicts St. Anne with the Virgin and Child flanked 

by St. Nicholas (left) and St. Anthony of Padua (right) (Plate 28). John Hand 

has described the altarpiece (4). Six pre della panels are accepted as part of the 

original altarpiece; the three panels at the National Gallery of Scotland illustrate 

three miracles from the life of St. Nicholas, while the three panels at the 

Toledo Museum of Art show three miracles from the life of St. Anthony. 

Another panel, Lamentation of Christ at the Foot of the Cross (The Art Institute 

of Chicago, Mr. and Mrs. Martin A. Ryerson Collection), is possibly part of 

this altarpiece. According to dendrochronological data, David painted the altarpiece 

around 1506, but the stylistic evidence for dating is less precise (5). 

The stiffness and lack of expression of the figures in certain altarpiece panels, 

especially the central panel and the St. Anthony pre della panels, have led to 

doubts regarding the attribution to David. Workshop participation in the St. 

Anne Altarpiece was proposed as early as 1905 (6) . Marlier suggested the participation 

of Ambrosius Benson in the altarpiece (7). Scillia has attributed the 

central panel and the St. Anthony predella panels to an assistant (8). Recently, 

it has been suggested that the entire altarpiece is by a follower of David (9). 

The support 

Figure 1. Secondary eiectroll image of the 

chalk ground of the center panel of Gerard 

David and Workshop's St. Anne Altarpiece, 

ca. 1506. The coccolith is typical of 

those found in northern European deposits. 

The two side panels of the altarpiece, which have been reduced in height, 

now measure 214 X 76 cm. If the semicircular upper profile is recreated 

based on the truncated arcs still visible at the upper extremity of each panel, 

the original height of these panels can be estimated at 236 cm. By comparison 

to contemporary Italianate altarpieces, the central panel can be assumed to 

have been even greater in height. The predella panels are 56 cm high. The 

altarpiece is the largest work in David's oeuvre and one of the largest surviving 

from the period (10). The scale alone suggests that David would have 

had assistance in this production. The patron fo r such a sumptuous and extravagant 

commission is not yet known. 

The central panel is assembled from fo ur boards, each side panel from three. 

From the relative widths of the central and side panels, it is clear that this 

was a fixed, not a closing, altarpiece. The oak panels were prepared in the 

traditional manner with coatings of chalk in glue. Coccoliths in the chalk 

ground prove a natural, and Northern, source (Fig. 1) (11). 


Metzger and Berrie 127

Figure 2. Infrared rriflectogram assembly if the cherub standing 01'/ the throlle (lift side, area if rriflectogram: 

23 X 10 C/'I'/). Type I drawing is visible, characterized by an even, relatively narrow width. 

Rriflectograph by M. Faries. 

The underdrawing 

Infrared reflectography reveals distinct campaigns of underdrawing that can 

be described in four groups (12). Type I is characterized by a narrow, even 

line that skips over the surface of the ground. It is the first drawing stage in 

all three of the main panels and the St. Nicholas predella panels. There is only 

a small area of type I drawing in one of the St. Anthony predella panels. The 

drawing made using this material is summary. An example of this drawing 

type in the cherub on the throne is illustrated in Figure 2. This underdrawing 

128 


Figure 3. Infrared reJlectogram assembly of the Virgin's head (area of reJlectogram: 23 X 17 em). 

Type II underdrawing characterized by variable density and line width. ReJlectograph by M. Faries. 

can be seen in the cross section illustrated in Plate 29a, obtained from the 

location marked in Figure 3. The underdrawing layer is only a few microns 

thick. 

Type II underdrawing is characterized by variable line width and intensity. 

These properties, together with the distinctly dropletlike forms visible in some 

places, suggest type II lines were produced using a liquid. Type II lines are 

found in the figural group of the center panel, where they describe the forms 

more firmly. Figure 3 shows the infrared reflectogram of St. Anne's head. 

Here sketchy type I lines have been overdrawn with the type II underdrawing. 

Nearly all the underdrawing in the St. Anthony predella panels is executed 

with a liquid type II drawing material that lays out the composition concisely. 

In the central figural group of the main panel, however, the line is often rigid 

and repetitive as, for example, in the shading to the right of the Christ Child. 

In addition to these kinds of underdrawing there is a third form, type III, 

that can be seen with the naked eye in St. Anne's red robe (Fig. 4). These 

lines are characterized by their bold appearance with handling qualities more 

similar to that of a paint than of an ink. A cross section shows type III was 

laid down thickly, and then covered with a translucent layer (Plate 29b). Type 

III has only been found in St. Anne's red drapery and no other place in either 

the side or predella panels. A fo urth type of marking of fo rm is evident in 

the reflectogram of the Virgin's sleeve (Fig. 5). Here, a dark irregular line lies 

over the first layers of paint (Plate 29c). The subsequent layer of paint follows 

the line denoted by this "correction" layer. Optical microscopy showed the 

pigment here is charcoal (13). Two panels, The Miracle if the Purse and The 

Metzger and Berne 129

Figure 4. Infrared niffectogram assembly of part of St. Anne's red drapery (area of riflectogram: 24 X 

21 em). Type III underdrawing, which is characterized by its painterly quality, is illustrated. Riflectograph 

by M. Faries. 

Miracle of the Dead Child, have wandering marks that recollect the appearance 

of the correction in the Virgin's sleeve. The dotted quality of these lines has 

led researchers to interpret them as having been produced by pouncing. However, 

microscopical examination of the paint surface showed that in these 

locations, the black paint protrudes through the uppermost white layers, giving 

an impression of a series of dots. These spotty "correction" lines in the 

predella panels were not followed in the final paint layers, as they were in the 

center panel. 

The looseness and lack of formality of the underdrawing in the two side 

panels and the predella panels, coupled with the fact that there are many 

repositionings from the underdrawn to the paint stage, suggests that someone 

with the authority to interpret the sketch and deviate from it, most likely the 

master himself, was working on the side panels and the predella panels. The 

careful additional delineation of outlines in the St. Anne suggests that someone 

who was less intuitive was to work on this panel. The "correction" mark 

in the Virgin's sleeve, which is followed in the final paint layer, also suggests 

a second hand in this process. The " correction" marks in the predellas, which 

are ignored in the final paint stage, suggest the finishing touches were completed 

by someone who could and did act independently. Participation of 

workshop members at the underdrawing stage has been noted. For example, 

Fairies found at least three hands in the underdrawing in the Crucifixion triptych 

in the Rijksmuseum Het Catherijnenconvent in Utrecht (14). The master,Jan 

van Scorel, drew only in the exterior wings; the drawing in the interior 

wings and the central panel was executed by two anonymous assistants. 

130 


Figure 5. Irifrared niflectogram assembly of the Virgin 's sleeve, showing the "correction" layer which 

appears as dotted lines (area of refiectogral1l: 24 X 21 an). Refiectograph by M. Faries. 

The paint layers 

The preparatory layers. Staining the cross sections shows that the chalk ground 

in the main panels was not coated with a layer of glue before application of 

the underdrawing. In some of the cross sections, however, there is a dusting 

of red and black particles in an oleaginous layer; these particles are especially 

visible under the green cloth of honor behind St. Anne and the Virgin. 

Scattered particles such as this are rarely noted in painting cross sections; 

however, Coremans, et al. and later Brinkman, et al. found a similar layer in 

the Ghent altarpiece (15, 16). There is a clearer indication of the use of a 

sealing layer over the underdrawing in St. Anne's red drapery. Here the hatching 

lines, which are visible to the naked eye, have a translucent layer over 

them, separating the red paint from the drawing lines. 

The X-radiographs of all six predella panels have clear indications for an 

overall, randomly brushed underlayer which must contain lead white to register 

in the X-radiograph. The few samples from the pre della panels show 


there is a lead-white underpaint or imprimatura. X-radiographs of the central 

panel show a similar vigorous, striated, and randomly applied layer in some 

areas. 

The relationship between the underdrawing and the final image. In the main and 

predella panels, the layers of paint follow the outlines of the underdrawing in 

a general fashion. There are many major and minor departures from the 

underdrawn outlines, however, and additional elements have been added to 

the composition in the paint stage without any redrawing. There are numerous 

examples in the predel1a panels. For example, in St. Nicholas Slips a 

Purse through the Window if an Impoverished Nobleman, the hand and purse 

were painted slightly lower than they are drawn, and a shoe and stocking that 

were not included in the underdrawing appear. The praying hands of the 

foremost boy in St. Nicholas Restores Three Dismembered Ch£ldren to Life were 

lowered in the paint stage and St. Nicholas's crosier was raised in the same 

scene. 

In the side panels, St. Anthony's foreground fo ot is drawn in a lower position 

than it was painted. The elaborate gold crosier that St. Nicholas grasps is 

changed from the drawn to the painted version. These changes were made 

at the initial stage of painting without any redrawing. 

In the central panel, there are fewer changes from the underdrawing to the 

final image, however, the Virgin's head was painted higher than it was drawn. 

The X-radiograph shows this change was made after the first layer of St. 

Anne's wimple had been laid down, after which the head was painted in its 

new position, without any detectable redrawing. 

The pigments and layer structure. The painting technique of the center panel of 

the altarpiece is similar to that found in the Ghent altarpiece (17). There is 

evidence fo r an oil-based sealing layer toned with red and black pigments 

over the underdrawing and under the paint. The paint structure is multilayered 

and the colors are worked from light to dark. The cross sections show 

that each large area of color was blocked in using a light tone of the final 

color. For example, the Virgin's mantle is underpainted in pale blue (Plate 

29d) ; a blue-green layer blocked in the whole area of the cloth of honor 

(Plate 2ge); and the throne, on which statues of putti stand, is underpainted 

with a lighter gray (Plate 29a) (18). Type III underdrawing provides the midtone 

in St. Anne's drapery, which was painted using two layers of vermilion 

and red lake covered by multiple layers of a transparent red glaze (Plate 29b). 

The paint layer structure in the side panels is not the same as in the center; 

the build-up of color and form is quite different. St. Anthony's robe has only 

a single violet-gray layer (Plate 29f). St. Nicholas's chasuble was painted by 

laying in a yellow ochre base, on which shadows were created with a thick 

dark layer of black pigment mixed with vermilion. Whereas in the center 

panel the layers of paint are always tonally related (e.g., red covering red) , in 

St. Nicholas yellow paint is used as a base for red and green in the saint's 

vestments. In the red embroidered chasuble, the cut velvet was painted with 

a layer of vermilion mixed with red lake. Red glazes were applied to depict 

the texture. 

The technique of Lamentation at the Foot if the Cross, as described by Butler, 

is similar to the technique employed on the predel1a panels in the use of an 

imprimatura containing lead white and in the relatively simple layer structure 

(19). Certain similarities to David's The Virgin and Child with Saints and a 

Donor are seen in the paint structure (20). 

Summary 

The central figural group of the center panel was produced using a different 

technique from the rest of the altarpiece. Its structure is similar to those found 

in earlier fifteenth-century paintings, such as in the Ghent altarpiece (21). 

This method of working-sealing the underdrawing, laying in light tones, 

132 


followed by modeling layers and glazes, with the addition of new drawing as 

guidelines-would have lent itself to the participation of the workshop. The 

repositioning of the Virgin's head during the painting process could indicate 

the continued involvement of the master. Painting with modifications and 

corrections fronl the underdrawing, and modeling of tone using very few 

layers of paint, as is seen in the side panels, requires self-assurance and authority; 

it would be harder to use assistants in the production of a painting 

when this was the method of working. Workshop participation in the center 

panel rather than the side and predella panels is not usually expected, although 

there is precedent for it. 


The authors thank Molly Faries, Maryan Ainsworth, and John O. Hand for many 

helpful discussions and the staff of the Department of Visual Services, National Gallery 

for their assistance. 

Notes 

1. Taubert,]. 1975. Pauspunkte in Tafelbildern des 15. und 16.Jahrhunderts. Bulletin 

de l'Institute Royale du Patrimoine Artistique (XV) :390. 

2. Bruges, State Archives, Aanwinsten 3804: Register van de burgerlijke vonissen 

van de Schepenkamer te Brugge 1518-1519, fols. 82v-83r. See also Parmentier, 

R. A. 1937. Bescheiden omtrent Brugsche Schilders in de 16e eeuw: 1. Ambrosius 

Benson. Handelingen van het genootschap Societe d'Emulation te Brugge LXXX, 92- 

3. See also Marlier, G. 1957. Ambrosius Benson et la peinture a Bruges au temps de 

Charles-Quint, Damme, 42-43. 

3. The St. Anne Altarpiece was the subject of the Curatorial Colloquy III of the 

Center fo r Advanced Study in the Visual Arts, 20-24 May 1991. 

4. Hand, ]. O. 1986. In Early Netherlandish Painting. Eds. ]. 0. Hand and M. Wolff. 

Washington: National Gallery of Art, 63-74. See also Hand,]. O. 1992. The Saint 

Anne Altarpiece. Exhibition brochure. Washington: National Gallery of Art. 

5. Klein, P 1986. Dendrochronological Report on The St. Anne Altarpiece. 23 April 1986. 

Klein determined that there are two groups of boards. Three of the boards, two 

from St. Anthony and one from St. Nicholas, come from the same tree. The 

felling date of this tree was estimated as 1496 (using the assumption of fifteen 

sapwood growth rings) . The other group of boards have rings that suggest a 

felling date of 1402. Assuming the wood was stored for ten years, Klein estimates 

the altarpiece could have been manufactured from 1506 onward. 

6. von Bodenhausen, E. E 1905. Gerard David und seine Schule. Munich: E Bruckman. 

7. Marlier, op. cit. (note 2), 46-50. 

8. Scillia, D. G. 1975. Gerard David and Manuscript Illumination in the Lowlands. 

Ph.D. diss. Case Western Reserve University, Ohio. 

9. van Miegroet, H.]. 1989. Gerard David. Antwerp: Mercatorfonds, 313-14. 

10. Verougstraete-Marcq, H., and R. Van Schoute. 1989. Cadres et supports dans la 

peinture fiamande aux 15e et 16e siecles. Heure-Ie-Romain, 9, 76-77. 

11. Gettens, R.]., E. W FitzHugh, and R. L. Feller. 1993. Calcium Carbonate Whites. 

In Artists' Pigments: A Handbook of their History and Characteristics, Vo!' 2. Washington: 

National Gallery of Art, 203-26. 

12. Infrared reflectography of the main panels was undertaken (1981-1982) by Molly 

Faries and in 1990 by Maryan Ainsworth. Reflectography of the Toledo panels 

was completed in 1984 by Faries, and of the Edinburgh panels in 1990 by the 

National Gallery of Art. Faries used a high resolution Grundig FA 70H video 

camera with a TV Macromar 1:2.8/36 mm lens and a Kodak 87 A Wratten filter 

on a Hammamatsu N214 vidicon detector. The assemblies were prepared by 

photographing the image as it appeared on the monitor and pasting the photographs 

together. 

13. A Leitz Orthoplan microscope was used with apochromat lenses. The samples 

were mounted in Cargille Meltmount and examined at xl00-630. 

14. Faries, M. 1986. Division of labour in the production of panel paintings in the 

workshop of Jan van Score!' In Kunst voor de beeldenstorm. Exhibition catalogue. 

's-gravenhage: 113. 

15. Coremans, P, R. ]. Rutherford, and ]. Thissen. 1952. La technique des Primitifs 

Flamands. Etude scientifique des materiaux, de la structure et de la technique 

picturale. Studies in Conservation (1):1-29. 


16. Brinkman, P W. E, L. Kockaert, L. Maes, L. Masschelein-Kleiner, E Robaszynski, 

and E. Thielen. 1984. Het lam Godsretabel van van Eyck. Een heronderzoek 

naar de materialen en schildermethoden. 1. De plamuur, de isolatie!aag, de tekening 

en de grondtonen. Bulletin Institut Royale Patrimoine Artistic-KIK (20):137- 

66. 

17. Brinkman, P W. E, L. Kockaert, L. Maes, E. M. M. Thielen, and J. Wouters. 1988- 

1989. Het lam Godsretabe! van van Eyck een Heronderzoek naar de materialen 

en schildermethoden. 2. De hoofdkleuren blauw, groen, gee! en rood. Bulletin 

Institut Royale Patrimoine Artistic-KIK (20):26-49. See also Coremans, et al., op. 

cit. and Brinkman, et a]., 1984, op. cit. 

18. A lEOL 6300 scanning electron microscope and a Link energy dispersive spectrometer 

with the Super ATW Si(Li) detector were used. Analysis of a green 

particle in the bottom layer of paint in the cloth of honor showed it was composed 

of copper, chlorine, and oxygen. 

19. Butler, M. H. 1976. A technical investigation of the materials and technique used 

in two Flemish paintings. Museum Studies (The Art Institute of Chicago) (8):59- 

71. 

20. Wyld, M., A. Roy, and A. Smith. 1979. Gerard David's "The Virgin and Child 

with Saints and a Donor." National Callery Technical Bulletin, Vol. 3. London: 

National Gallery, 5-65. 

21. Brinkman, 1988-1989, op. cit. See also Coremans, et al., op. cit. and Brinkman, 

et al., 1984, op. cit. 

134 


Abstract 

Based on extensive information from 

cross sections and infrared reflectography, 

this paper presents some aspects 

of painting technique that were 

held in common by Jan van Scorel, 

the head of a productive sixteenthcentury 

North Netherlandish workshop, 

and Maarten van Heemskerck, 

his best-known assistant. Some of the 

idiosyncrasies in Heemskerck's painting 

technique differ from Scorel's 

studio routine and are more apparent 

in this artist's early works. 

Maarten van Heemskerck and Jan van Scorel's Haarlem 

Workshop 

Molly Faries* 

Henry Radford Hope School of Art 

Indiana University 

Bloomington, Indiana 47405 

USA 

Christa Steinbuchel 

Wallraf-Richartz-Museum 

Bischofsgarten 1 

W-5000 Kaln 

Germany 

J. R. J. van Asperen de Boer 

Groningen University 

Department for the History of Art and Architechture 

Oude Boteringestraat 

p. 0. Box 716 

9700 AS Groningen 


Introduction 

From the records of the Mariakerk in Utrecht where Jan van Scorel held 

clerical office, the precise dates of his stay in Haarlem are known: 29 April 

1527 to 28 September 1530. Karel van Mander's Schilder-Boeck, the primary 

source on the workshop Scorel established in Haarlem during that time, reports 

that Scorel rented a house in order to take on students. In the biographies 

of other sixteenth-century North Netherlandish painters, including 

Jan Swart van Groningen or Jan Vermeyen, Mander implies that these painters 

were either Scorel's students or were somehow in contact with his shop. 

Modern scholars have added other likely (and not so likely) names to Mander's 

list, including Comelis Buys, Herman Postma, and Jan Stephan van Calcar, 

but the assistant whom Mander described at some length was Maarten 

van Heemskerck. According to Mander, Heemskerck applied himself diligently 

in Scorel's Haarlem studio, eventually producing works that were indistinguishable 

from those of his master. In Mander's rather dramatic account, 

Heemskerck finally surpassed the de Const (artistry) of his master, only to be 

dismissed from the shop, ostensibly because of Scorel's jealousy (1). However 

anecdotal this story might seem, art historians have indeed had difficulties 

distinguishing early works by Heemskerck from Scorel's work. For most of 

the twentieth century, early Heemskerck artworks were attributed to Scorel; 

it was not until the 1980s that several key attributions were changed, primarily 

in the 1986 Art Before Iconoclasm exhibition (2). Jeff Harrison's recently 

published dissertation on Heemskerck is the first to outline a new and plausible 

chronology for his early works (3). 

Jan van Scorel's technique 

Full technical studies have been carried out by Molly Faries and J. R. J. van 

Asperen de Boer on a number of paintings by Jan van Scorel from this period, 

including examination by binocular microscope, infrared reflectography, X 

radiography, dendrochronology, and sampling. Until the new shifts in attribution, 

few early works of Heemskerck had been studied as thoroughly. This 

situation changed with the recent cleaning of Lamentation (Wallraf-Richartz 

Museum, Cologne), a painting attributed variously to Scorel, Heemskerck, or 

an anonymous artist from the same period. The research in conjunction with 

this restoration, carried out under the direction of Christa Steinbuchel, has 


Faries, Steinbuchel, and van Asperen de Boer 135

Figure 1. Jan van Scoref, Baptism of Christ, ca. 1527-1530. Photograph by M. Faries (after 

Dingjan), courtesy of Frans Hafsmuseum, Haarlem. 

provided critical comparative evidence (4). The results help to define the art 

historical attributions and suggest some changes to Harrison's chronology. 

This material not only brought greater clarity regarding the typical painting 

procedures used in Scorel's shop, but also signaled the steps Heemskerck took 

in his evolution away from Scorel's shop. 

During the Haarlem years, Scorel standardized his painting technique in response 

to the needs of an active workshop (5). The most overt clue to the 

standardization is revealed in the layout of underdrawings from this period. 

The Baptism if Christ in the Frans Halsmuseum, Haarlem, must have been 

one of the most prestigious commissions the artist received 1527-1530 (Fig. 

1). The layout of this painting is fully worked out, from the assured contours 

and loosely marked shaded zones in the figures of the main scene to the light 

and dark bands in the landscape and other background detail. Areas where a 

known motif was to be placed were no longer just left as a blank space in 

the underdrawing, but marked with indicative shapes referring to the motif. 

The underdrawing was laid out as a more recognizable and emphatic guide. 

The Baptism was also painted following what had evolved into Scorel's usual 

practice. This practice comprised the following: (a) an application of lead 

white as a continuous intermediate layer between the ground and paint, (b) 

an underdrawing in black chalk on this layer, and (c) a preference fo r certain 

paint-layer structures and color combinations. A white intermediate layer certainly 

occurs in artistic groups other than that of Scorel's workshop. It has 

been found in some early German panels, and sporadically in the Hans Memling 

and Gerard David groups, for instance. Scorel, however, could not have 

learned the use of such a layer from his master in Amsterdam. It begins to 

show up in some of Scorel's early works produced during his journey to Italy, 

but only appears consistently after his return north. It was the three-fold 

fu nction of this layer that was peculiar to the efficacy of Scorel's painting 

technique: it isolated the ground, added to the painting's luminosity, and provided 

"tooth" for the underdrawing, a fu nction almost unique to the Scorel 

group (6). In many infrared documents, the underdrawing can be seen to skip 

or crumble on top of this ridged surface. One cross section from Baptism 

shows this typical paint layer structure: a first layer consisting of lead white 

at a maximum thickness of 12 fL, a second layer composed of clumps of black 

136 


o . (Vv..-.< 

1 2. 

underdrawing, and a third layer consisting of a dark green mixture of green, 

blue, and white; no ground is used (Fig. 2). In addition, Scorel often paints 

wet in wet with the usual run of pigments, except for his use of natural 

ultramarine and a blue-over-rose structure. The latter superimposition of colors, 

with ultramarine on the surface (as seen again in a cross section from 

Baptism), has been linked to Scorers observation of Italian painting technique 

(Plate 30) (7). 

Figure 2. Cross section from Baptism with 

measurements in microns: layer 1, traces of 

the intermediate white layer; layer 2, particles 

of black chalk underdrawing; layer 3, mixture 

of light green, blue, and white. Photograph 

by M. Faries (after J. R. J. van Asperen de 

Boer). 

Marteen van Heemskerck's technique compared 

Before the conservation of the Cologne Lamentation and the recent revisions 

in art historical opinion, what was known of Heemskerck's painting technique 

showed little connection to Scorel (Fig. 3). The current acceptance of Lamentation, 

long attributed to Scorel, as a Heemskerck by scholars Faries, Harrison, 

and W Th. Kloek is by no means unchallenged. A study of the painting's 

technique shows why a separation of "hands" has been so difficult. 

Lamentation was begun as a Scorel studio piece and finished as a Heemskerck, 

as seen in both the underdrawing and the painting technique. Infrared reflectography 

examination, undertaken by Faries in 1991, disclosed exceptionally 

complicated compositional change in this work (8). The painting was 

taken through as many as six stages while the composition was changed from 

a profile Scorelesque Entombment scene to a Heemskerck frontal presentation 

of the Lamentation. 

The numerous samples required because of the piece's complex restoration 

history provided more than enough evidence about Lamentation's painting 

procedure. The initial preparation of the ground and position of the underdrawing 

in the paint layer structure match the standard practice of Scorers 

Haarlem shop. A thin layer of lead white covers the entire surface of the 

ground; it appears consistently in sections that show the entire paint layer 

structure. The black chalk underdrawing is found on top of the intermediate 

layer (9). Although some of the lines of the different compositional stages 

must have crossed, no noticeable overlapping or disjunctures in any samples 

that include the underdrawing layer can be seen. The presence of the lead 

white intermediate layer coupled with the black chalk underdrawing can be 

Figure 3. Maarten van Heemskerck, Lamentation of Christ, ca. 1530. Photograph by C. Steinbachel, 

courtesy of Wallraj-Richartz-Museum, Cologne. 

Faries, Steinbiichel, and van Asperen de Boer 137

considered a kind of technical signature, although not of an individual hand: 

It is the signature of Scorel's shop. Most of the pigments used in completing 

the image are, of course, also common to those found in Scorel's paintings. 

No ultramarine has been found, however, and the blue used throughout the 

painting is azurite. This no doubt proves not only that Jan van Scorel is the 

only north Netherlandish artist documented to have used natural ultramarine 

in his works, but that he also used it selectively, saving it for his most important 

commissions (10). It was, therefore, not available to his assistant, or at 

least not for this painting. Not unexpectedly, the blue-over-rose paint layer 

structure is also lacking in Lamentation. 

Figure 4. Infrared reflectogram detail if 

Lamentation, showing the dark undermodelil1g 

strokes if Nicodemus's turbal1 in the upper 

background. I'!frared reflectography by M. 

Faries. 

;; 

I 

I 

s) (.10 

'1) 10-'3£ 

3) q -2-1 

Z) g-I& 

Figure 5. Cross sectiol1 from Maarten van. 

Heemskerck's St. Luke with measuremen.ts in 

microns: layer 1, possible trace of ground; 

layer 2, gray intermediate layer; layer 3, 

compact layer if black with black particles; 

layer 4, admixture if white, black, and red; 

layer 5, white layer with one large azurite 

crystal; layer 6, varnish. Photograph by M. 

Faries (after J. R. J. van Asperen de Boer). 

Other ways in which the painting practice in Lamentation deviates from ScoreI's 

standard require a different explanation, since these differences form critical 

links with other paintings that can now be considered early Heemskerck. 

Dark modeling strokes executed with a brush have been detected in the 

finishing stages of Lamentation. Some strokes that underlie the surface colors 

can be disclosed by infrared (Fig. 4). Other black-colored modeling is applied 

in the final paint layer, especially in reds. The cross section from John the 

Evangelist's drapery clearly shows the black particles in the vermilion and red 

lake admixture (11). 

Although Scorel's Baptism and Heemskerck's Lamentation are the focus of this 

study, their techniques may be compared to a substantial amount of related 

technical material. Ample evidence proves that the paint-layer structure described 

above typifies the Scorel group paintings during Scorel's time in Haarlem. 

In addition, since nine paintings now attributed by some scholars to the 

early Heemskerck period have also been examined by various technical 

means, the evolution of Heemskerck's painting technique can be surveyed 

(12). Of the nine paintings, only four have been studied with the binocular 

microscope and sampled so far. At least three paintings use the blue-overrose 

structure typical of the Scorel group although, in each case, the blue 

pigment is azurite. Other aspects of the paint layer structure can be documented 

by X-ray or infrared reflectography. The infrared vidicon distinguishes 

pure colors that look black to our eyes but are transparent to the eye of the 

vidicon from "true" blacks, which remain opaque in reflectograms. This instrumental 

technique can therefore locate black modeling or under modeling 

in the same way it locates underdrawings. Dark modeling strokes used to 

block in forms can also be found in at least three other works in the hypothetical 

early Heemskerck group. The technique is unknown in the Scorel 

group. The presence of a thin layer of lead white over the ground can be 

verified in cross sections, as well as by X-radiography where it appears as 

broadly-brushed streaking. This practice is evident in a majority of works in 

the early Heemskerck group. It is the coupling of underdrawing with this 

layer, however, which undergoes a change; and finally, the character of the 

layer itself changes. In Lamentation, the black chalk underdrawing is supplemented 

by dark modeling and undermodeling in paint. Several other paintings 

have a fo rm of underdrawing, but it is executed in a dark, paintlike 

substance rather than chalk. In four works, the underdrawing is undetectable. 

No underdrawing could be detected in the St. Luke Painting a Portrait of the 

Virgin and Child (1532), an indisputable early Heemskerck, and no underdrawing 

appeared in samples. In this work Heemskerck used a gray intermediate 

layer, a very different basis upon which to build up the fo rms of his 

painting (Fig. 5) (13). 

Conclusion 

Heemskerck must have worked in Scorel's studio for all or part of the period 

from 1527 to 1530, and after Scorel left Haarlem in September 1530, Heemskerck 

continued to paint in the city as an independent master until he himself 

left for Rome in 1532. The artist moved from a graphic black-on-white 

layout to more subtle forms of shading and undermodeling; based on changes 

in painting technique, an evolution is proposed for Heemskerck in this period. 

138 


The key position of Lamentation in this proposed development is obvious, for 

it encapsulates both painting methods. Generally, those paintings that still rely 

on underdrawing (on the intermediate white layer) as part of the basic layout 

have an early date. Whether or not the paintings were produced in Scorers 

shop probably cannot be proven, but they were painted following Scorers 

shop routine. The paintings without underdrawing but with black-colored 

modeling appear to be later works, closer to Heemskerck's dated works of 

1531 and 1532. A clearer art historical picture of the early Heemskerck is 

now possible, but surely the blending of Scorers painting practices into the 

build-up of some of the earliest works explains the difficulties art historians 

have had in the past in attributing these paintings. 

Notes 

1. van Mander, K. 1604. Het Schilder-Boeck. Haarlem. Davaco ed. 1969, fol. 245. 

2. Kunst voor de beeldenstorm. 1986. Exhibition catalogue. Eds. J. P. Filedt Kok, W. 

Halsema-Kubes, W. Th. Kloek. Amsterdam: Rijksmuseum. Especially catalogue 

number 69, Rest on the Flight into Egypt (National Gallery of Art, Washington, 

D.c.) and catalogue number 72, Portrait oj a Young Scholar (Boymans van Beuningen 

Museum, Rotterdam), which were definitively assigned to Heemskerck. 

Around this same time, Adam and Eve (Hatfield House, Marquess of Salisbury) 

was being shifted from Scorel to Heemskerck by the research of Faries and 

Harrison. 

3. Harrison, J. C. 1992. The Paintings oj Maerten van Heemskerck. (Ph.D. diss., 1987, 

University of Virginia.) Ann Arbor, Michigan, esp. 11-34. 

4. For the restoration, see C. Steinbuchel, "The investigation and restoration of 

Maarten van Heemskerck's Lamentation oj Christ. " In Le dessin sous-jacent dans la 

peinture, Colloque X: Le dessin sous-jacent dans le processus de creation, eds. R. Van 

Schoute and H. Verougstraete-Marcq. Universite Catholique de Louvain, Institut 

superieur d'archeologie et d'histoire de l'art. Forthcoming. 

5. Faries, M. 1987. Some results of the recent Score! research: Jan van Scorel's 

definition of landscape in design and color. In Color and Technique in Renaissance 

Painting, Italy and the North. Ed. M. B. Hall. New York, 89-103, esp. 96. 

6. van Asperen de Boer, J. R. J., M. Faries, and J. P. Filedt Kok. 1986. Painting 

technique and workshop practice in northern Netherlandish art of the sixteenth 

century. In Kunst voor de beelderlstorm. Exhibition catalogue. Eds. J. P. Fi!edt Kok, 

W. Halsema-Kubes, W. Th. Kloek. Amsterdam: Rijksmuseum, esp. 108. 

7. Faries, op. cit., 92. 

8. For these infrared reflectography results, see M. Faries, "Attributing the layers of 

Heemskerck's Cologne Lamentation oj Christ." In Le dessin sous-jacent dans la peinture, 

Colloque X: Le dessin sous-jacent dans le processus de creation. Eds. R. Van 

Schoute and H. Verougstraete-Marcq. Universite Catholique de Louvain, Institut 

superieur d'archeologie et d'histoire de l'art. Forthcoming. 

9. Sample numbers 1,2, and 27.1 show this structure. Communication from Christa 

Steinbuchel to Molly Faries, 7 October 1994. 

10. van Asperen de Boer, et aI., op. cit., 109. 

11. Sample number 18. Communication from Christa Steinbuchel to Molly Faries, 

7 October 1994. 

12. The works studied, in presumed chronological order, are as follows: ca. 1527- 

1530: (1) Lamentation oj Christ, private collection, New York; (2) Crucifixion, 

Detroit Institute of Arts; (3) Rest on the Flight into Egypt, National Gallery of Art, 

Washington, D.c.; (4) Lamentation oj Christ, Wallraf-Richartz-Museum, Cologne; 

(5) 1531: Portrait oj a Scholar, Museum Boymans-van Beuningen, Rotterdam; (6) 

1531?: Baptism oj Christ, Staatliche Museen Preuf3ischer Kulturbesitz, Berlin; (7) 

1532: Christ as the Man oj SorroUls, Museum voor Schone Kunsten, Ghent; (8) 

1532: St. Luke Painting a Portrait oj the Virgin and Cltild, Frans Halsmuseum, 

Haarlem; (9) ca. 1530-1532: Family Portrait, Staatliche Kunstsanmilungen, Kassel. 

For the lack of underdrawing revealed in the 1529 so-called Bicker Portraits in 

Amsterdam, Rijksmuseum, see Kunst voor de beeldenstorm, op. cit., catalogue number 

71.1-2. 

13. van Asperen de Boer, J. R. J. 1987. A technical study of some paintings by 

Maarten van Heemskerck. In Color and Technique in Renaissance Painting, Italy and 

the North. Ed. M. B. Hall. New York, 105-1 14. 

Faries, Steinbiichel, and van Asperen de Boer 139

Abstract 

In the 1620s, a generation of Dutch 

landscape artists began to work in a 

naturalistic mode very different from 

that of the earlier generation of 

Flemish mannerist landscape artists, a 

number of whom had recently emigrated 

to the Northern Netherlands. 

The change from fantastic landscape 

to representations of Dutch scenes 

reflected political and economic 

changes as the Northern Netherlands 

established independence from Spanish 

domination. This stylistic change 

is reflected in changes in the painting 

materials and practices of the realist 

painters. In the 1620s, Dutch 

painters of naturalistic landscape 

adapted the efficient working practices 

of the Flemish landscape 

painters. They replaced the refined 

handling of paint and bright colors 

of the mannerist painters with limited 

tonalities and an abbreviated handling 

of paint to create convincing 

views of the Dutch landscape. 

Style and Technique in Dutch Landscape Painting 

in the 1620s 

E. Melanie Gifford 

Scientific Research Department 



USA 

Introduction 

The early years of the seventeenth century saw a striking change of style in 

landscape painting in the Netherlands. At this time, Flemish landscape specialists 

produced paintings characterized more by fantasy than by close observation 

of nature. In the 1580s and 1590s, large numbers of Flemish artists, 

some of them landscape specialists, had emigrated to Holland to escape the 

political and economic hardships of the rebellion against Spanish rule (1). 

Around 1620, a very different, naturalistic style of landscape painting developed 

in the Northern Netherlands, particularly around Haarlem. These paintings, 

often based on drawings after nature, created distinctive images of the 

Dutch landscape (2, 3, 4). 

Such a dramatic change of style raises fascinating questions about how style 

develops. Was the new, naturalistic landscape style sparked by the arrival of 

the Flemish immigrants, or does it reflect an indigenous artistic sensibility? Is 

artistic style dependent on painting practice learned from other artists, or do 

artists modifY their practice to meet the demand for new styles? 

In an ongoing technical study, the author has been looking for material evidence 

for landscape artists' artistic concerns (5). This study seeks to characterize 

the differences of technique between Flemish mannerist landscapes of 

the turn of the sixteenth and seventeenth centuries, landscapes by Flemish 

immigrants to the Northern Netherlands during the same period, and the 

first naturalistic Dutch landscapes, which date from the second and third 

decade of the seventeenth century (6). In defining the artists' choices of 

working methods and painting materials, the study seeks to expand our 

understanding of the motivations, both fo rmal and practical, behind the 

development of the naturalistic Dutch landscape in the 1620s. The present 

contribution centers on developments in the handling of paint in these landscapes. 

From the paintings studied to date, I have chosen five paintings as 

illustration: a Flemish mannerist landscape, a mannerist landscape by a Flemish 

immigrant, two naturalistic landscapes by Dutch artists, and a somewhat later 

Dutch "tonal" landscape. A later, more complete publication will present full 

analyses of the paintings' materials. 

Flemish mannerist landscape 

In the Flemish landscape painting tradition, space is defined by formal conventions 

that convey an illusion of recession. Dark and light passages alternate, 

often with shadowed foreground repoussoirs sharply outlined against a brightly 

lit area beyond. Optical phenomena such as atmospheric perspective are represented 

not illusionistically, but almost symbolically, by a space organized into 

three zones defined by distinctly different tonalities. The darkened foreground 

zone is typically a rich brown, the middle ground is green, and the distance 

is a clear blue. 

Study of the technique of Flemish mannerist landscapes reveals refined handling 

combined with efficient working methods, which contributed to the 

high output of busy workshops. The compositions were usually laid out with 

a fairly complete underdrawing, most often including the figures, which were 

planned as integral parts of the composition. With the underdrawing as guide, 

140 


Figure 1. Workshop of Jan Bruegel, Noah's Ark. Panel, 65. 4 X 94.5 em. The Walters Art Gallery, Baltimore (3 7. 1998). 

space for the figures was left in reserve as the landscape was painted. Some 

landscape specialists collaborated with figure painters, in which case the figures 

would be painted over a fully completed landscape. Other workshops 

divided up production in a sort of assembly line, with junior associates filling 

in minor details. 

The painting materials in the Flemish mannerist group, while not unusual, 

were used in ways that enhanced the clear colors. Typically, a light-colored, 

opaque ground maintained the luminous tonalities of the paint, but rarely is 

the ground itself visible in the final image. Instead, the three dominant zones 

of the composition were laid out with broad areas of underpaint: warm brown 

in the foreground, soft green in the middle zone, and light blue in the distance. 

The final image was worked up over these base colors. Working from the 

back of the composition to the foreground, each zone was painted in turn 

in a limited range of colors harmonizing with that of the underpaint. 

A painting of Noah 's Ark at the Walters Art Gallery was produced by the 

workshop of Jan Bruegel (Fig. 1) (7, 8). In this painting, the techniques described 

above were adapted for team production in a workshop setting. As 

shown in a paint cross section from high in the central group of trees, a thick 

white chalk ground was sealed and lightly toned by a translucent imprimatura 

(layers 1 and 2) (Plate 31) (9). Over the preparation and underdrawing, the 

sky was laid in first, followed by a base color loosely brushed for each particular 

area of the composition; the upper two layers of the cross section show 

the sky (layer 3), which extends under the foliage base tone in the upper part 

of the trees, and the clear green underpaint (layer 4) of the lightest passage 

of the central group of trees. A skilled painter painted the main foliage elements 

and the figures and animals, which are the primary subjects of the 

painting, onto the broad areas of underpaint. Only after these major elements 

GijJord 141

Figure 2. Roelandt Savery, Landscape with Animals and Figures, signed and dated 1624. Panel, 54.6 X 91.4 cm. National Gallery if Art, 

Washington, D. C. (1989.22. 1). Gift of Robert H. and Clarice Smith in honor if the fiftieth anniversary if the National Gallery of Art. 

were completed did a workshop assistant fill in minor foliage details. The 

assistant's work has a mechanical, repetitive quality, and shows a cautious respect 

for the work of the master; the monkey on a low branch at the left is 

surrounded by a halo of the base color, where the assistant scrupulously avoided 

overlapping the master's work with his blades of grass. 

Landscapes by Flemish immigrants 

Landscapes by Flemish immigrants such as Gillis van Coninxloo, who arrived 

in Amsterdam from Frankenthal in 1595, and Roelandt Savery, who arrived 

with his family in 1591, are clearly in the Flemish tradition, both compositionally 

and technically (10). There are, however, variations of emphasis in 

the compositions that have consequences for the technique. The compositions 

retain the conventions of Flemish landscape. Dark and light passages are 

strongly juxtaposed and the three zones in brown, green, and blue organize 

the recession into space, but there is a much greater emphasis on the nearer 

zones. In Coninxloo's forest landscapes, such as the Landscape with Hunters of 

1605 in Speyer, the second and third zones are reduced to glimpses caught 

through the dense growth of monumental trees that fills the foreground zone 

(11). The result is a newly limited tonal range, one that emphasizes the browns 

and deeper greens of the middle and foreground. This self-imposed limitation 

is in itself a first step toward naturalism. 

The techniques of these immigrant artists are consistent with contemporaries 

still working in Flanders. The compositions were planned in underdrawings 

that made provision for all but the minor figures. Over a light-colored ground 

the artists established a base color for each area with a brushy underpaint; 

each area was then worked up with loosely painted, final details in a color 

that harmonized with the underpaint showing through from below. To avoid 

the danger of monotony in their emphasis on the brown foreground and 

green middle zones, the artists introduced a range of subtle variations within 

the harmonies of browns and greens. 

142 


Figure 3. Camelis Vroom, Landscape with a River by a Wood, signed and dated 1626. Panel, 31.3 X 44.2 em. National Gallery, London (3475). 

In Roelandt Savery's Landscape with Animals and Figures (1624), an area for 

the foreground group of figure and animals was held in reserve while the 

landscape was painted; only the small figures to the rear are painted over the 

landscape (Fig. 2). The foreground of this work is painted with final details 

in a typical transparent brown over a warm brown underpaint. In the lighter 

middle ground of this painting, the same transparent brown is used for the 

details over passages of tan underpaint; in a passage underpainted in gray, such 

as the stream at the left, the details are worked up in rapidly sketched strokes 

of darker gray paint. 

Landscape in Haarlem in the 1620s 

By the 1620s, a number of Dutch painters had taken up landscape as a specialty. 

In contrast to the fantasy landscapes of the Mannerists, these paintings 

represented recognizable scenes of the local countryside. The colors are limited, 

the compositions rely on subtle atmospheric effects to create almost 

continuous recession into space, and the figures no longer play a significant 

role. The techniques of landscape painting in this early period are varied, but 

there is evidence that those Haarlem artists who were most adventuresome 

compositionally also incorporated new ways of handling paint, methods 

which became integral to the effects of the "tonal landscape" painters in the 

1630s. 

Comelis Vroom painted the Landscape with a River by a Wood, a work now 

in London, in 1626 (Fig. 3) . This Haarlem landscape artist did not participate 

in the tonal style in the next decade, but in this early landscape, Vroom 

modified Flemish practice to create a more subtle spatial recession with his 

characteristic color range dominated by browns, yellows, and grayish greens. 

The color of each area is established using an underpaint, following the Flem- 

Giff ord 143

Figure 4. Esaias van de Velde, Winter Landscape, signed and dated 1623. Panel, 25. 9 X 30.4 em. The National Gallery, London (6269). 

ish practice, but this underpaint is more varied, not restricted to sharply 

defined bands of color receding into the composition. Throughout the foreground 

and middle ground the underpaint is a web of grayish green and 

yellow passages, worked wet-into-wet. 

The landscape paintings of Esaias van de Velde, produced during his years in 

Haarlem (1609-1618) and in the Hague until his death in 1630, are central 

to the development of the naturalistic landscape. Though Esaias was trained 

by the Flemish inunigrant Gillis van Coninxloo, he developed most of the 

elements of technique that were to define the work of the naturalistic tonal 

landscapists, who included his pupils Jan van Goyen and Pieter de Neyn. This 

is particularly apparent in his winter scenes, where the bare landscape encourages 

a particularly limited tonality. The Winter Landscape of 1623, now 

in London, is prepared with a thinly rubbed ground that barely fills the grain 

of the oak panel (Fig. 4). A cross section from the upper edge of the sky 

shows this extremely thin layer, textured on the underside by the grain of 

the wood (layer 1) (Plate 32). This preparation is barely perceptible on the 

painting; the warm pinkish tone that dominates passages, such as the distance 

at the right and parts of the fo reground, is created by the wood showing 

through the slightly tinted ground (12). In a loose and suggestive underdrawing, 

the artist situated the main elements of the landscape and indicated the 

foliage along the horizon with a few looped strokes but made no provision 

for the figures (Fig. 5). Following the guide of the underdrawing, he toned 

144 


Figure 5. Infrared photograph of Esaias van de Velde's Winter Landscape. National Gallery, London (6269). 

broad areas of the cOmpOSltlOn, leaving the panel's thinly applied ground 

visible in places. At this stage, the sky was painted with varying concentrations 

of pale smalt (layers 2 and 3), and the details of the horizon were painted 

wet-into-wet; fo rms in the foreground were sparsely indicated with tan and 

gray-green underpaint. Over this underpainting the artist sketched out the 

details of the image in a monochrome paint ranging from gray to black, then 

completed the painting with a few touches of creamy mid-tones and final 

highlights painted wet-into-wet. When the entire landscape was complete, 

the artist superimposed the small figures over it, the human presence almost 

incidental to the depiction of a raw winter's day. 

The tonal landscape painters 

At the end of the 1620s, Jan van Goyen and two other Haarlem artists, Pieter 

Molijn and Salomon van Ruysdael, initiated a landscape style in which a 

completely convincing representation of the Dutch countryside was rendered 

in a deliberately limited, almost monochromatic palette. A work as late as van 

Goyen's View oj Dordrecht Jrom the Dordtse Kil (1644) reveals the economy of 

handling that typifies these works: a wet-into-wet application of the sky and 

horizon, followed by a sketchy monochrome design only occasionally amplified 

by rapid touches of lightly colored paint (Fig. 6) . The tone of the 

wood and the grain pattern can be seen through the thin ground; both play 

a role in the fmal image. The results of an earlier technical study established 

Gifford 145

Figure 6. Jan van Coym, View of Dordrecht from the Dordtse Kil, sigrled and dated 1644. Panel, 64. 7 X 95. 9 eln. National Callery of Art, 

Washil1gton, D. C. (1978. 1.11). Alicia Mellon Bruce FUl1d. 

that both the visible wood grain and the monochromatic palette, especially 

the use of almost colorless smalt in the grayish skies, resulted from conscious 

artistic choices rather than accident (13). The rapidity with which these works 

must have been painted may have had an economic motive as well, lowering 

the cost of production and hence enlarging the market for such works (14). 

Conclusion 

The transition from Flemish mannerist landscape to a new Dutch form of 

naturalistic landscape in the first decades of the seventeenth century represents 

a dramatic transition in style. The mode of mythological or biblical episodes 

set before brightly colored, imaginary vistas, which Flemish immigrants had 

brought north, was replaced in Holland in just a few years by subdued and 

sympathetic renderings of the Dutch countryside. 

The preliminary results of this study of the painting techniques of the period 

suggest that the change in style was accomplished by deliberate changes in 

painting technique. Dutch artists modified the conventionalized Flemish landscape 

technique of light grounds, complete underdrawings, and colored underpaints, 

which established three schematic zones of space. Instead, working 

from drawings made from nature, they developed a sketchy shorthand style 

in which the properties of their painting materials were exploited to evocative 

effect. Thinly applied grounds and spare applications of underpaint allowed 

the panels themselves to play a role in the image. Underdrawings became 

quick notations upon which a more complete painted "sketch" was developed, 

and this monochrome image, only partially colored and worked up in 

the final stage of painting, was a dominant part of the finished work. 

These developments are a fascinating illustration of the ways in which political 

and social changes fo ster new artistic markets. As members of Dutch society 

146 


at the beginning of the seventeenth century, naturalistic landscape painters 

must have shared with their fellow citizens a newly awakened appreciation 

for the beauties of the local scene. As business people, they supplied and 

developed a new market governed by this taste. As practitioners of painting, 

they developed new working methods, which were both economical and 

superbly suited to expressing their new aesthetic. 

Notes 

1. Briels,J. G. C. A. 1976. De zuidnederlandse inmugrantie in Amsterdam en Haarlem 

omstreeks 1572-1630. PhD. diss. Rijksuniversiteit Utrecht. 

2. Bengtsson, A. 1952. Studies on the Rise of Realistic Landscape Painting in Holland 

1610-1625. Figure 3. 

3. Brown, C. 1986. Introduction. In Dutch Landscape, The Early Years: Haarlem and 

Amsterdam 1590- 1 650. Exhibition catalogue. London: The National Gallery, 11- 

34. 

4. Freedberg, D. 1980. Dutch Landscape Prints of the Seventeenth Century. London. 

5. I am grateful fo r the opportunity to have studied paintings in the collections of 

the Frans Halsmuseum, Haarlem; the National Gallery, London; the National 

Gallery of Art, Washington; and the Walters Art Gallery, Baltimore. I would 

particularly like to thank Ashok Roy of the National Gallery and Ella Hendriks 

of the Frans Halsmuseum for their assistance. I appreciate very much the many 

discussions of this material that I have had with Arthur K. Wheelock,Jr. Paintings 

included in the study were exanuned with a stereonLicroscope, and with X 

radiography and infrared reflectography, where available. Paint analysis was carried 

out on a linuted number of dispersed pigment samples and paint cross sections 

by light nLicroscopy and scanning electron nucroscopy with energy dispersive X 

ray spectroscopy (SEM/EDS). The compositions of paint mediums were estimated 

only using biological stains on the cross sections. Additional paintings were 

exanLined visually to study the handling of paint. 

6. A survey of several Dutch landscapes from the 1620s and the "tonal" period can 

be found in Bomford, D. Te chniques of the early Dutch landscape painters. In 

Brown, op. cit., 45-56. 

7. This painting was sampled by Feliciry Campbell during conservation treatment 

in 1988. 

8. This painting closely replicates the primary version now at the J. Paul Getry 

Museum, Malibu (92.PB82). 

9. From the samples, it is not clear whether the underdrawing lies under or over 

the imprimatura. 

10. Coninxioo died in 1607. Savery left Amsterdam for Prague, where he served the 

Emperor Rudolf II until about 1613, when he returned to Amsterdam. In 1619 

he moved to Utrecht, where he worked until his death in 1639. 

11. Historisches Museum der Pfalz, Speyer (H. M.1957/122). Illustrated in Sutton, 

p. C. 1987. Masters of Seventeenth-Century Dutch Landscape Painting. Exhibition 

catalogue. Amsterdam: RijksmuseuIT1; Boston: Museum of Fine Arts; Philadelphia 

Museum of Fine Arts. Plate 1. 

12. A ground of black, toned with a little yellow earth and rubbed into the poplar 

support, was observed on the original section of Farms Flanking a Frozen Canal, 

1614, North Carolina Museum of Art, Raleigh (52.9.61). See Goist, D. C. 1990. 

Case study of an early Dutch landscape. lCOM Committee fo r Conservation preprints, 

9th Triennial Meeting, 648-52. 

13. Gifford, E. M. 1983. A technical investigation of some Dutch seventeenth-century 

tonal landscapes. AlC preprints, 39-49. 

14. Montias,J. M. 1987. Cost and value in seventeenth-century Dutch art. Art History 

10:455-66. 

Gifford 147

Abstract 

Through extensive study of Vermeer's 

paintings, the author demonstrates 

that the artist must have used 

a chalk line attached to a pin at the 

vanishing point in the painting to 

create the central perspective in his 

pictures. By studying the changes in 

the design of the central perspective 

throughout his oeuvre, a certain 

chronology appears. This conclusion 

contradicts the previously accepted 

beliefs that Vermeer's interiors were 

faithful portraits of actual rooms or 

that the use of a camera obscura explained 

the realism of his interiors. 

Johannes Vermeer (1632-1675) and His Use 

of Perspective 

J orgen Wadum 

Mauritshuis 

Korte Vijverberg 8 

p. O. Box 536, 2501 eM Den Haag 


Introduction 

After visiting Vermeer on 21 June 1669, the art collector Pieter Teding van 

Berckhout noted in his diary that, among the examples of Vermeer's art he 

had seen, the most extraordinary and curious were those showing perspective 

(1). Three-dimensional interiors, depicted on two-dimensional canvases in 

such a way that the eye is deceived into believing the spatial illusion, were 

greatly admired by litjhebbers (connoisseurs) in the seventeenth century. The 

appreciation of perspective was underlined by the fact that these paintings 

had to be executed by artists who were sufficiently technically competent to 

be able to create these effects convincingly (2). Architectural pictures or "perspectives" 

were therefore often much more expensive than other genres. 

Montias documented that around 1650, the price fo r a perspective was fairly 

high, at an average of25.9 guilders apiece, while landscapes sold for an average 

of 5.6 guilders each (3). The Delft architecture painter Hendrik van Vliet (ca. 

161 1-1675) could have observed that one of his perspectives, in the estate of 

the art dealer Johannes de Renialme in Amsterdam at his death in 1657, was 

valued at 190 guilders (4). 

Montias states that despite the fact that interior scenes could also rightfully 

be called perspectives, he never came across a description of one in the many 

inventories he examined (5). All the more interesting then is the comment 

van Berckhout made after visiting Vermeer's atelier. 

When the inventory of Vermeer's estate was carried out after his premature 

death in 1675, a number of books in folio were found in his back room 

together with twenty-five other books of various kinds. Among the easels 

and canvases in his atelier, three bundles of all sorts of prints were found (6). 

It might be interesting to speculate what these books and prints were about, 

but we shall never know. It is conceivable, however, that some of them were 

guides to perspective drawing, works either by Hans Vredeman de Vries 

(1526 or 1527-1606) or those published by S. Marolois (ca. 1572-1627), 

Hendrik Hondius (1573-1649), and F. Desargues (1593-1662) (7, 8, 9, 10). It 

can be seen in the perspective design extrapolated from his paintings that 

Vermeer was certainly familiar with the principles laid out in these manuals 

on perspective. 

Unfortunately, nothing is known about Vermeer's apprenticeship. Therefore 

one must turn to an extensive examination of his paintings in order to gain 

an impression of the development of his method of rendering space. 

State of research 

Over the years many studies have been made of Vermeer's use of perspective 

and his spatial constructions, only a few of which shall be referred to here. 

Early this century, Eisler made an extensive study of Vermeer's use of space, 

and he describes the complicated use of triangles, circles, squares, and diagonals 

that may have fo rmed the basis fo r Vermeer's pictures (11). 

Probably inspired by Wilenski, who in 1928 wrote about some special effects 

in photography, " . .. perhaps one of the ironies of art history [is] that with 

a Kodak any child might now produce by accident a composition that a great 

148 


artist like Vermeer had to use all his ingenuity ... to achieve ... ," Swill ens 

published his first detailed study of the paintings in 1929. Swillens concluded 

that Vermeer had painted his oeuvre in five different rooms, all thoroughly 

recorded (12, 13). In another publication on Vermeer in 1950, he further 

elaborated his views on Vermeer's use of spatial illusion and his realistic recording 

of space (14). Swillens illustrated how Vermeer depicted his interiors 

with great accuracy. The position or eye level of the artist was established 

and thus the height of Vermeer himself and of the chair on which he almost 

always sat when painting. The work of Swillens, and belief that Vermeer 

rendered what he actually saw in front of him, has had a major influence on 

the scholarly research on Vermeer. 

Even in the 1940s, Hyatt Mayor records that highlights in the foreground in 

some of Vermeer's paintings "break up into dots like globules of halation 

swimming on a ground glass," and a few years later Gowing reached the 

conclusion that Vermeer had used a camera obscura (15, 16). 

In his studies, Seymour continues with this thought, which, apart from halation 

around highlights, he based on specific phenomena in the paintings, 

such as the diffusion of the contours (17). He also found that the perspective 

in certain paintings resembled the distortion obtained when using a wideangle 

lens. 

Prompted by Seymour's article, Schwarz fu rther suggested that Vermeer may 

have used the camera obscura as a technical aid in his painting process (18). 

Bearing testimony to Vermeer's use of technical devices for rendering his 

images, wrote Schwarz, is the fact that the mathematician and physicist Balthasar 

de Monconys (1611-1665) attempted to visit Vermeer during his stay 

in Delft in 1663, and that a friendship probably existed between Vermeer and 

fellow townsman Anthony van Leeuwenhoek (1632-1723), a specialist in 

microscopes and lenses. 

There is a tendency to consider mathematicians otherwise uninvolved in the 

creation of visual arts as responsible for developing an intellectual interest in 

perspective. However, this view was probably not shared by the artists, who, 

fo r their part, were using the simplest and at the same time most convincing 

methods to create their spatial illusions. We are therefore entitled to believe 

that, as de Monconys was also an art collector, it would be much more likely 

that he wanted to pay Vermeer a visit in order to see his renowned paintings 

or perspectives (19). Vermeer was also the Headman of the Guild of St. Luke 

at the time, and he would naturally be the person for an art collector to see 

when visiting the town. 

In 1968 Mocquot suggested that Vermeer might have used double mirrors 

to create his perspectives, both in his Allegory of Painting and in The Concert 

(20, 21, 22). 

Finck claimed in 1971 to be able to prove that twenty-seven paintings by 

Vermeer must have been made with the aid of a camera obscura (23). The 

arguments presented here will make it clear that this highly ambitious thesis 

has no basis in reality. 

Wheelock undertook the most detailed study of the optics and perspectives 

used by Delft painters around 1650 (24). It is argued that some of Vermeer's 

pictures (although far fewer than is asserted by Finck) do indeed have many 

effects similar to that which can be achieved using lenses or a camera obscura, 

and therefore the use of these devices seems highly probable. Wheelock does 

make clear, however, that the use of a camera obscura would be extremely 

difficult indoors because the light levels were generally insufficient to obtain 

an image. In more recent publications, Wheelock increasingly argues that 

Vermeer probably did not trace images seen through a camera obscura, but 

that he must have been aware of the device and used certain special effects 

seen through it for his paintings. 

Wadum 149

Based on intensive studies of various means used to create " church portraits," 

de Boer concluded in 1988 that optical devices were not generally used by 

artists in the Netherlands around 1650 (25). Interestingly, he notes that the 

reason for this would probably be the difficulty of combining the use of a 

camera obscura (for tracing an image) with actually painting a painting. 

Recently, Arasse made a general comparison of the position of the horizon 

and the viewpoint in Vermeer paintings (26). He notes that the viewpoint 

gradually lowered between 1656 and 1661. According to Arasse, Vermeer 

tended to combine a low viewpoint with a high horizon. Arasse considers 

the often very low viewpoint in relation to the depicted figures to be a special 

effect that Vermeer deliberately wanted to create in order to draw the viewer 

into the scenes. This statement shows that Arasse considers Vermeer's intention 

to be the creation of an illusionistic spatial setting as an imaginative 

process rather than the rendering of a known space, an opinion this author 

shares. 

Present research 

In the following paragraphs, results from the author's latest research on this 

aspect of Vermeer's painting technique are presented. Through a thorough 

study of the actual paintings, mostly out of their frames and placed under a 

stereomicroscope, certain surface phenomena in the paint layer have been 

observed. Together with X-radiographs and other photoanalytical means such 

as ultraviolet and infrared photography, a compilation of information has been 

possible, leading to the conclusion that Vermeer did not paint "naar het leven" 

(after life), as suggested by the majority of scholars mentioned above, but that 

as a craftsman he created a spatial illusion with the masterly hand of an 

outstanding artist. 

In 1949 Hulten was the first to actually record a discernible vanishing point 

in one of Vermeer's paintings. He observed that just below the left knob of 

the map hanging on the rear wall in The Allegory of Painting there was a small 

irregularity in the paint layer which coincided with the central vanishing 

point of the composition (27). 

Indeed, fo r Vermeer the central perspective was the main guideline for his 

interiors. Current examinations reveal that the vanishing point can still be 

fo und in most of his interior scenes (28). It can be seen (with the naked eye 

or more easily with a stereomicroscope) that Vermeer must have attached a 

pin at the vanishing point in the painting, resulting in the loss of minuscule 

amounts of paint and ground. X-radiographs can be used to find the black 

spot where the ground containing lead white is missing between the threads 

of the canvas. Having inserted the pin at the vanishing point, Vermeer would 

have used a string to reach any area of his canvas to create perfect orthogonals 

for the perspective. 

Vermeer's method is far from unique; among the architectural painters of his 

time it was well known. Gererd Houckgeest (1600-1661) and Emanuel de 

Witte (1617-1 692) practiced this method, which Pieter Saenredam (1597- 

1665) had already brought to perfection (29, 30). Vermeer's slightly older 

colleague Pieter de Hooch (1629-1683) also used a single vanishing point. 

Similarly, in paintings by Gerard Dou (1613-1675), Gabriel Metsu (ca. 1629- 

1667), and others, we again find irregularities in the paint where a pin was 

placed at the vanishing point. 

The method of using a chalk line to indicate lines is still used by painters 

and other artists when planning illusionistic interiors (e.g., with marbling, a 

specialty developed during the Baroque period) . That this kind of illusionistic 

painting was known to Vermeer is clear from the virginals in the two London 

paintings, both of which have been "marbled." 

In order to transfer the line indicated by the string, chalk is applied to the 

string. Holding the string taut from the pin inserted at the vanishing point, 

150 


Figure 1. Perspective drawing of Vermeer's Glass of Wine, 1658-1660. 

the painter draws the string up a little from the surface, using the free hand, 

and lets it snap back onto the canvas. The powdery chalk is thus applied to 

the surface of the painting; the line can be traced with a pencil or brush. The 

remaining dust can be gently wiped or blown away, leaving little or no trace 

of the method except the pin point. 

The distance points, positioned on either side of the vanishing point on the 

horizon, provide the basis fo r the diagonals, which in turn form the basis of 

the tiled floors. Distortion occurs at the corners if the horizon is placed 

relatively high and the distance points are close to the vanishing point. Examples 

of this are observed in the Glass of Wine (1658-1660), with its viewing 

angle of approximately 43° (Fig. 1), and the Girl with The Wine Glass (1659- 

1660), which has an even smaller angle of approximately 35° (31, 32). The 

Music Lesson (1662-1665), in which Vermeer again returns to a large angle, 

(approximately 44°), is the last picture in his oeuvre to show a certain distortion 

of the floor tiles due to the short interval from the distance points 

(33). In The Concert (1665-1666), the angle again returns to about 34°; in The 

Allegory of Painting (1666-1667), the viewing angle has decreased to around 

30° (Fig. 2). In The Love Letter (1667-1670) the angle declines to about 28°, 

and in the last painting executed by his hand, A Lady Seated at the Virginals 

(1673-1675), Vermeer reduces the viewing angle to only 22° (Fig. 3) (34, 

35). 

A growing tendency can clearly be observed over the years to let the distance 

points move further away from the scene. By doing this, Vermeer eliminated 

the distortion of the floor tiles in the foreground corners, particularly as he 

moved his vanishing point toward the edge of the painting at the same time. 

As the vanishing point can still be identified in many of Vermeer's pictures, 

his method of using threads attached to a pin inserted at the central point is 

evident. The distance points, however, could constitute a problem. Would he 

be able to determine the position of the diagonals on the edge of his canvas 

when space recedes towards the back walls in his interiors? This is hardly 

Figure 2. Perspective drawing of Vermeer's Allegory of Painting, 1666-1 667. 

Wadum 151

Figure 3. Perspective drawing of Vermeer's A Lady Seated at the Virginals, 1673- 1675. 

likely because it would imply doing unnecessary calculations, and indeed no 

trace of marks on the edges of his paintings has so far surfaced. 

If there were a simple method of creating perfect central-point perspective, 

painters would surely have used it. By placing the canvas against a board (most 

of his paintings are small) or a wall, between two nails on either side of the 

painting, the painter would be able to use strings for the diagonals as well. 

Indications of the use of such a simple method may be deduced from books 

on perspective that might have been known to Vermeer. Desargues writes in 

his introduction that a painter who wants to know more about the Meetkonst 

(the art of measurement) should consult the Landmeeter (the cartographer) 

in order to make use of his expertise (36) . This, he writes, would lead 

to a better understanding of perspective or Doorsicht-kunde. Desargues further 

suggests that the painter should look around him in other guilds to take 

advantage of the knowledge of carpenters, bricklayers, and cabinetmakers. 

Furthermore, it appears that constructors of perspective in the seventeenth 

century were using drawing tables almost as sophisticated as the ones we use 

today. With strings attached to the upper corners of the drawing table, the 

draftsman could create any orthogonals he wanted on his paper. The horizon 

could be plotted using a sliding ruler at a fixed 90° angle to the horizontal 

bottom edge of the table. A horizon would be chosen at the desired level on 

this ruler, and by sliding the ruler across the paper, a line could be drawn 

(37). 

Vermeer also worked in this way, as is proven by the presence of the clearly 

distinguishable needle point found in the paint in paintings throughout his 

whole oeuvre (38). 

Conclusion 

The extraordinary and curious perspectives, so much admired by van Berckhout 

in 1669, therefore appear to have been carefully constructed. This leaves 

the impression that Vermeer should be regarded first and foremost as a practical 

and skilled master in creating space just the way he wanted it. This 

approach departs from the previous conception of the artist as reproducing 

the scenes he saw in front of him, either by careful copying using drawing 

frames or a camera obscura. The author believes that Vermeer was completely 

aware of the spatial illusion he wanted to create, which he produced by 

combining his skill in constructing space with his artistic talent fo r composition, 

color, technique, and iconography (39). He thereby created his images 

in such a way that viewers are deceived into believing that the scenes were 

real. This was the highest level of artistic ambition to which a seventeenthcentury 

painter could aspire, a level Vermeer surely attained. 


Most sincere thanks are due to the institutes that have been more than generous in 

supplying information about their Vermeer paintings and letting the author examine 

most of them in the conservation studios. The amount of information is overwhelming 

and goes well beyond the scope of this article; the reader is referred to the 

fo rthcoming exhibition catalogue on Johannes Vermeer (Washington, 1995; The 

152 


Hague, 1996). The author is also grateful to Rob Ruurs of the University of Amsterdam 

for a fruitful and stimulating discussion. 

Notes 

1. Van Berckhout may have been able to see the two town views, The Little Street 

and The View of Delft, and certainly interior scenes such as The Music Lesson, A 

Woman Holding a Balance, The Concert, and The Allegory of Painting. He might 

have also seen The Love Letter, which is dated 1667 (Blankert) or 1669-1670 

(Wheelock). 

2. Ruurs, R. 1991. Functies van architectuurschildering, in het bijzonder het kerkinterieur. 

Perspectieven: Saenredam en de architectuurschilders van de 17e eeuw: Boymans-van 

Beuningen Rotterdam, 50. 

3. Montias, J. M. 1991. "Perspectieven" in zeventiende-eeuwse boedelbeschrijvingen. 

Perspectiven: Saenredam en de architectuurschilders van de 17e eeuw: Boymansvan 


4. Bredius, A. 1915-1922. Kunstler-Inventare, Urkunden zur Cesichte der Hollandischen 

Kunst des XVIten, XVIIten und XVIIItenjahrhunderts 1-7. Den Haag.Vol. 1:238. 

5. Montias, op. cit., 24. 

6. Montias, J. M. 1989. Vermeer and his Milieu: A Web of Social History. Princeton, 

339 (doc. 364). 

7. Vredeman de Vries, H. 1604. Perspective. Dat is de hoochgheroemde cOl'lste . . . . Den 

Haag. 

8. Marolois, S. 1628. Perspective contenant la tMorie, pratique et instructionfondamentale 

d'icelle. Amsterdam. 

9. Hondius, H. 1647. Crondige onderrichten in de optica of de perspective konste. Amsterdam. 

10. Bosse, A. 1664. Franfois Desargues. Algemene manier tot de practyck der perspective 

gelyck tot die der meet-kunde met de kleyn voet-maat [ ... J by-een-gevoeght door. Amsterdam. 

11. Eisler, M. 1916. Der Raum bei Jan Vermeer. jahrbuch der Kunsthistorischen Sammlungen 

der Allerhachsten Kaiserhauses. Wi en-Leipzig: Band XXXIII (4):213-91. 

12. Wilenski, R. H. 1928. An introduction to Dutch Art. New York, 284-86. 

13. Swillens, P. T. A. 1929. Een perspectivische studie over de schilderijen van Johannes 

Vermeer van Delft. Oude Kunst (7): 129-61. 

14. Swillens, P. T. A. 1950.joharmes Vermeer, Painter of Delft 1632-1675. Utrecht 

Brussels. 

15. Hyatt Mayor, A. 1946. The photographic eye. The Metropolitan Museum of Art 

Bulletin (5):15-26. 

16. Gowing, L. 1953. Vermeer. 

17. Seymour Jr., C. 1964. Dark chamber and light-filled room: Vermeer and the 

camera obscura. The Art Bulletin XLVI (3):323-31. 

18. Schwarz, H. 1966. Vermeer and the camera obscura. Pantheon (XXIV): 170-82. 

19. De Monconys visited numerous artists during his travels all over Europe to meet 

with fellow scientists. 

20. Mocquot, M. 1968. Vermeer et Ie portrait en double mirror. Le club franfaise de 

la medaille 18 (1):58-69. 

21. Kunsthistorisches Museum, Vienna. 

22. Isabella Steward Gardner Museum, Boston. 

23. Finck, D. 1971. Vermeer's use of the camera obscura: a comparative study. The 

Art Bulletin. (LIII):493-505. 

24. Wheelock Jr., A. K. 1977. Perspective, Optics, and Delft Artists Around 1650. New 

York. 

25. de Boer, P. G. 1988. Enkele Delftse zeventiende eeuwse kerkportretten opnieuw bekekert. 

Dissertatie: Delft. 

26. Arasse, D. 1994. The Art of Painting. Paris. 

27. Hulten, K. G. 1949. Zu Vermeer's Atelierbild. Konsthistorisk Tidskrift (XVIII):90- 

98. 

28. The first pin hole we discovered in the vanishing point of a Vermeer painting 

was in the Lady Writing a Letter with her Maid from the National Gallery of 

Ireland. We are grateful to Andrew O'Connor, chief conservator of the gallery, 

fo r letting us examine the painting just after its recent rediscovery. 

29. Ruurs, R. 1987. Saertredam: The Art of Perspective. Amsterdam, Philadelphia, Gronmgen. 

30. Giltaij, J. 1991. Perspectiven, Saenredam en de architectuurschilders van de 17e 

eeuw. Perspectiven: Saemedam en de architectuurschilders van de 17e eeuw: Boymansvan 


Wadum 153

31. Herzog Anton Ulrick-Museum, Braunschweig. 

32. Staatliche Museen zu Berlin, Preul3ischer Kulturbesitz, Gemaldegalerie, Berlin- 

Dahlem. 

33. H. M. Queen Elizabeth II, Royal Collection, London. 

34. Rijksmuseum, Amsterdam. 

35. The National Gallery, London. 

36. Bosse, op. cit. 

37. Ibid. 

38. From Offi cer and Laughing Girl, ca. 1658 (The Frick Collection, New York) , to 

A Young Woman Standing at the Virginal, ca. 1673-1675 (The National Gallery, 

London). 

39. See fo rthcoming exhibition catalogue on Johannes Vermeer (Washington, 1995; 

The Hague, 1996), in which various aspects of Vermeer's paintings are examined. 

154 


Abstract 

The paper presents some results of 

research on the painting materials 

and methods used in Latvian 

churches in the seventeenth century. 

The technical research of historical 

painting techniques in Latvia is rather 

preliminary. The authors concentrate 

on the polychromy of the interior 

decorations. Results of the 

analysis of materials and techniques 

used in the decorations are given, 

stressing the varieties as a result of 

the various interpretations of the Baroque 

style and technique in different 

regions of Latvia. 

A Technical Study of the Materials and Methods 

Used by the Painters of the Latvian Churches 

in the Seventeenth Century 

Ilze Poriete* 

Restoration Centre History 

Museum of Latvia 

Vecpilsetas 7 

Riga, LV 1050 

Latvia 

Dace Choldere 

State Culture Monuments Department of Latvia 

Klostera 5 

Riga, LV 1050 

Latvia 

Introduction 

In the second half of the seventeenth century, the territory of Latvia was 

ruled by several different monarchs and was therefore divided into areas of 

various influences. From the middle of the sixteenth century, the western part 

was ruled by the Duke ofKurzeme an Zemgale; after the Polish-Swedish war 

the southeastern part (Latgale) was subject to the Polish king and the northeastern 

part (Vidzeme) to the Swedish king. These conditions influenced the 

production of Latvian art. Created by artists coming from various parts of 

Europe, general styles and tendencies were interpreted in various ways (1). 

Therefore, the objects of this survey-the churches and their pulpits in Nurmuizha 

and Burtnieki, both decorated in the 1680s and representative of the 

Baroque-illustrate two different styles using very different techniques and 

materials. 

Figure 1. The pulpit and the altar of the 

Nurmuizha church bifore restoration. Photograph 

by M. B. Vanaga and R. Kanins. 

Figure 2. The pulpit of the Burtnieki church 

bifore restoration. Photograph by M. B. 

Vanaga and R. Kanins. 

The church of Nurmuizha is located in the territory of Kurzeme and was 

founded in 1594; however, the present interior dates from the 1680s. Twisted 

columns with rich decorative carvings and many sculptures portraying disciples 

of Christ decorate the pulpit and the church interior. A sculpture of 

Moses functions as support for the pulpit. The artist who executed the wood 

carving is unknown but his style indicates an eastern Prussian origin. Although 

written contemporary sources do not shed any light on when the 

decorations were painted, one must assume that it was carried out in the 

1680s. Later repairs are relatively extensive, including additional wood carvings 

and paintings executed by different craftsmen, such as gilders and interior 

painters (Fig. 1). 

The church of Burtnieki is located in the northeastern part of Latvia which 

was under the rule of the Swedish king. Here, the basic construction of the 

pulpit is very similar to construction of the pulpit in Nurmuizha, but the 

decoration is more restricted, rational, and even classical. Baroque influences 

arrived here by many different routes, thus resulting in a large variety of 

interpretations. In both churches, pulpits are situated under the arch of triumph. 

In Burtnieki, however, the pulpit's construction is polygonal with 

straight stairs, and the columns are not twisted but straight and narrow. The 

sides of the stairs are decorated with panels separated by pilasters instead of 

columns. Instead of sculptures, there are paintings on the panels between the 

columns, all representing scenes from the New Testament. In contemporary 

documents (1691), it is noted that the altar and the pulpit were painted and 

outlined with silver. During the present restoration, a rather neutral overpainting 

was removed, revealing the original decorative painting, which shows 


Poriete and Choldere 155

marbled columns, multicolored paintings on the panels, and blue and silver 

paintings on the podium. The identities of the artists are unknown (Fig. 2). 

Methods 

Analysis of the pigments was carried out combining optical microscopy, microchemical 

tests, and emission spectroanalysis (2, 3). Media were determined 

by the use of thin-layer chromatography, infrared spectroscopy, and microchemical 

tests (4, 5, 6) . 

Materials and results 

Technical analysis from paint samples was executed to reveal the original layer 

of the pulpit of the Nurmuizha church. Samples were taken from the background 

and profiles of the pulpit's twisted columns (now black) and from the 

decorative wood carvings of grapes, masks, and reliefs (now gold) . The results 

indicate that the ground layer consists of an unpigmented calcium carbonate 

bound with an animal glue, as was shown by staining tests. In the cross section, 

the ground layer is shown as a white layer with some small brown glue 

particles. On a thin transparent layer in which protein has been fo und, there 

is a black layer with occasional particles of a blue pigment. The few blue 

particles present are transparent; a positive identification of the pigment was 

impossible. Later overpaintings are executed in black, containing charcoal and 

oil. The decorative vines are gilded. The underlayer is composed of calcium 

carbonate and glue. In the cross section, a layer of brown hematite is visible 

and the presence of glue particles was determined with staining tests. On top 

of the gilding there is a layer of mordant gilding: oil pigmented with ochre 

and minium, the latter probably acting as a drying agent. The bronze layer 

on the pigmented oil layer was applied much later. 

Figure 3. Cross section if paint sample from 

part if the flesh color on a sculpture in the 

Nurmuizha church. The layers are as follows: 

(1) thin pink ground, (2) main paint layer 

containing lead white and ochre, (3) transparent 

layer containing oil, (4-8) overpaints. 

Photograph by M. B. Vanaga and R. Kanins. 

Samples were taken from the clothes, hair, and flesh color of one sculpture 

on the pulpit (Fig. 3) . The analyses showed that the gilded wrap was executed 

in a water-gilding technique, while the dress itself was originally blue. The 

blue layer consists oflead white and smalt with tempera as a binding medium 

on a chalk-glue ground layer (7). Analysis of the samples taken suggests that 

the pulpit had water-gilded, wood-carving details on a blue background. The 

same blue color is also fo und in other details of the church's decoration (the 

altar and the decorative ledge). Later overpaintings, however, have penetrated 

the original, damaged layer, changing it and making a correct analysis difficult. 

During the technical investigation of the pulpit of Burtnieki church, samples 

were taken from parts that were well conserved: the background of the pulpit's 

construction, the podium's decorative ledge, the decorative grapes, the roof, 

and the pelican. The results of the analysis show that the pink background 

contains hematite and occasional particles of calcium carbonate (CaC03) in 

oil (Fig. 4). Colors from the original marbling are revealed in the background 

(smalt, indigo, hematite, charcoal black, and copper resinate). Decorative elements 

are silvered, such as the silver leaves, which were glazed with copper 

resinate over the pink underlayer (hematite, oil). 

Figure 4. Cross section if paint sample from 

part if the podium in the Burtnieki church. 

The layers are as follows: (1) thin pink 

ground, (2) layer of silver leaves, (3) green 

layer containing cupric resinate, (4-8) overpaints. 

Photograph by M. B. Vanaga and 

R. Kanins. 

The decorative bunch of grapes is painted in a dark blue layer containing 

indigo, smalt, and oil, and the green leaves are executed in a green glaze 

(copper resinate, oil, gum) over silver leaf applied in an oil-based mordant 

technique. The decorative details of the pulpit's baldachin have the same 

preparatory ground layer as the podium. The marbling is carried out as described 

above. The silvering is done using an oil mordant and subsequently 

glazed. In some places the silver is covered with colorful glazes (i.e., the drops 

of blood on the pelican are executed in an organic red glaze on silver). In 

summary, the decorative painting of the Burtnieki church shows a rich polychromy 

and is executed in an oil medium. 

156 


Summary 

We can conclude after this preliminary investigation that many differences 

between the two pulpits are present, although they both date from the same 

period. Two types of grounds were found; in the Nurmuizha pulpit, chalkglue 

ground is used, while in the Burtnieki pulpit a ground of ochre in oil 

is present. The pigments are similar in that they are probably the most commonly 

used pigments in Latvia during that period. During previous investigations 

of blue colors in the polychromy of seventeenth-century Latvian 

churches, two pigments were always present: indigo and smalt. Natural ultramarine 

has been found in very few cases. 

Details in the decoration of the Nurmuizha church are executed using a 

water-gilding technique. It is impossible to determine if a varnish or glaze is 

present, as the next gilding layer is applied in an oil-based mordant technique 

that fully blends with the first gilded layer. 

The decorative wood carvings of the Burtnieki church are silvered and in 

some places covered with colored glazes. 

Different hands are present in both churches, as during the seventeenth century 

in Latvia there was a continuous migration of craftspersons. A gilderpainter 

could have been called from any workshop, especially in rural areas, 

introducing that workshop's typical techniques and materials. It is not suggested 

that the described methods were the most common in Latvia. Any 

concrete conclusions can only be made after fu rther investigations in Latvia, 

and after a comparison with methods used in other parts in Europe. 


The authors are grateful to polychrome wooden sculpture restorers S. Rugevica and 

1. Pukse fo r the samples and for their help with the work. Thanks are due to M. 

Harris for editing the manuscript. 

Notes 

1. Grosmane, E. 1981. Ventspils koktelnieki 17.gs otra puse- 18.gs.sakums. Riga, Zinatne, 

46-61. 

2. Microscope MBI-15 (USSR), magnification x160. 

3. Emission spectroscopy was carried out by A. Deme, an assistant of the spectroscopy 

laboratory of Latvia University. 

4. Striegel, M. E, and D. Stulik. 1993. High performance thin-layer chromatography 

fo r the identification of binding media: techniques and applications. In AIC Preprints. 

5. Infrared spectroscopy was carried out by 1. Gudele, an assistant at Riga Technical 

University. 

6. Martin, E. 1977. Some improvements in techniques of analysis of paint media. 

Studies in Conservation 22:63-67. 

7. The definition is in process. 

Poriete and Choldere 157

Abstract 

The shift in the nineteenth century 

from a tradition of artist-prepared 

materials to an industry of mass-produced 

commercial products greatly 

endangered the artistic community 

through the widespread distribution 

of products of inferior quality and 

unstable properties. For over fifty 

years, the British Pre-Raphaelite 

painter William Holman Hunt 

waged a campaign fo r the reform of 

the manufacture of artists' materials 

and the rights of the artist as a consumer 

to expect materials of consistent 

quality and uniformity. The reevaluation 

of the Pre-Raphaelite 

technique, in conjunction with an 

exploration of Hunt's advocacy on 

behalf of artist-consumers, places in 

perspective his focus on artistic 

traditions at a crucial transition time 

in the history of materials and tech 

TUques. 

William Holman Hunt and the 

"Pre-Raphaelite Technique" 

Melissa R. Katz 

Davis Museum and Cultural Center 

Wellesley ColJege 

Wellesley, Massachusetts 02181 

USA 

Introduction 

The late-eighteenth and early nineteenth centuries in British painting constitute 

a period noted for the rise of a major school of national painting, 

dominated by masters such as Sir Joshua Reynolds, Thomas Gainsborough, 

and J. M. W. Turner, who engaged in technical experiments of dubious value. 

Their reliance on gelled mediums, bituminous paints, and fugitive pigments, 

respectively, has left a body of work disfigured by sunken patches, wide craquelure, 

and faded color. In search of shortcuts to achieve the luminous glow 

of the old masters, they produced, instead, paintings whose technical inadequacies 

were known and seen by the generation that followed. 

This next generation of painters included the Pre-Raphaelite Brotherhood, 

formed in 1848 in radical opposition to the training and taste imposed on 

British art by the Royal Academy. Hunt was a fo unding brother, and the only 

one of the group to maintain a lifelong adherence to their principles of 

fidelity to nature minutely observed, boldness in color and lighting, emulation 

of early Italian painting, and depiction of contemporary or literary subject 

matter. Yet the rebellion was short-lived, and the Pre-Raphaelites rapidly 

became the leading painters of their day, with Hunt as one of the most 

popular. 

By the 1870s Hunt's position was assured as the celebrated painter of such 

Victorian icons as The Light oj the World, The Awakening Conscience, and The 

Finding if the Saviour in the Temple (Figs. 1,2, 3). Periods spent in the Middle 

East seeking Biblical authenticity alternated with spells in pleasant, wellequipped 

London studios where, liberated from hand-to-mouth struggle, he 

was free to contemplate other aspects of art as a career (Figs. 4, 5). The stability 

and longevity of his paintings were of primary concern to Hunt, who observed 

not only the technical inadequacies of the preceding generation, but 

also the poor aging qualities of the artworks of his contemporaries. 

Nineteenth-century commercial artists' materials 

Hunt was among the first artists to note the increasingly poor quality of the 

artists' materials fo r sale in the mid-nineteenth century, and the most vociferous 

in calling attention to their faults and advocating their improvement. 

From the platform of leading painter of his day, he set out on a crusade to 

reform the manufacture of artists' materials, to impose standards of quality 

and workmanship, and to ensure access to consistently reliable products from 

colormen informed of and interested in the durability of the goods they were 

selling. 

Figure 1. William Holman Hunt in his studio, 

painting the St. Paul's versioll q{The 

Light of the World, ca. 1900. 

The market fo r artists' materials in the early and mid-nineteenth century had 

changed overwhelmingly with the advent of industrial production and mass 

marketing, and the rapid introduction and adaptation of newly developed 

materials whose aging properties were unknown. Prior to the nineteenth 

century, artists had used materials prepared fo r them in their own studios or 

by local artisans who followed the exacting standards of their clients, allowing 

the artist to determine the materials used and methods of preparation. With 

the rise of the commercial colorman in an age of burgeoning capitalism, a 

middleman was introduced between the manufacturer and the consumer of 

158 


the product, a salesman motivated by the need for profit and immediate shortterm 

customer satisfaction rather than long-term stability. 

The evolution of the art nurket also brought with it an expansion of materials 

available fo r sale, aided by advances in scientific fields that greatly expanded 

the range of available pigments and dyestuffs. The artist's palette roughly doubled 

in the nineteenth century, each decade bringing new colors, starting 

with cobalt blue and lemon yellow. In the early part of the century, iodine 

scarlet, chrome yellow, and emerald green were introduced; and, in the 1820s, 

synthetic ultramarine became available. Zinc oxide (Chinese white) and viridian 

came in the 1830s, and cadmium yellow and orange followed in the 

1840s. The 1850s brought the aniline dyes, brilliantly colored and wildly 

impermanent coal-tar derivatives. The 1860s saw the dawn of aureolin yellow, 

chromium oxide green, synthetic alizarin, manganese violet, cerulean blue, 

and so on (1). The colors produced by these new pigments were often dazzling, 

but also alarmingly unstable. 

Figure 2. William Holman Hunt sculpting a 

model of Christ's head to aid him in the 

painting

with colormen and hired scientists to analyze paint samples from various 

suppliers. He badgered the Royal Academy into appointing their first professor 

of chemistry to research and teach materials science. He fo rmed an 

artists' cooperative to secure hand-ground pigments and pure materials. He 

monitored the condition of his own paintings, attentive to the conditions of 

their display and handling. He made test panels and stored them in his studio 

ten, fifteen, and twenty years to observe the effects of aging. With ever increasing 

obsession, he investigated material permanence, compatibility, and 

composition. Like a true conservator, he fo rsook the artwork for the artmaking, 

producing fewer paintings over longer intervals in a painstakingly slow 

technique. 

Figure 5. William Holman Hunt, Self-Portrait, 

1875. Courtesy of the Ujfi zi Gallery, 

Florence, Italy. Note the length of the brushes 

on the table. The tunic is the same striped 

garment used in The Finding of the Saviour 

in the Temple. 

Figure 6. Nineteenth-century containers for 

oil paints (animal bladder with ivory tacks, 

piston tuhe, and collapsible metal tube). 

Courtesy of the Forbes Collection, Harvard 

University Art Museums. 

Hunt's passions flared in the mid-1870s with the realization that Roberson's 

orange vermilion, a favorite commercial tube paint, was being adulterated 

with red lead, and thus blackening on the canvas. In frustration, Hunt wrote 

to his friend and fellow artist, John Lucas Tupper (5): 

It seems as tho [sic] I were struggling against Fate. Every day sometimes 

including Sundays I have been toiling every hour, and just as I have got 

my task nearly completed the whole thing has fallen into disorder again for 

at least five or six times and I have had to begin again. At last I have 

fo und out what has been the cause oj this: Roberson's tube oj Orange 

Vermilion, which I used without suspicion because 25 years ago they sold 

this color absolutely pure, is adulterated with 10 percent oj villainy, the 

greater part lead, which has blackened so rapidly that when it had got dry 

enough fo r the f 

inal glazings the flesh had got to such a color [sic] that I 

nearly went crazy ... I have had the color analysed and at the same time 

have taken the opportunity to have others investigated and find that the 

fraudulent habit is exercised in many other cases. What is to me more 

discouraging than this is that many artists I have spoken to about [it] are 

quite satisfied to go on dealing in these spurious colors saying "Oh, they 

will last my time, " and "I never fo und my pictures change" and with 

base humility "they, " the colors, "are good enough for my work. " Leighton, 

when I proposed a little co-operative society for importing and grinding 

pure colors said, "And what's poor Roberson to do?" 

The culmination came on Friday, 23 April 1880, when Hunt addressed the 

members of the Royal Society of Arts on the subject, "The Present System 

of Obtaining Materials in Use by Artist Painters, as Compared with that of 

the Old Masters." At a conservative estimate, the talk lasted at least two-anda-half 

hours. Concerned as much with the decline in knowledge of artists' 

techniques as well as materials, Hunt observed, "In the old days the secrets 

were the artist's; now he is the first to be kept in ignorance of what he IS 

using" (6). Eloquently he informed his audience (7): 

I feel called upon to avow that I regard the artists' colourmen oj London 

as gentlemen of intelligence, of character, and great enterprise, to which 

qualities we are much indebted for the comparatively safe positions we enjoy; 

for indeed, at the worst, it must be recognised that we might have gone 

fu rther astray. It is needful, however, that we should be not only in good 

hands, but we should give strong prooJ that we can distinguish between 

that which is fa ulty and that which is perfect; and it is the want of discriminating 

power in the painter which produces all the indiff erence on the 

part oj the preparer to the permanent character of the materials he supplies. 

The painter has really not the power to trace the causes of difects. The 

colourman naturally judges of the character of the materials he vends by the 

condition they are in while under his own eye. To him, the evils revealing 

themselves in the work which has passed through his shop do not exist if 

he never sees them; and if he hears of them only, as evils untraceable in 

their cause which have occurred to one of his customers (who may, sometimes, 

have obtained materials elsewhere), his sense oj responsibility is quieted, 

when he has received the assurance of his men in the workshop that 

the usual rules, which have hitherto resulted in work of a kind not eliciting 

160 


complaints, have been strictly adhered to. The workmen, too, in these shops 

are not permanent, and there is virtually no responsibility fo r any one 

preparation. In most cases the complaint is never made, for the evil may 

be a very serious one, and yet it may not manifest itself bifore the death 

of the artist. 

In conclusion, he stated, "The cure that we have to seek is one it is possible 

to defme compactly. It is to establish a means of transmitting the practical 

wisdom of one generation of painters to another" (8). 

With striking prescience, he addressed the need "to fo und a society for looking 

after the material interests of painting . .. composed of important members 

of the profession of painting ... [joined by] gentlemen of reputation in 

chemical science." This society would form "a library ... of all works of 

literature which exist on the subject of artistic practices . .. ,establish a workshop 

for the preparation of materials . .. ,arrange for the importing of colours 

from abroad for collecting specimens of experiments. An important aspect of 

the society will be to cultivate the opportunities of obtaining further samples 

of every variety of colours existing in the far East, of proving these, and 

putting them on record in our museum fo r all generations to see." A technical 

school would be established, and artists trained, "that thus we should be the 

inheritors, not only of our immediate predecessors, but the heirs of all the 

ages, and that, though our pretensions would not be ostentatious in our humble 

way, we might be proud that we should be repeating the chosen tasks of 

the gods, the directing of inert matter to a spiritual end" (9). 

The Pre-Raphaelite technique 

Fully aware of the frailty of the materials of painting, Hunt became a master 

technician. Most of his paintings have remained well preserved, retaining the 

exceptional brilliancy of color which he strove so hard to achieve via successive 

layering of minute strokes of transparent and semi-transparent paint. 

He chose quality linen canvas stretched on paneled stretcher supports as "protection 

against accidental injuries, such as a push from a corner of a picture 

frame in the confusion which precedes and follows exhibitions, a kind of 

injury which, if not visible at the time, may show years afterwards in starred 

cracks in the hard paint" (10). 

With some notable exceptions (such as the Liverpool Triumph of the Innocents, 

painted on handkerchief linen from an Arab bazaar because the artist was too 

impatient to await a delayed shipment of supplies from England), many of 

Hunt's paintings remain unlined today and on their original stretchers, in the 

same gilt frames he designed for them. On other occasions, he had the paintings 

lined as a prophylactic measure, either during execution or shortly after 

completion (11). 

Hunt chose his medium and surface coating with the same view to permanency 

as his pigments and support. Instead of the popular megilp, a gelled 

preparation of linseed oil and mastic varnish much used and abused by nineteenth-century 

British painters, Hunt used amber colors, smooth-flowing 

tube paints with pigments bound in a drying oil and copal resin. For once, 

Hunt diverged from sound technique, failing to foresee the embrittling effect 

to the paint layer over time from the addition of copal to the medium, along 

with the eventual yellowing of his colors due to the oxidation of the varnish 

component. With the best of intentions, he defended his choice of medium, 

noting that "amber varnish ... protects the colours very perfectly, but has 

two slight disadvantages, as it lessens the brilliance of the white by the richness 

of the yellow tone in the varnish, and permits each touch to spread, though 

very slightly. Both these difficulties, however, occur immediately and may be 

calculated for" (12). 

In spite of his profound interest in stability of materials and techniques, if 

Hunt is remembered at all today as a technical innovator, it is as the author 

of the Pre-Raphaelite technique, a technique actually attempted by very few 

Katz 161

Pre-Raphaelite painters and never to any great extent. The myth of the Pre 

Raphaelite technique arose from a single paragraph in Hunt's 1,000-page, 

two-volume autobiography. Citing as an example the painting Valentine Rescuing 

Sylvia from Proteus (Fig. 7), Hunt stated that he would (13): 

Figure 7. William Holman Hunt, Valentine 

Rescuing Sylvia from Proteus (Two 

Gentlemen of Verona), 1851. Courtesy if 

the Birmingham Museums and Art Callery, 

Birmingham, England. This is the painting 

Hunt was working on when he described the 

wet-ground technique ill his autobiography (I, 

276): "The heads of Valentine alld if Proteus, 

the hands if these figures, and the brighter 

costumes in the same painting had been 

executed in this way . ... In the country we 

had used it, so fa r, mainly for blossoms of 

flowers, for which it was singularly valuable. " 

Paint daubs if van'ous mixtures of red are 

visible on the spandrels. 

Select a prepared ground, originally Jor its brightness, and renovate it, if 

necessary, with Jresh white when first it comes into the studio, white to be 

mixed with a very little amber or copal varnish. Let this last coat become 

oj a thoroughly stone-like hardness. Upon this suiface complete with exactness 

the outline oj the part in hand. On the morning Jo r the painting, 

with Jresh white (from which all supeifluous oil has been extracted by means 

oj absorbent paper, and to which a small drop oj varnish has been added) 

spread a Ju rther coat very evenly with a palette knife over the part Jor the 

day's work, oj such consistency that the drawing should Ja intly show 

through. In some cases the thickened white may be applied to the Jo rms 

needing brilliancy with a brush, by the aid oj rectified spirits. Over this 

wet ground, the colour (transparent and semi-transparent) should be laid 

with light sable brushes, and the touches must be made so tenderly that 

the ground below shall not be worked up, yet so Jar enticed to blend with 

the superimposed tints as to correct the qualities oj thinness and staininess 

[sic], which over a dry ground transparent colours used would inevitably 

exhibit. Painting oj this kind cannot be retouched except with an entire 

loss oj luminosity. 

In spite of the attention given to this quotation, however, it is rarely pointed 

out that Hunt describes this as a technique with which he experimented at 

one point in his career, used fo r specific design areas rather than entire canvases 

and not as a wholesale working method, as has been interpreted. Indeed, 

until the appearance of the autobiography in 1905, Hunt appears to repudiate 

his early experimental technique. It remains unmentioned in his first version 

of his memoirs, nor does it crop up in the frequent handwritten inscriptions 

in which he recorded details of technique on bare sections of canvas or support 

(14) (Fig. 8). Hunt dismisses its relevance to his career in an article on 

his painting technique that appeared in the magazine Porifolio in 1875, stating 

that he used wet grounds only from 1850 to 1854 to capture the effects of 

sunlight (15). While these five early years are viewed today as Hunt's period 

of greatest productivity and success, to his contemporaries his highest achievements 

came later in a career that spanned seven decades (16). 

The sudden prominence accorded to the technique in 1905, may be due not 

to Hunt himself, but to his wife, Edith, seeking to enhance his reputation 

through the implication of technical innovation. Suffering from glaucoma, 

Hunt dictated much of his memoir to Edith who, according to their granddaughter, 

among others, took liberties with the text, "deleting passages ... 

she considered unsuitable for posterity" (author's emphasis) (17). Edith's urge 

to improve went so far as taking advantage of her husband's blindness to have 

a studio assistant repaint Hunt's portrait of her, secretly slimming her waist 

and reddening her lips (18). It may well have been at her suggestion that the 

wet-ground technique, after fifty years of oblivion, abruptly became of importance 

in her husband's career, serving as one more opportunity to assert 

his innovation, skill, and pivotal role in the movement he had helped to shape. 

Along with the disproportionate attention given to the Pre-Raphaelite technique 

comes a misunderstanding of the technique itself, due to a fu ndamental 

misreading of the word "ground." In Hunt's writings, the term often refers 

to an imprimatura, or underpaint, layer, rather than the intermediary priming 

that prepares a solid or fabric support to receive paint. In his diary, fo r example, 

Hunt refers to "lay[ing] a new ground for the left shoulder, which I 

do of white, cobalt green, and cadmium" (19). Further investigation confirms 

that Hunt's "wet ground" was actually a layer of paint, the "fresh white" of 

his writings referring to flake white oil paint, as he indicated in a letter of 

1878 (20): 

162 


F(gure 8. Illfrared rriflectograph of the hal1dwritten notatioll under the paint layer il1 the upper right 

corner of The Miracle of the Sacred Fire, Church of the Holy Sepulchre at Jerusalem, at the 

HalVard University Art Museums. Inscriptions such as this one are frequel1tly fo und in paintillgs by 

HUllt, either on the reverse of the supports, 011 bare spandrels, or ul1derneath the pail1t layers. Image 

provided by the author. 

I am obliged to wait long Jo r the drying oj the paint I put on to Jo rm a 

Jresh ground, and while patience is being exercised as a necessity it seems 

but a little virtuous to summon hope to help me over the completing oj the 

parts where the ground has finally become workably [sic] even and dry. 

Indeed, Hunt complains here that the underpaint layer is wet, rather than 

seeking to exploit its freshness as part of his technique. 

This reevaluation of the Pre-Raphaelite technique is consistent with the results 

of analyses carried out on cross sections taken from various paintings by 

Hunt, which indicate the blending of paint layers wet-into-wet, a common 

technique throughout the history of painting, rather than the swirling of paint 

layers into wet primings (21). This clarification not only places Hunt's 

achievements in a far more plausible technical frame, but also emphasizes, as 

he did throughout his career, his connection with and indebtedness to the 

Katz 163

paintings techniques of the past masters, rather than the faddish disregard for 

tradition prevalent among his contemporaries. 

Conclusion 

In seeking reform, Hunt sought not to stop the wheels of progress but to 

channel them. His advocacy for the rights of the consumers in a time of 

rampant laissez-faire capitalism had ramifications throughout his century and 

our own, contributing directly to the enactment of regulations governing the 

safety of commercial goods and legislation regarding truth in advertising. His 

sophisticated understanding of the science of painting, in spite of limited 

formal schooling, and the impact of environment in the preservation of art 

place him as a key figure in the development in the professions of art conservation 

and museology. His understanding of historic painting techniques, 

well advanced fo r his day as well as our own, generated a revival of the craft 

of painting after a generation notorious for technical inadequacies. Too often 

he has been dismissed by art historians as a minor painter of deeply tasteless 

religious scenes. His contribution to nineteenth-century British painting and 

to the current stability of commercial artists' materials must not be underestimated. 


The author wishes to thank Kate Olivier (Fogg Art Museum, Harvard University) 

for having encouraged the undertaking of this research; Jacqueline Ridge (National 

Museums and Galleries on Merseyside, Liverpool), Zahira Veliz (private practice, London), 

and David Bomford (National Gallery, London) fo r facilitating research in Britain;Judith 

Bronkhurst (Witt Library, Courtauld Institute, London) and Leslie Carlyle 

(Canadian Conservation Institute, Ottawa) for their valuable insights into nineteenthcentury 

British art and techniques; and the staff of the following institutions fo r 

generous access to curatorial and conservation files pertaining to William Holman 

Hunt: Fogg Art Museum, Cambridge, Massachusetts; Wadsworth Atheneum, Hartfo 

rd, Connecticut; Tate Gallery, London; Courtauld Institute of Art, London; Guildhall 

Art Gallery, London; Walker Art Gallery, Liverpool; Lady Lever Art Gallery, Port 

Sunlight; Manchester City Art Galleries, Manchester; Birmingham Museums and Art 

Gallery, Birmingham; and the Roberson Archives, Hamilton Kerr Institute, Cambridge 

University, Cambridge. 

Notes 

1. Harley, R. D. 1970. Artists' Pigments c. 1600-1835: A Study in English Documentary 

Sources. London: Butterworth Scientific. Also, Gettens, R., and G. Stout, 1966. 

Painting Materials: A Short Encyclopaedia. New York: Dover Publications, Inc. 

2. Bomford, D., et al. 1990. Art in the Making: Impressionism. London: The National 

Gallery. 

3. Standage, H. C.,1886. The Artists' Manual of Pigments showing their Composition: 

Conditions of Permanency, Non-Permanency, and Adulterations; Effects in Combination 

with Each Other and with Vehicles; and the Most Reliable Tests of Purity . .. . London: 

Crosby, Lockwood & Co. 

4. Standage, op. cit., 53-54. 

5. A Pre-Raphaelite Friendship: The Correspondence of William Holman Hunt and John 

Lucas Tupper, no. 114 (11 August 1875). 1986. Eds. J. H. Coombs, et al. Ann 

Arbor, Michigan: UMI Research Press, 20l. 

6. Hunt, W H. 1880. The Present System of Obtaining Materials In Use By Artist 

Painters, As Compared With That of the Old Masters. Journal of the Society of 

Arts 28 (23 April 1880), 492. 

7. Hunt, op. cit, 492. 

8. Ibid., 493. 

9. Ibid. 

10. Hamerton, P. G. 1875. Technical Notes-W Holman Hunt, an interview with 

the artist. The Porifolio: An Artistic Periodical VI (1):45. 

11. Original paint extends over onto the tacking margins of the lining canvas and 

the gummed tape edging the lining of the Birmingham Finding of the Saviour in 

the Temple. Conservation files, Birmingham Museums and Art Gallery. 

12. Hamerton, op. cit., 45. 

164 


13. Hunt, W. H. 1905-1 906. Pre-Raphaelitism and the Pre-Raphaelite Brotherhood. London: 

Macmillan & Co., (1):276. 

14. Hunt, W H. 1886. The Pre-Raphaelite Brotherhood: A fight fo r art. The Contemporary 

Review 49 (April):471-88, (May) :737-50, Oune):820-33. 

15. Hamerton, op. cit., 46. 

16. Of course, the Pre-Raphaelite Brotherhood in its strictest sense only endured 

from 1848 to 1854, so a purist might claim it as a Pre-Raphaelite technique, 

although all the Brothers quickly returned to traditional painting techniques, and 

only Hunt and John Everett Millais had ever developed it to any great extent. 

17. Holman-Hunt, D. 1969. My Grandfather, His Wives and Loves. London: Hamish 

Hamilton, 18. 

18. AITlOr, A. C. 1989. William Holman Hunt: The True Pre-Raphaelite. London: Constable, 

266. 

19. Hunt, W. H. 1872. Personal diary (May 25, 1872). Original manuscript in the 

John Rylands University Library, Manchester, England. 

20. A Pre-Raphaelite Friendship, no. 136 (16 October 1878), 258. 

21. Consistent results have been achieved in cross-section analysis carried out by the 

author, and by conservators at the Tate Gallery and University College, London, 

among others. 

Katz 165

Abstract 

The production of a painting in early 

nineteenth-century France followed 

a clearly defined sequence of 

steps. After drawings had been made, 

the composition was outlined on the 

prepared canvas and the modeling 

was indicated, often with a reddishbrown 

"sauce." Local color, light, and 

shade were laid in; this sketch was 

elaborated using a full range of tones 

laid out individually on the ebauche 

(palette). The final stage of painting 

refined this process further. This is 

demonstrated in Paul Delaroche's 

The Execution of Lady Jane Grey 

(1833). A pupil of Watelet and Gros, 

Delaroche received much popular 

acclaim during the 1830s, and a 

number of eminent painters passed 

through his studio. Thus he occupies 

a central position in the history of 

academic painting. 

Paul Delaroche: A Case Study of Academic Painting 

Jo Kirby* and Ashok Roy 

Scientific Department, National Gallery 

Trafalgar Square 

London WC2N 5DN 


Introduction 

The July Monarchy of Louis Philippe, who came to the throne in 1830, was 

a period of technological advance and increasing industrialization, marked by 

the rise of a wealthy and influential middle class. It was distinguished by its 

adherence to the philosophy of eclecticism, not only in politics, but also in 

the realm of the arts. Official art followed a middle course between the two 

dominant trends, Classicism and Romanticism, showing the careful composition, 

drawing, and modeling of the former and an interest in the subject 

matter and emotional content of the latter. The painting of the juste milieu 

could justly be described as the art of the bourgeoisie. Paul Delaroche, who 

rose to prominence at this time, was one of its most popular and successful 

exponents (1, 2). 

Born in 1797, the son of an art dealer, Delaroche (christened Hippolyte) 

entered the studio of Antoine-Jean Gros, a disciple of David, after early training 

with the landscape painter Louis-Etienne Watelet and with Constant 

Desbordes. The recipient of many honors, royal patronage, and several official 

commissions, Delaroche achieved early Salon success. In 1833, he inherited 

Gros's studio and became a professor at the Ecole des Beaux-Arts (3, 4, 5, 6). 

His atelier was perhaps the busiest and most effective of the period; his pupils 

included Gerome, Daubigny, Millet, Monticelli, and Thomas Couture, himself 

the master of Edouard Manet (7). 

The Execution of Lady Jane Grey 

Delaroche's The Execution if Lady Jane Grey, finished in 1833, achieved considerable 

success at the Salon exhibition of 1834. The subject, drawn from 

English Tudor history and depicted with ostensible accuracy, appealed to popular 

taste. The scene depicted-the moment immediately before the beheading-was 

that of the greatest dramatic tension; it also touched the sensibilities 

of the public without disgusting them. As Etienne-Jean Delecluze wrote, 

"The spectator can contemplate the axe ... without horror" (8) (Plate 33). 

The blindfolded Lady Jane fu mbles for the block; a figure who is probably 

intended to be Sir John Brydges, the Lieutenant of the Tower of London, 

gently guides her hand. On the left, a despairing lady-in-waiting turns her 

face towards the massive column; the other lady-in-waiting, her mistress's 

discarded dress across her lap, has fainted. The impassive executioner stands 

on the right. Delaroche's historical sources for the painting included the Martyrologue 

des Protestans of 1588, quoted in the Salon catalogue (9) . There were 

a number of other publications, as well as the works of other painters, upon 

which he could have drawn, including Hans Holbein the Younger's painting 

of Anne of Cleves, which was at the Louvre (10, 11). Delaroche undertook 

exhaustive research before any painting project (12). 

Adored by the crowds, the painting was praised and condemned in almost 

equal measure by the critics. The criticism that Delaroche's treatment was 

theatrical rather than dramatic, voiced by Gustave Planche among others, is 

interesting as it may reflect an aspect of the artist's practice: Delaroche used 

small model rooms, within which he arranged wax figures to assist in the 

composition of his paintings (13, 14). According to Edward Armitage, a for- 


166 


Figure 1. The Execution of Lady Jane Grey, after cleanirlg a/1d before restoration. 

mer student, the use of similar "boxes" to study the fall of light upon small 

draped nlOdels was not uncommon (15). It is fair comment that the scene 

depicted takes place uncomfortably close to the picture plane within a stagelike 

space. 

The 1834 Salon catalogue states that the painting was owned by Prince Anatole 

Demidoff, who paid 8,000 francs for it. It remained in private ownership 

until it was bequeathed to the National Gallery by the second Lord Cheylesmore 

in 1902 (16). In 1928 it suffered flood damage while at the Tate 

Gallery and was later described as destroyed (17). In 1973 it returned to the 

National Gallery, where examination showed that the condition of the painting 

was not as bad as had been feared, consisting principally of tears in the 

canvas (the largest tear through the executioner's feet) and associated paint 

losses. Figure 1 shows the picture before restoration after it had been repaired, 

double lined with wax resin, and cleaned (1974-1975). Evidently its construction 

must have been relatively sound, and this was confirmed by its recent 

technical examination (18). 

Method and materials of painting 

The highly fmished Salon painting or Academic picture was the result of a 

well-established procedure; the artist would make preliminary studies, an esquisse 

(a sketch of the intended composition), and detailed drawings before 

transferring the design to canvas and beginning to paint (19, 20, 21,22). These 

Kirby and Roy 167

Figure 2. Paul Delaroche, Study for The Execution of Lady Jane Grey, ca. 1833. Watercolor 

and body color over pencil, with varnish, 18.4 X 14.3 em. University of Manchester, f;IIhitworth Art 

Gallery. 

steps may be observed in The Execution if Lady Jane Grey and other works 

by Paul Delaroche. It must be remembered that the artist may have been 

assisted by students at certain stages in the production of a painting of this 

size, even though there is no obvious indication of this here (23). 

Compositional sketches for several of Delaroche's paintings survive; Joan of 

Arc in Prison (London, Wallace Collection), for example, was painted as the 

sketch fo r Joan if Arc . . . Interrogated in Prison by the Cardinal of Winchester 

(1824, Rouen, Musee des Beaux-Arts) (24, 25). Most are smaller than the 

fmal versions and all are more freely painted. Delaroche felt strongly that a 

preliminary sketch embodied the artist's imaginative process and inspiration 

(26). The only known compositional study for The Execution if Lady Jane 

Grey is a small watercolor in the Whitworth Art Gallery, Manchester (Fig. 2). 

Here, Delaroche sets the figures within a Romanesque interior and only the 

executioner, who stands in profile holding a sword, is significantly different 

from the large final version (27). Many minor alterations can be attributed to 

scaling up and improving the composition: modifYing the background arcade 

and staircase, for example, and simplifYing the costumes. The striking transformation 

is in the color, which is so resonant and warm in the finished 

version in contrast to the cooler and less coherent tonalities of the watercolor. 

In the sketch, the executioner is dressed in dull green and red; the attendant 

facing the column is portrayed in deep blue, rather than rich dark purple. In 

the sketch, light plays evenly across the room; in the painting, it is more 

concentrated on the figures, although the pattern of light fall is similar. The 

squat, oddly appealing figures of the tiny watercolor have been transformed 

into an elegant, theatrical "tableau." 

The next step was to make drawings for the composition's elements. Much 

importance was traditionally attached to drawing, and Delaroche produced 

hundreds of drawings during his career (28). Several must have been made 

for The Execution if Lady Jane Grey. Two studies for it on paper certainly 

survive, one in the Musee du Louvre in Paris, the other in the British Museum 

in London (Fig. 3) (29). The Louvre sheet shows the figure of the 

executioner on the left, squared up fo r transfer: It is very close to that in the 

168 


Figure 3. The Execution of Lady Jane Grey. Pencil drawing for the central group, 18 X 16.5 

cm. London, British Museum {fo rmerly National Gallery, London}. 

finished painting. The two other figures on the sheet are studies for Lady 

Jane's ladies-in-waiting; neither has been squared up. That for the standing 

figure has been used with little modification. The other figure in the Louvre 

drawing appears to kneel and pray facing the central group; this composition 

was not used in the final painting, in which the slumped position of the 

attendant is very much as it is in the watercolor study-to greater dramatic 

effect. The British Museum drawing shows Lady Jane and Sir John Brydges; 

it has been used with little modification. There are small differences in details 

of costume between the drawings used and the finished painting. 

Apart from his use of draped figures in room sets, discussed above, the artist 

was in the habit of making wax or plaster models of elements of the composition 

when he felt there was a problem to be solved; a plaster model was 

made of the two princes, for example, in The Children if Edward IV (1830, 

Musee du Louvre) (30). This appears not to have been necessary for the Lady 

Jane Grey. 

The chosen drawn versions of the figures would then have been transferred 

to the prepared canvas. The canvas for a painting of this size would have been 

stretched to order, apparently from a single piece of medium-weight linen; 

the largest stock canvas was a toile de 120 (about 1.3 X 1.9 m) (31). The roll 

of canvas used was already primed with a ground of lead white in linseed oil. 

Over this first priming, Delaroche then had a second ground-also of lead 

white, but this time bound in walnut oil-applied to the stretched canvas. 

Delaroche seems to have transferred designs for the single figures or groups 

individually; infrared reflectography has revealed "squaring-up" lines under 

the paint of Lady Jane's dress, her seated attendant, and in the area of the 


Figure 4. Il1frared reftectogram from Lady jane Grey's skirt in The Execution of Lady Jane Grey. 

block, but the scale of these various grids suggests that there is no single 

system of squaring across the entire picture surface. The infrared images show 

many changes in the underdrawing; lines drawn across Lady Jane's wrists 

suggest that her dress was to be long-sleeved, as in the watercolor sketch (Fig. 

4). At the drawing stage, Delaroche also indicated such fe atures as shadows 

in drapery folds, as seen in the standing woman's dress. 

In many unfinished French paintings of this period it can be seen that the 

underdrawing has been strengthened with a translucent brownish wash, sometimes 

known as "sauce" (32, 33). In this case, no such material could be 

detected with certainty in any of the cross sections examined, and only the 

charcoal drawing was fo und; possibly a grayish wash, for example, was applied 

in the shadows of folds. 

Subsequently the figures and background were laid in using brownish or 

grayish shades of paint, composed of lead white combined with a variety of 

tinting pigments (Cassel earth, ochres, and other natural earth pigments, as 

well as small quantities of cobalt blue, a red lake pigment and black) . The 

tonality of this underpaint bears some relation to the color of the paint that 

was to be applied on top, thus the background and black garments are underpainted 

in shades of gray, while the underpaint of the flesh varies from a 

warm beige fo r the executioner to a grayish white for Lady Jane. The underpaint 

also indicates light and shade, by varying the proportion of ochres, 

black, and cobalt blue in the mixture, the color of the underpaint of Lady 

Jane's dress is changed from a pale beige in areas of highlight to a dark brown 

in the deepest shadow. Occasionally, the underpaint is similar in color to the 

intended local color; for the cushion it is a translucent green consisting of 

black, verdigris, Pruss ian blue, yellow ochre, and perhaps a yellow lake. 

At this point the painting would have had an appearance not unlike a grisaille 

version of the final composition, an element of Delaroche's practice mentioned 

by Delaborde (34). Infrared reflectography has revealed many pentimenti, 

however, bearing out comm.ents by students that Delaroche frequently 

reworked passages during painting (35). The executioner appears originally 

to have had a sash around his waist, for example, and the position of the pike 

behind the balustrade and the angle of the banisters have been altered. The 

underpaint for the executioner's tights is mauve-gray (composed of lead 

170 


Figure 5. Detail of Lady Jal1e Grey from The ExeclItion of Lady Jane Grey. 

white, red lake, and black), similar in color to that used in the watercolor 

sketch; the presence of a thin layer of black paint above it, darkening or 

obliterating it, suggests a change of mind. Above it is a much warmer version 

of the underpaint layer, containing more red lake pigment, followed by the 

desired red paint (Plate 34a) . 

This grisaille probably represents the ebauche stage of the painting (36). 

When it was dry, the local color was applied. Bouvier describes the use of a 

series of premixed, graduated middle tones applied side by side and then 

blended to give an smooth passage from light to dark across each part of the 

painting, for both the ebauche and the final paint layers (37). The sim.plicity 

and precision of the layer structure and the generally close relationship between 

the depth of tone in the grisaille and the chiaroscuro of the finished 

painting, as demonstrated in the painting of Lady Jane's dress, suggests that 

something approaching this may have been done (Fig. 5). Only one or two 

quite thin, even layers of paint may be present over the underpaint, with 

perhaps an additional glaze or highlight. Observation of the paint surface, 

however, and the occasional presence of extrem.ely thin scumbles or glazes of 

paint (or sim.ply a greater concentration of paint medium) at the top of a 

paint layer reveal the careful blending of tones. One paint layer frequently 

appears to have been applied over another while the layer below was still 

fresh, enabling one mid-tone to be merged into the next. Plate 34b shows a 

cross section of paint from the shadow of a fold in Lady Jane's dress. The 

grayish paint of the shadow, consisting of lead white with a little cobalt blue 

and Cassel earth, merges with the creamier paint of the layer below so com- 


pletely that it is difficult to distinguish between them. The lowest paint layer 

(above the grayish wash of the drawing and the white ground) is the dark 

undermodeling of the ebauche, containing Cassel earth, charcoal, and cobalt 

blue. This is probably sufficiently dark in color to contribute to the observed 

chiaroscuro of the dress (a similar mixture is used as a final glaze fo r the 

deepest shadows) . The creamy mid-tone of the dress consists principally of 

lead white, with traces of yellow ochre, Cassel earth, and cobalt blue; the 

proportion of the tinting pigments is altered as one tone blends into the next, 

however. The blending process is also reflected in the presence of a number 

of paint layers in subtly different shades of cream in different parts of the 

dress. Only in the lightest highlights is lead white used almost pure, ground 

in walnut oil; even here, the presence of a trace of cobalt blue gives it coldness. 

This general pattern of paint construction is repeated elsewhere in the painting. 

Passages of paint containing a pigment used more or less unmixed are 

very rare indeed, but in the red glazes supplying the purplish red of the 

brocade dress, on the lap of the seated lady-in-waiting in Figure 6c, and the 

shadows on the executioner's tights, a crimson lake pigment was used in this 

way. In both cases, the dyestuff present was that extracted from a cochineal 

insect, Dactylopius coccus Costa, on a substrate consisting largely of hydrated 

alumina. The brown paint of the brocade, containing yellow ochre with other 

iron oxides and black, contained ordinary linseed oil; the red glaze, however, 

contained heat pre-polymerized linseed oil and a little mastic resin (Plate 34c). 

This indicates the use of a varnish of the type recommended as a painting 

medium, perhaps the jellylike vernis des Anglais described by Merimee as being 

particularly suitable fo r glazes because it could be brushed on so easily (38). 

The presence of a resin-containing medium is also suggested by the whitish 

fluorescence exhibited by the glaze layer in ultraviolet illumination under the 

mIcroscope. 

The paint used for the red of the executioner's tights is perhaps surprisingly 

complicated, as it contains two red lake pigments mixed with vermilion and 

lead white (Plate 34a). Examination under the microscope suggests that one 

is the cochineal lake used in the glaze; the other, less crimson in color, was 

not present in sufficient quantity fo r analysis, but its pronounced orange-pink 

fluorescence in ultraviolet illumination suggests that the dyestuff may have 

been extracted from madder root, the use of which was being developed in 

France at the time (39). Curiously, the same lake (mixed with black and cobalt 

blue) is used rather than the more crimson cochineal lake for the attendant's 

purple dress. Even the velvety black of Sir John Brydges' garment is a mixed 

color: it contains a subtle combination of black, Prussian blue, red lake, and 

a translucent yellow pigment (Plate 34d). This combination is similar to 

Edouard Manet's tinted darks in Music in the Tuileries Gardens, painted thirty 

years later (40). Quite marked brushwork is visible in the black garments of 

Sir John and the executioner; analysis shows that Delaroche used a varnishtype 

painting medium of similar composition to that used for the red glaze, 

which retained the texture of the brush strokes. 

To summarize, Paul Delaroche painted a grisaille of his composition and then 

colored it in; the labor necessary to produce the finished painting was, however, 

considerable barely revealed on its bland surface. The painter's craftsmanship 

and understanding of materials cannot be denied; his reliance on 

lead white and ochres in particular and the absence of bitumen have resulted 

in a paint film in remarkably good condition, considering the recent history 

of the painting. The use of a varnish-type paint medium has often proved to 

be a recipe fo r disaster; in this case the paint shows few of the defects often 

caused by its use. Delaroche appears to have added only cobalt blue and 

synthetic ultramarine (and possibly an improved madder lake) to what could 

be described as a conventional eighteenth-century palette. 

The production of the Salon painting was not the end of the story. Public 

awareness of successful Salon paintings was increased by means of reproductions 

in the press; several of Delaroche's paintings, including Children of Edward 

172 


Figure 6. Paul Delaroche, The Execution of Lady Jane Grey, ca. 1834. Canvas, 45. 7 X 53.3 

e/11. Corporation of London, Guildhall Art Gallery. 

IV, were reproduced by lithography. Artists were also able to keep their works 

in the public eye by means of high-quality engravings and Delaroche had an 

arrangement with the print publisher Adolphe Goupil to produce and publish 

engravings, after his paintings (41). For large paintings, including Lady Jane 

Grey, The Children of Edward IV, and Joan of Arc, he painted a reduced-scale 

copy from which the engraving could be made (42) . In the case of Lady Jane 

Grey, it is possible that the painting now in the collection of the Guildhall 

Art Gallery, Corporation of London, is the reduced-scale copy (Fig. 6) (43). 

The engraving, by Mercurij, one of several engravers who engraved the painter's 

works fo r Goupil, was begun in 1834, but was only completed in 1857, 

the painter himself having died the previous year (44). 


The authors would like to thank John Leighton, curator of nineteenth-century paintings 

at the National Gallery, fo r generously sharing the information he has gained 

during his research with the authors; Rachel Billinge, Leverhulme Research Fellow 

at the National Gallery, for infrared reflectography carried out on the painting; Raymond 

White fo r examination of the paint medium by gas chromatography mass 

spectrometry; Sara Hattrick and the Photographic Department for photography. 

Notes 

1. Boime, A. 1971. The Academy and French Painting in the Nineteenth Century. London: 

Phaidon, 9-21. 

2. Boime, A. 1980. Thomas Couture and the Eclectic Vision. New Haven and London: 

Yale University Press, 3-35. 

3. Delaborde, H., and J. Godde. 1858. Oeuvre de Paul Delaroche, reproduit en photographie 

par Bingham, accompagne d'une notice sur la vie et les ouvrages de Paul Delaroche 

par Henri Delaborde, et du catalogue raisonne de l'oeuvre par Jules GoddlE. Paris: Goupil. 

References to the essay are cited as Delaborde, 1858; those to the catalogue as 

Godde, 1858. 

4. Ziff, N. D. 1977. Paul Delaroche: a study in nineteenth-century French history painting. 

New York and London: Garland Publishing, Inc. 

5. Boime, 1980. Op. cit., 40-41, 52-60, 70-75. 


6. Foissy-Aufrere, M.-P 1983. La jeartlle d'Arc de Paul Delaroche, Salon de 1824: 

dossier d'une oeuvre. Rouen: Musee des Beaux-Arts, 82-87. 

7. Hauptman, W 1985. Delaroche's and Gleyre's teaching ateliers and their group 

portraits. Studies in the History of Art (18) :79-82, 89-97. 

8. Delecluze, E.-J. 1834. Salon de 1834 (2nd article). joumal des debats, politiques et 

litteraires 5 March 1834. 

9. Paris Salon de 1834. Explication des ouvrages de peinture, sculpture, architecture, gravure 

et lithographie des artistes vivalls exposes au Musee Royal Ie 1" mars, 1834. Paris: 

Vinchon. Facsim.ile edition, 1977. New York and London: Garland Publishing 

Inc., 52 (no. 503). 

10. Ziff, op cit., 126-35. 

11. Ganz, P 1950. The Paintings of Ha1/5 Holbeill. London: Phaidon, 251 (no. 107) 

and plate 148. 

12. Mirecourt, E. de, [C-J.-B.Jacquot]. 1871. Histoire co/I.temporaine: portraits et silhouettes 

au XIX" siMe. 119. Delaroche, Decamps. 3rd ed. Paris: Librairie des contemporains, 

16-17. 

13. Planche, G. 1855. Salon de 1834. In Etudes sur I 'ecole franfaise (1831-1852), peinture 

et sculpture. Paris: Michel Levy Freres, Vol. 1, 241. 

14. Boime, 1980. Op. cit., 56. 

15. Armitage, E. 1883. Lectures on painting, delivered to the students of the Royal Academy. 

London: Tri.ibner & Co., 254-56. 

16. Gould, C 1975. Delaroche and Gautier: Gautier's views on the ((Execution of Lady 

jane Grey" and on other compositions by Delaroche. Leaflet accompanying the exhibition 

of the restored painting. London: Trustees of the National Gallery, Publications 

Department. 

17. Alley, R. 1959. Tate Gallery catalogues. The fo reign paintings, drawings and sculpture. 

London: Trustees of the Tate Gallery, 298. 

18. Examination of the paint cross sections and pigments was carried out by Ashok 

Roy using optical mjcroscopy, scanning electron microscopy coupled with energy-dispersive 

X-ray analysis, and X-ray diffraction. Identification of the paint 

media by gas chromatography mass spectrometry was performed by Raymond 

White. Identification of lake pigment dyestuffs by high performance liquid chromatography 

was carried out by Jo Kirby. The conservation treatment is described 

in the National Gallery conservation records. 

19. Boime, 1971. Op. cit., 22-47. 

20. Bouvier, P-L. 1832. Manuel des jeunes artistes et amateurs en peinture. 2nd ed. Paris: 

F. G. Levrault. 

21. Paillot de Montabert,J.-N. 1829. Traite complet de la peinture. Paris: Bossange pere, 

Vol. 9, 36-65. 

22. Arsenne, L.-C 1833. Manuel du peintre et du sculpteur [etc.]. Paris: Librairie encyclopedique 

de Roret, Vol. 2, 195-413. 

23. For example, Godde states that the ebauche of Straffo rd (1835, location unknown) 

was painted by Henri Delaborde from Delaroche's watercolor sketch and a wax 

model. Godde, 1858. Op. cit. (note 3), plate 19 and commentary. 

24. Ingamells,J. 1986. The Wallace Collection. Catalogue of Pictures, II, French Nineteenth 

Century. London: Trustees of the Wallace Collection, 114 (no. P604). 

25. Foissy-Aufrere, op. cit., 21-23. 

26. Boime, 1971. Op. cit., 45, 102-3. 

27. Godde, 1858. Op. cit. (note 3), plate 16 and commentary. 

28. Ziff, N. D. 1975. Dessins de Paul Delaroche au cabinet des Dessins du musee du 

Louvre. La Revue du Louvre et des Musees de France, XXV: 163-68. A further 

donation of drawings was made to the Louvre in 1982. 

29. Etude pour la 'Jane Gray" [sic], 1727RF. Pencil and red chalk, 23. 1 X 20 cm. 

See Guiffrey, J., and P Marcel. 1910. Inventaire general des dessins du Musee du 

Louvre et du Musee de Versailles. Ecole franfaise. Paris: Librairie centrale d'art et 

d'architecture. Vol. 5, 2-3 (no. 3535). 

30. Delaborde and Godde, op. cit., 15, plate 14 and conU11entary. 

31. Paillot de Montabert, op. cit., vol. 9, 144-47. 

32. Bouvier, op. cit., 236-37. 

33. Couture, T. 1879. Conversations on Art Methods lMetlLOde et entretiens d'atelierJ. 

(French edition, 1867). New York: G. P Putnam's Sons, 7. 

34. Delaborde, op. cit., 18. 

35. Armitage, op. cit., 79. 

36. Boime, 197 1. Op. cit., 36-41. 

37. Bouvier, op. cit., 207-74. 

174 


38. Merimee,J.-F.-L. 1830. De la peinture a I'lwile [etc.]. Paris: Mme. Huzard (facsimjle 

reprint, Puteaux: Erec, 1981), 70-71. 

39. Paillot de Montabert, op. cit., 258-71; see also Merimee, op. cit., 144-65. 

40. Bomford, D., J. Kirby, J. Leighton, and A. Roy. 1990. Art in the Making: Impressionism, 

London: National Gallery and New Haven: Yale University Press, 117, 

119. 

41. Mirecourt, op. cit., 24. 

42. Ingamells, op. cit., 99-100 (no. P276), 103-104 (no. P300). 

43. Knight, V 1986. The Works of Art of the Corporation of London: A Catalogue of 

Paintings, Watercolours, Drawillgs, Prints alld Sculpture. Cambridge: Woodhead 

Faulkner, 72 (no. 1052). 

44. Godde, op. cit. (note 3) , plate 16 and conunentary. The engraving measured 29 

X 36 cm. 


Abstract 

Turner's use of sketches on paper, his 

development of successful oil sketches 

into finished paintings, his preference 

for absorbent primings, and his 

modified oil media are described. 

His oil painting techniques circa 

1800-1850 are illustrated by studies 

of several works. His use of megilps 

(varnish-modified oil media) is outlined, 

along with his use of newly 

available manufactured pigments, and 

is compared with analyses from other 

works painted by British artists 

1775-1875, ranging from Reynolds 

to Whistler. 

Painting Techniques and Materials of Turner and Other 

British Artists 1775-1875 

Joyce H. Townsend 

Senior Conservation Scientist 

Tate Gallery, Millbank 

London SW 1P 4RG 


Introduction 

This paper presents important aspects of J. M. W Turner's technique, discussed 

before by the author in greater detail, by examining several paintings not 

previously described in this context (1, 2). Turner painted in oil fo r over fifty 

years (ca. 1798-1850), and it is interesting to compare his materials to those 

used in British paintings of the preceding and fo llowing twenty-five years (in 

the Tate Gallery collection, unless otherwise stated). Thus, this paper presents 

a comparison of the use of modified paint media and the adoption by various 

artists of new pigments produced between 1775 and 1875. 

Turner's oil painting techniques, compared to others' techniques 

Turner spent at least ten years as a watercolorist before he used oil as a paint 

medium, developing a range of techniques that he would utilize ever after in 

oil. In the earlier watercolors, transparent washes overlie the white paper 

except where Turner reserved highlights. The greens were made by mixing, 

overlaying, or physically mixing with his fingers, washes of brown and blue. 

There is very little underdrawing, and generally it is free rather than detailed. 

The lTlental image was transferred directly to the support. As Farington wrote, 

"Turner has no settled process but drives the colour about till he has expressed 

the ideas in his mind" (3). Turner produced over 20,000 pencil sketches and 

watercolors, now at the Tate Gallery, but few have a direct counterpart in oil. 

He hardly ever produced a detailed oil sketch, even for a commission, and 

when he sketched in oil he developed the best sketches into completed and 

exhibited works, rather than repainting them on a new canvas. 

Turner seems to have tried nearly all materials and methods once. The following 

descriptions apply to many of his paintings, if not all. Once he began 

to work in both media (always independently) , he utilized new application 

techniques and pigment in both media at about the same time. His early 

work in watercolor gave him an understanding of and liking fo r light-toned, 

absorbent surfaces, and honed his skills in the application of optical greens 

and blacks. Some eighteenth-century artists such as Wright of Derby (4) used 

white grounds to lend luminosity to their oil paintings, but many of Turner's 

immediate contemporaries were using thicker paint and warm-toned grounds 

up to 1820. Constable produced oil sketches upon mid-toned buff, red, or 

blue grounds, though his exhibited works do not show quite such a variety 

of grounds (5). After 1820, more British artists tended to use white grounds. 

Many commercially primed canvases are white during this time and up to 

the 1890s, when Sargent and Whistler used gray ones of varying tones in 

England (6) . Turner used white primings for a fair proportion of his oils, both 

exhibited and unfinished, in the first decade of the nineteenth century; in 

later decades most of his supports had white grounds. In a fair num.ber of the 

paintings examined, the primings consist of lead white in whole egg medium 

(7), most of the others being lead white in oil, not sized on the surface or 

between applications of priming, as had previously been usual. Absorbent 

surfaces gave the impetuous Turner a very rapid indication of the final color 

of the paint, and allowed him to develop the composition rapidly over fastdrying 

paint if it looked promising. 

176 


F(gure 1. J. M. W. Turner, Goring Mill and Church, ca. 1806-1807. Oil sketch on canvas, 

857 X 1162 111111. Courtesy of the Tate Gallery, London (N02704). 

Some works were abandoned at this stage, for example Goring Mill and Church 

(ca. 1806-1807), in which the buildings were lightly drawn in pencil before 

washes were applied of thinned linseed oil paint in green and brown for grass, 

buildings, and cattle, and in highly thinned, pale blue paint to suggest the 

clouds (Fig. 1). Other abandoned oil paintings from this date and later have 

little or no pencil underdrawing. Turner used it mainly to outline buildings 

or ships, where accuracy mattered. By contrast, landscapes and trees were 

usually freely painted. In later years, Turner used increasingly brightly colored, 

thinned paint as a first lay-in, generally brown for the landscape and blue for 

the sky, leaving white priming in areas that would later be depicted as yellow 

sunlight. He glazed down the colors as he worked. Cross sections from most 

oils show thin transparent washes, overlaid by thin paint layers in the same 

colors, lightened with lead white. Some abandoned works have a patch of 

bright red or blue in the foreground, which Turner would have developed 

into something appropriate as the image evolved, perhaps a red buoy in the 

sea, or a brightly dressed figure in a landscape. 

Turner's earliest oils look thickly and conventionally painted at first glance or 

when viewed through accumulated yellow varnish, but this is deceptive. Dolbadern 

Castle, North Wales is a good example (Fig. 2). Turner presented it to 

the Royal Academy, London, when he obtained full membership at the age 

of twenty-eight. Recent cleaning revealed thin glazes of Mars orange (i.e., 

strongly colored synthetic iron oxide) and localized scumbles of black, and 

Mars orange or red, with white in the landscape. The surface was, in fact, 

vulnerable and potentially sensitive to solvents. Clouds were applied with a 

thicker, creamy-looking paint that retained brush marks, small areas were then 

scumbled over with quite bright yellows or pinks, applied rather lean. Naples 

yellow, reddish brown ochres, a purplish ochre, and ivory black provided these 

highlights, while the sky itself was painted in ultramarine, lightened with lead 

white. Turner continued to use ultramarine for finishing skies throughout his 

life, having often done the initial lay-in with smalt, though not in this case. 

He used Prussian blue glazes and washes over thicker layers of brown ochres 

and umbers for the greener parts of the landscape and the stream. The fastflowing 

water is indicated by small flecks of textured white paint. These may 

be lean strokes of paint applied in a pure oil medium. In other paintings, such 

areas were found to contain linseed oil but very rarely walnut oil, despite the 

fact that the latter was known to yellow less. It is much less likely that the 

Townsend 177

Figure 2. J. M. W. Tumer, Dolbadern Castle, North Wales, 1802. Oil 011 canvas, 1190 X 889 /11111. Courtesy of the 

Royal AcadelllY of Arts, LOlldo/l. 

textured clouds were painted in pure oil medium; they more resemble a 

megilp, a medium discussed in detail later. 

Many of Turner's oils are disfigured by wide contraction cracks long associated 

with natural bitumen or asphaltum. The paint of Dolbadern Castle is free 

of such defects, possibly because the subject did not require the extreme 

contrasts between sunny and shaded landscape that Turner achieved so effec- 

178 


Figl/re 3. Detail if lower left, s/lOwillg wlltractioll cracks ill vitulllell-rich paillt, jrolll J. M. W. TurlIer's 

The Hero of a Hundred Fights, ca. 1800, rel/,orked 1847. Oil 01/ cal1vas, 908 X 1213. 

COl/rtesy oj the Tate Gallery, LOlldoll (N00551). 

tively elsewhere through the use of bitumen, megilp, and resinous glazes. 

Bitumen has been detected in his Italian landscapes of 1820-1840 in particular, 

and appears at its worst when used as an underlayer for the foreground. 

In other works, Turner added to bitumen's poor drying characteristics by 

painting it directly over old damaged paint or by exposing canvas to extreme 

dampness and dripping water, so the support stressed the paint and caused 

cracking. It is not certain whether he ever used synthetic bitumen, a material 

severely criticized by his contemporaries, and reputed never to dry at all (8) . 

Figure 3 shows a detail from Turner's The Hero of a Hundred Fights that has 

bitumen-rich paint applied over oil paint a few decades old, showing severe 

cracking typical of this artist's oeuvre. Figure 4 shows even more disfiguring 

cracking, attributed to synthetic bitumen, in a study of Editha's head fo r 

Hilton's Editha and the Monks Searching fo r the Body of Harold. 

The full complexity of Turner's later paintings can be represented by The 

Dawn oj Christianity, The Flight into Egypt, exhibited in 1841. Turner used a 

square canvas with a white, absorbent priming of lead white and oil, and 

planned the work fo r a non-square frame. He sketched in octagonal cut-offs 

in pencil, drew a circle with pencil and compasses, and then painted a more 

or less circular image. He prudently made use of the corners of the canvas 

to try out colors, and quite large brush loads of paint survive there, uncontaminated 

by later layers. Figure 5 shows the framed painting, and Plate 35 

shows a detail of the lower right corner. The most textured paint, nearly free 

of pigment and by now honey colored, proved to be a megilp. The canvas is 

Townsend 179

Figure 4. Detail (showing cracks) from William Hilton the Younger's study of Editha's head for Editha 

and the Monks Searching for the Body of Harold, ca. 1834. Oil on canvas, 335 X 244 

mm. Courtesy rf the Tate Gallery, London (N00333). 

unlined. The paint dribbling down the tacking margins at an angle suggests 

Turner used a sloping easel, probably the tripod type depicted in his watercolors. 

There are very few descriptions of Turner painting, but observation 

of the paintings and cross sections makes it clear that he thinned paint excessively, 

until it contracted into islands as it dried (visible in the foreground); 

at other times, he mixed paint in drying oil on the palette so rapidly as to 

leave recognizable blobs of the added oil. Almost certainly, Turner completed 

The Dawn of Christianity at the Royal Academy in the three days required of 

other artists of his era fo r retouching sunken areas or appling varnish. The 

sky paint was applied rather thickly with a palette knife, as was white impasto 

in the lower right. Both paints consisted of good quality lead white with few 

impurities. Turner sometimes modified the sky with opaque scumbles, but 

rarely glazed it, emphasizing the contrast between the sky and the highly 

glazed landscape in the foreground. The middle ground was painted rapidly 

180 


Figure 5. J. M. W. Tumer's The Dawn of Christianity, the Flight into Egypt, exhibited 1841. Oil on cal1vas, 787 X 787 111m. 

Courtesy of Ulster Museum, Belfast, Northern Ireland. 

with megilped paint, and hence has high impasto that did not slump as it 

dried. Turner may even have mixed the megilp into the oil paint on the 

canvas, rather than on the palette. Beeswax and spermaceti wax, added to oil 

and both found in this painting (9), yielded a more flowing medium used to 

good effect in the more distant trees, though the prominent tree on the right 

has a more strongly textured trunk painted with megilp. The very thin, later 

glazes that ran off the edges and soaked into the corners of the absorbent 

canvas would have soaked less into medium-rich paint. Turner used emerald 

green here (and quite frequently by this date) to provide a stronger contrast 

with chrome yellows and orange than that provided by an optical green. 

Curiously, green mixtures of opaque blues and yellows are virtually unknown 

in his paintings. The original varnish may have survived on this painting in 

the hollows of textured paint, where it has entrapped hogs' hairs, as in other 

Turner oils in which the varnish may be original. Little can be said conclu- 

Townsend 181

sively about the varnishes Turner used. He may have varnished later works 

only when they were sold. 

Megilps compared to oil medium 

There are numerous references in artists' manuals and critics' reviews to megilps, 

combinations of mastic varnish and drying oil that gelled on mixing, 

and could then be mixed into pure oil paint on the palette (10). Megilps 

(thixotropic medium modifiers) had excellent handling characteristics both 

for impasto and glazing, but a severe tendency to darken, and to cause cracking 

whenever varnish was applied. Such materials had been used and criticized 

at least since Reynolds' time. Recent studies have shown that megilps 

can be made successfully from linseed oil cooked with lead acetate or litharge, 

cooled, and later mixed with mastic spirit varnish (11). These megilps subsequently 

show different behavior on aging, and Turner's paint is somewhat 

closer to lead acetate megilp. Pure megilp samples from the corners of The 

Dawn of Christianity certainly contain lead and behave when heated like artificially 

aged lead acetate megilps made from nineteenth-century recipes. 

Megilped white paint from this painting and the considerably earlier Dolbadern 

Castle has rather a different chemical composition from pure drying oil 

(12), being more hydrolyzed and less oxidized, as though a drier were present 

from the beginning. Documentary evidence indicates that Turner used lead 

acetate in copious amounts. 

Megilps are two-component materials whose properties vary significantly as 

the oil:resin proportion changes from 1:3 to 3:1. Film-forming capability, the 

tendency of the megilp to segregate afterwards, its tackiness, and its ability to 

absorb more or less dust than oil paint all depend on the exact proportions. 

Turner, who worked fast and fu riously, and never even paused to grind his 

pigments finely, must have used a variety of fo rmulations and proportions of 

megilp, albeit unconsciously. This has led to variations in degree of yellowing 

and solubility in Turner's paintings, and made them very sensitive to cleaning. 

Megilps, as well as paints with a startling variety of melting and softening 

points, have been fo und in numerous samples from Reynolds's later paintings 

of the 1780s. Reynolds's The Death if Dido, one of the "fancy pictures" 

wherein he is said to have used paint media that he never allowed his pupils 

to use, is particularly rich in megilplike layers and has some very striking 

surface defects. They arise when a modified oil layer is applied to fairly pure 

oil, whereupon microwrinkles form in the previously stable film, and the 

surface later exhibits a rough texture with drying cracks cutting in deeply. 

Figure 6. Detail if Anna's veil, painted over clouds and sky, from Joshua Reynolds's The Death of 

Dido, ca. 1781. Oil on cal1vas, 1473 X 2407. Courtesy of The Royal Collection, Her Majesty the 

Queen. 

182 


TabLe 1. Dates of manufa cture of inorganic pigments introduced 1775- 1875, with the earliest 

instance of their use fo und to date in oil paintings at the Tate Gallery. 

Pigment Firat manufacture Earlieat dete Artlat Source 

Prussian blue early18th c. by mid-18th c. varioos 

Marred mid or later 18th c.? 1755 1760 Reynolds a 

Mars yellow mid or later 16th c.? 1781 Reynolds a 

Mars brown mid or later 18th c.? 1781 Reynolds 

patent yellow published ca.1 ns 1781 1 Reynolds a 

Scheele's green used 1775 ca. 1606-1807 Turner a 

published 1778 

b 

cobalt green pub 1780 not found yet a 

made mid-19th c.? or 1825 

C 

barium sulphate used 1782 1840s Turner d 

Indian yellow mentioned 1766 1781 Reynolds a 

Brunswick green "nawn in 1795 not found yet e 

opaque oxide of chromium scld by Reid 1815 not found yet a 

cobalt blue 1802 in France ca.1801>-1807 Turner f 

chrome yellow 1814 1815 exh. 1814 Turner a 

pale lemon chrome posl 1814-1815? ca.1822 1823 Turner 

chrome orange post 1814-18151 ca.1822 1823 Turner 

emerald green disc. 1814 ca.1828 Turner I 

synthetic ultramarine 1826-1827 c8.1851 Turner a 

1857 Campbell j 

ultramarine green prepared 1828, made 1854-6 not found yet 9 

Chinese white 1834 in England 1835-1840 Turner a 

1852 Hunt 

viridian 1830s in France exh. 1842 Turner f 

cobalt yellow 1831 in Germany 18731 Whistler a 

made 1851. 1860 Eng 

d 

barium chromate French pat 1840s exh. 1843 Turner h 

strontium chromate sold by Reid 1835? Mulready 8 

1857 Campbell j 

orange vermilion ''new in 1835 exh. 1843 Turner i 

chrome scarlet 1840 W&N ca.1851 Turner e 

cadmium yellow 1843 1855 Millais j, i 

antimony orange patented in England in 1847 not found yet c 

zinc chromate 1850 not found yet 9 

cobalt violet dark mentioned 1859 not found yet e 

cadmium orange 1860 Rowney 1872 Whistler e 

unramarine red dev 1870 1880 not found yet 9 

Sources: (a) Harley, R. D. 1982. Artists' Pigments c11835. A Study of English Documentary Sources. 

London: Heinemann-Butterworth ; (b) Reid, G. 1835. Chromatography. London: Tilt; (e) Friedstein, H. 1981. 

Journal of Chemical Education (58):291-95; (d) Artists' Pigments. A Handlx>ok of their History and Characteristics. 

1986. Ed., A. L Feller. Washington: National Gallery of Art; (8) l. Carlyle, personal communication; (1) Bomford, D., 

J. Kirby, J. Leighton, and A. Roy. 1990. Art in the Ma king:Impressionism. London: National Gallery; (g) Gattens, R. 

J., and G. l. Stout. 1966. Painting Materials: A Short Encyclopedia. New York: Dover Publications; (h) D. 

Saunders and J. Kirby, personal communication ; (i) Harley, A. D. 1987. Some new watercolours in the nineteenth 

century. The Conservator (11):46-50; m E. Sheldon, personal communication. Paintings are privately owned. 

Figure 6 shows a detail of this painting, which was lined early in its history. 

Turner's techniques were similar, in that he would apparently apply any paint 

medium over any other to gain a beautiful but short-lived visual effect. 

Some of Turner's imitators achieved effects that today appear similar to his, 

but without the shrinkage and disruption of the paint. Turner's paintings may 

have changed greatly with time, as the critic John Ruskin thought. Several 

of Callcott's paintings have numerous thin glaze layers and less wet working 

than Turner used. Callcott, Etty, and other contemporaries of Turner did not 

have the patience to wait until previous paint had dried before they added 

another layer, but they did use varnish interlayers so that later paint could be 

applied safely. The Pre-Raphaelites favored a disciplined approach, too, as is 

well known. Hunt's Strayed Sheep (Our English Coasts) has very detailed, localized 

layers applied to already dried paint throughout the foreground, and, 

like works by Mulready and Collins, has a very well-preserved surface today 

(Plate 36). 

The other durable method is to use such slow-drying oil paint that wet 

working is possible, or to thin the paint so it fo rms a single layer. This was 

Whistler's method in the early 1870s, when he was producing nocturnes and 

harmonies (13). Instead of allowing the paint layer to grow thick, he scraped 

it back vigorously and began again, sometimes using paint so wet that the 

canvas had to be laid flat until it dried. 

British artists' adoption of new pigments, 1775-1875 

Table 1 summarizes dates of invention, first publication, and so on, of pigments 

that became available between the eighteenth century and 1875, and 

notes the earliest occurrence yet found in oil paintings at the Tate Gallery. 

Townsend 183

Reynolds was known to experiment with paint media, and he might be 

supposed to have tried out new pigments too. Five of his works have been 

examined, including three at the Tate. Not surprisingly, all five include traditional 

pigments such as lead white, ivory black, asphaltum, vermilion, and 

Naples yellow, while one or more included red lead, orpiment, blue verditer, 

smalt, ultramarine, and green earth (14). Prussian blue, invented early in the 

eighteenth century, was used frequently by Reynolds. Some of this Prussian 

blue has a different microscopical appearance from the modern variety; similar 

material, previously illustrated in color by Welsh (15), was identified in Turner's 

paintings up to about 1840, when Turner began using the "modern" 

variety as well. George Jones used the modern variety in 1832. Reynolds also 

used Mars red from 1755-1760, and Mars orange, red, and brown in 1781. 

He used Indian yellow in a work dated to 1788, and wrote about a material 

that may have been Indian yellow in 1784, two years before it has been noted 

in the literature (16). Patent yellow has also been tentatively identified in a 

painting of 1781. Table 1 shows that all these instances demonstrate early uses 

of these pigments. Reynolds' organic pigments (red, blue, and a green made 

from yellow and blue dyes) are still being investigated. 

Turner's early use (1800-1850) of new pigments in oil, summarized in Table 

1, has been described elsewhere (17). While it is true that more paintings by 

Turner than other individual artists have been analyzed, the inference that he 

was more innovative than his contemporaries is inescapable. Turner used Mars 

colors frequently, as did many of his fellow British artists, including Constable 

from circa 1810 (18). In contrast, Arnald, Farington, Hilton, and Callcott used 

only well-established pigments such as ultramarine, Prussian blue, Naples yellow, 

and vermilion. Artists who used new pigments quite soon after their 

introduction include the following: Constable-cobalt blue in 1817-1818, 

chrome yellow in 1816, and opaque oxide of chromium in 1837 or earlier; 

Briggs-chrome yellow and orange in 1826; and Mulready-emerald green 

in 1842 (19). Barium chromate and a pigment tentatively identified as strontium 

yellow were found in a Mulready of 1835. Cadmium yellow has been 

fo und in a Millais of 1855, and strontium yellow in a Campbell of 1857 (20). 

The latter included synthetic ultramarine, rarely used before 1850 except by 

Turner, because it had a poor reputation (21). Whistler used two shades of 

cadmium yellow regularly from 1864 (the earliest of his works at the Tate), 

and strontium yellow in two works circa 1864-1871 and in 1872, respectively 

(22) . The earliest tentative identification of cobalt yellow (mixed with barium 

chromate) is in one of Whistler's oils from the fo llowing year. No examples 

have yet been found of the cobalt violet shades that were available by the 

end of this period or by 1900. 

White pigments and their fillers are also interesting. Artists from the time of 

Reynolds and Romney tended to use lead white with pipe clay or china clay 

extenders, both for priming and paint. Gypsum has been found in many paint 

samples from Reynolds and Wright of Derby in the later eighteenth century 

(23). Zinc white was very rarely found in Turner's oils, and not yet in his 

fellow Royal Academicians works painted before 1847. But Hunt used it in 

1852, with lead white for a local imprimatura under the sky of Strayed Sheep 

(Our English Coasts) and with Prussian blue, presumably supplied as a tube of 

paler "Antwerp blue." Barium, attributable to barium sulfate, has been fo und 

as an extender in Turner's paintings of the 1840s, in the Hunt piece, and very 

frequently in Whistler's white paint from 1864-1 875, and beyond. 

A discussion of pigments that fell out of use during this period would be out 

of place here, but work is continuing in this area (24). Analyses of the paint 

media used by the artists mentioned here also continue, but as yet there are 

insufficient results for comparisons between Turner and many of his contemporanes. 


The Turner research was fu nded by the Leverhulme Trust and supported by the Tate 

Gallery, through Stephen Hackney. My thanks to the following people: Brian 

184 


Kennedy of the Ulster Museum, Viola Pemberton-Piggott of the Royal Collection 

and Helen Valentine of the Royal Academy, who arranged the examination of the 

Turners in their collections, Anna Southall and Gillian Osmond, who pointed out 

interesting nineteenth-century pigment samples, and Stephen Hackney, Tom Learner, 

Rica Jones, and Anna Southall for editing. 

Notes 

1. Townsend, ]. 1993. Turner's Painting Techniques. London: Tate Gallery. Information 

on Turner is published here, except where stated otherwise. 

2. Townsend,]. H. n.d. Turner's use of materials, and implications for conservation. 

In Turner's Paintings Techniques in Context. Ed. ]. H. Townsend, et al. London: 

UKIC. Forthcoming. 

3. The Diary ofjoseph Farington. 1978. Eds. K. Garlick and A. Maclntyre. London: 

Yale University Press, Vol. I (1799), 273. 

4. Jones, R. 1991. Notes for conservators on Wright of Derby's painting techniques. 

The Conservator (15):13-32. 

5. Cove, S. 1991. Constable's oil painting materials and techniques. In Constable. 

Eds. L. Parris and 1. Fleming-Williams. London: Tate Gallery, 493-529. 

6. These Whistler canvases are in the Birnie Philip Gift, Hunterian Art Gallery, 

University of Glasgow. 

7. Townsend,]. H. 1994. The materials of]. M. W. Turner, 2: primings and supports. 

Studies in Conservation (39):145-53. 

8. Carlyle, L. A., and A. Southall. 1993. No short mechanic road to fame: the 

implications of certain artists' materials for the durability of British painting: 

1770-1840. In Robert Vernon's Gift: British Artfor the Nation 1847. London: Tate 

Gallery, 21-26. 

9. Townsend, n.d., op. cit. 

10. Carlyle and Southall, op. cit. 

11. Townsend, n.d., op. cit. 

12. Rainford, D., personal communication. 

13. Hackney, S. 1994. Colour and tone in Whistler's nocturnes and harmonies 1871- 

1872. The Burlington Magazine CXXXVI (10):695-99. 

14. Some identifications were made by Rica Jones, Tate Gallery. 

15. Welsh, F. S. 1988. Particle characteristics of Prussian blue in an historical oil paint. 

journal of the American Institute fo r Conservation (27):55-63. 

16. Dubois, H. 1992. Aspects if Sir joshua Reynolds' Painting Technique. Unpublished 

typescript. Cambridge: Hamilton Kerr Institute. 

17. Townsend, n.d. , op. cit. 

18. Cove, op. cit. 

19. Ibid. 

20. Sheldon, E., personal communication. Paintings are privately owned. 

21. Townsend,]. H. 1993. The materials of]. M. W Turner, 1: Pigments. Studies in 

Conservation (38) :231-53. 


23. Jones, op. cit. 

24. Townsend, ]. H., L. Carlyle, S. Woodcock, and N. Khandekar. Unpublished research. 

Townsend 185

Abstract 

Several of Whistler's paintings were 

examjned and, where possible, analyzed; 

specific examples of pigment 

and media analyzes are given. Literature 

sources were searched fo r references 

to the artist's materials and 

methods. This paper describes Whistler's 

interest in texture, his use of 

dark gray grounds, his limited color 

ranges and careful preparation of the 

palette, his frequent erasure of his 

unsuccessful work, and details of his 

studio practice and portrait techniques. 

Probable changes in the 

appearance of his work are also discussed. 

Art for Art's Sake: The Materials and Techniques of 

James McNeill Whistler (1834-1903) 

Stephen Hackney 


Tate Gallery 

Millbank 

London SW 1P 4RG 


Introduction 

Several of Whistler's oil paintings were studied in detail; works in the Tate 

Gallery, in particular, were analyzed and examined before conservation treatment 

using a fu ll range of analytical techniques. Paintings in the H unterian 

Art Gallery, Glasgow from the Birnie Philip bequest were also examined in 

detail and in some cases analyzed (1). Paintings in the National Gallery of 

Art, Washington were examined in conjunction with their conservation records. 

Several other works were examined or viewed in isolation; finally, the 

paintings in the Whistler Exhibition (1994) were inspected on arrival at the 

Tate Gallery (2). It has therefore been possible to make some interesting 

observations about Whistler's materials and techniques, based on detailed examinations 

of selected works and supported by the wider survey, and to compare 

them with the literature on Whistler's methods. This interpretation may 

be useful in understanding the present condition of his works and in helping 

to decide appropriate conservation treatments. It is also of general interest for 

the viewer wishing to appreciate a particular painting. 

Supports 

From the 1880s, Whistler frequently carried small panels for sketching, focusing 

particularly on street scenes, views of the sea, and figure studies. Some 

of his nocturnes and early studies were also on larger panels. The scale of 

these works was small because he painted distant views approximately the 

size that they appeared to him. For his major finished paintings and portraits, 

however, he worked on canvas, which provided a texture he liked. He often 

chose quite heavy canvases and applied thin grounds in order to preserve 

their texture. On other occasions he might paint on a fine canvas, then have 

the painting glue-lined onto a coarser one very soon after completion or 

even during the time of painting. He did this quite deliberately and there is 

evidence that he wished to imitate the lined appearance of old master paintings 

(3). His later work, such as Mother of Pearl and Silver: The Andalusian 

(1888-1900), illustrates this desire to express the canvas texture. 

Grounds 

Feurc 1. James Whistler, Note in Red: 

The Siesta, 1883- 1884. Oil 011 palle/. 

Daniel j. Terra Collcetioll. 

Whistler's grounds were crucial to his methods. He came to London shortly 

after the Pre-Raphaelites had taken the use of smooth white grounds and 

intense color to extremes; some of his early grounds appear to be the white, 

commercially available grounds of the mid-century. Yet Whistler's training in 

Paris with Charles Gleyre (an academician who specialized in subjects taken 

from his travels in the Middle East), his early interest in the methods of the 

French Realists, and his experience with pastel, etching, and drypoint were 

more important influences. Whistler frequently applied a light gray imprimatura 

of oil paint on top of the ground to allow him to paint directly in a 

mid-ground technique. This technique is most easily observed in his small 

sketches on panel, for example, Note in Red: The Siesta (1883-884) (Fig. 1). 

After 1871 he increasingly used darker gray, exploiting the ground to develop 

full chiaroscuro effects in his later works, often setting his figure against a 

dark background. A photograph of his studio reveals that he had a black cloth 

186 

Historical Painting Techniques, Materials, alld Studio Practice

Figure 2. James Whistler, Nocturne in Blue and Silver: Cremorne Lights, 1872. Oil on canvas. 

The Tate Gallery, Londol1. 

to drape over an easel behind the sitter. Dark grounds were first successfully 

exploited by Whistler to produce his nocturnes of the 1870s. Similarly, fo r 

his watercolors, designs, and etchings, Whistler also sought old mid-toned or 

dark papers to form the basis of his images. 

To make dark grounds, he mixed ivory black and lead white, frequently modified 

with other pigments. In his Chelsea studio, the Greaves brothers helped 

him prepare his materials. Walter Greaves describes his ground as a very absorbent 

distemper, indicating that at this stage Whistler was not only modifYing 

commercial grounds but also preparing his own (4). The production of 

nocturnes was a period of experimentation. Nocturne in Blue and Silver: Cremarne 

Lights (1872) (Fig. 2) is on a partially scraped down, reused canvas from 

an earlier series with an unmodified, lighter commercial ground, whereas 

Nocturne in Blue Green: Chelsea (1871) has a dark ground applied over a white 

commercial priming on a mahogany panel. Nocturne in Black and Gold: The 

Fire Wheel (1875) has no commercial ground, only an artist's priming over 

the sized canvas. This ground consists of ivory black, chrome or cadmium 

yellow, and small amounts of lead white. 

Palette 

Figure 3. Brushes, paints, palettes, and other 

painting materials. Courtesy of the Hunterian 

Art Gallery, University of Glasgow, Birnie 

Philip Bequest. 

Several of Whistler's palettes survive, along with his brushes, charcoal, engraving 

tools and tube paints, presumably dating from the last period ofWhistler's 

life before they were donated to the Hunterian Art Gallery (Fig. 3). 

These have been examined and analyzed during the present study (5). For a 

short time (1898-1901), Whistler taught at the Academie Carmen. Reports 

by students of his methods demonstrate Whistler's concern fo r the preparation 

and layout of his palette. Colors were laid out in a specific order across the 

top of the palette from left to right: Prussian blue, cobalt blue, raw umber, 

burnt sienna, raw sienna, yellow ochre, a large blob of lead white, vermilion, 

Venetian red, Indian red, and black. Flesh tones were mixed just below the 

white, using the appropriate surrounding colors, and in turn these tones were 

modified by the black which was spread in a broad band curving downward. 

A preparation for the background color was mixed at the left. He then 

worked out all the colors and tones for his composition on his palette before 

placing any paint on his canvases. So obsessed was he with this "scientific" 

method that he would frequently examine his students' palettes yet ignore 

their paintings, wishing (he claimed) not to interfere with their individuality 

and free expression (6) . 

Hackney 187

Although this is a description of Whistler's late work, the artist acknowledges 

that he learned his approach from Gleyre and therefore would have employed 

similar methods throughout his career, with either a wider range of pigments 

or less mixing of pigments in his earlier work. Menpes confirms the attention 

to the palette but attributes to him the use of more intense lemon and cadmium 

yellows placed next to the ochre, and the use of rose madder instead 

of the Venetian red (7). 

Whistler kept a great number of brushes, many of which he used at a single 

sitting in order to prevent his dominant hues from becoming fu rther mixed. 

He spent much time cleaning and preparing these brushes, sometimes trimming 

and changing their shape. He had various types, including decorator's 

brushes to lay in grounds and backgrounds and extremely long-handled 

brushes for his portrait technique. 

Pigments analyzed 

Figure 4. James Whistler, Harmony in 

Grey and Green: Miss Cicely Alexander, 

1872- 1874. Oil on canvas. The Tate Gallery, 

London. 

Analysis of pigments fo und on Whistler's paintings confirm the use of complex 

mixtures of pigments usually involving at least some ivory black. 

Frequently, the same pigments are found mixed in different proportions 

throughout a painting. In nocturnes, mixtures of lead white, ivory black, cobalt, 

Prussian blue, and ultramarine have been found. By these means, Whistler 

produced his harmonies. For instance, Harmony in Grey and Green: Miss Cicely 

Alexander (1872) has a gray background, but this is an optical mixture of 

several colors based on lead white and ivory black (Fig. 4) . The same colors 

occur in the superficially white dress. By restricting the saturation and hue 

of his colors throughout the scheme, Whistler could model his fo rms without 

introducing discordant hues typical of the works of many nineteenth-century 

painters influenced by the Pre-Raphaelites and exploiting a range of unmixed 

pigments. This restriction allowed him to explore subtle color effects; fo r 

example, he could choose to highlight a red as in the Arrangement in Flesh 

Color and Black: Portrait

5). There are many such examples of scraped down paint that have not been 

entirely painted over (i.e., Cicely Alexander) and there must be even more that 

are no longer visible. Although tiresome for the sitter, this approach was 

essential to the artist. The harmony would not survive if the pigments were 

not all mixed in a single operation on the same palette. His paint is essentially 

opaque with little or no deliberate glazing. Discolored glazes would also destroy 

the harmony. Therefore, as a rule, all his transparent pigments were 

mixed with opaque. 

A further constraint in the studio is described by Walter Sickert (8): 

Figure 5. Detail from Cicely Alexander, 

showing areas scraped down to the canvas. 

Whistler was a great portrait painter, and prided himself on the precision 

oj his portraiture. Whistler also had extensive knowledge, and knew that 

the eye could only see at a glance an object which in size is one-third oj 

the distance between the eye and that object. In other words, if you are 

painting a man six Jeet high you should be 18 Jeet away Jrom him. 

Whistler had a very long studio, and he was accustomed to place his model 

against a black velvet background, and alongside his model he placed his 

canvas. His painting table was 18 Jeet away. He would stand at the 

painting table, carifully survey the model, then charging his brush with the 

requisite pigment he ran at Ju ll tilt up to the canvas and dropped it on the 

spot. 

This unlikely scenario is consistent with the small scale of much of his work 

and the flatness of the space around and behind his models. 

In his nocturnes, Whistler's paint was extremely dilute. He mixed his oil paint 

with large amounts of turpentine and also added mastic varnish to produce 

a paint that could be brushed freely and did not dry too matte. He called it 

his "sauce" and analyses at the Institute for Atomic and Molecular Physics in 

Amsterdam confirm its formulation from mastic and a drying oil. In his later 

work he may have used newly available petroleum oil, as recommended to 

him by Sickert in 1885. In portraits such as Arrangement in Yellow and Grey: 

Effi e Deans (1876-1878), he has allowed paint to run down the canvas, indicating 

just how diluted it was. His nocturnes were painted quite quickly 

(in a day) with minor modifications the next day, as Whistler freely admitted 

in his libel trial with Ruskin. The dark passages were first brushed in umber 

and black, or were simply the part of the imprimatura left exposed. Then a 

layer of "sauce" was applied using a large brush to scumble over the shadows 

and develop the lighter parts. The color and tone were controlled mainly by 

the thickness of application. The paint was worked wet-in-wet, scraped, 

rubbed, and even dragged across the wet surface, as in the reflections in Nocturne: 

Blue and Silver (1872) (Plate 38). Finally, details such as highlights were 

applied, after which the nocturne was put outside to dry in the sun. When 

it was dry, perhaps the next day, fu rther details could be applied, preferably 

from the same palette, but reworking was not possible at this stage. 

Condition and changes with time 

Changes in appearance have occurred on many of his works and, despite his 

frequent scraping down, minor adjustments to compositions-such as the 

positions of arms and feet-are now evident. Most significantly, many works 

have become darker and cooler as the thin paint has increased in transparency 

over the dark ground. Similarly, the small sketches with medium-toned 

grounds have lost contrast in the mid-tones. His earlier, more decoratively 

colorful work on light grounds or thickly painted work, although flawed in 

detail, has preserved its appearance much better and is more readily appreciated. 

The darkness of much of Whistler's later work is due, in part, to 

changes resulting from dark grounds and also possibly from the darkening of 

medium, but to some extent the effect was intended. 

Whistler appears to have always varnished his work and now many of his 

varnishes are excessively darkened and yellowed. In particular the discolora- 

Hackney 189

tion of varnish has dramatically affected the cool tones of nocturnes and some 

portrait backgrounds. 


This research has been a collaboration between the author and Dr. Joyce Townsend, 

who also carried out all the pigment analysis and investigated the media. Medium 

analysis was also carried out by the Institute fo r Atomic and Molecular Physics, 

Amsterdam. The author expresses his gratitude to Prof. Jaap Boon and Jos Pureveen 

fo r this work. Appreciation is also extended to the Hunterian Art Gallery, Glasgow, 

and to the National Gallery of Art, Wa shington, fo r allowing access to their collections; 

and to colleagues at the Tate GaLlery. 

Notes 

1. McLaren Young, A., M. F MacDonald, and R. Spencer, with the assistance of 

A. Miles. 1990. The Pailltings oj james McNeill Whistler. New Haven and London. 

2. Dorment, R., and M. MacDonald. 1994. james McNeill Whistler. Exhibition catalogue. 

London: Tate Gallery. 

3. Hackney, S. 1994. Colour and tone in Whistler's nocturnes and harmonies. TIle 

Burlington Magazine. Vol. CXXXVI, (1099): 695-99. 

4. Pennell, E. R.., and J. Pennell. 1911. The Life oj james McNeill Whistler. 5th rev. ed. 

London: Heinemann, 113-17. 

5. Townsend, J. H. 1994. Whistler's oil painting materials. The Burlington Magazine. 

Vol. CXXXVI (1099):690-95. 

6. Pennell, op. cit., 375-78. 

7. Menpes, M. 1904. Whistler as I Knew Him. London: Adam and Charles Black, 69- 

71. 

8. Sickert, W. 1930. Walter Sickert's class: methods of instruction. Manchester Guardian, 

31 November. 

190 


Abstract 

This paper discusses aspects of paintings 

executed on a photographic 

substrate. Inspired by Mervyn Ruggles's 

research into the use of this 

practice in the nineteenth century, 

this discussion presents examples of 

materials and techniques that have 

been proposed, manufactured, or 

employed fo r this purpose over the 

last 100 years. Information on this 

subject was gathered from the available 

literature and through personal 

communication with artists engaged 

in producing this type of artwork, as 

well as through discussions with 

manufacturers and museum personnel. 

North American and European 

instances are included, and details regarding 

the works of contemporary 

artists, such as Lynton Wells, Shirley 

Wiitasalo, Anselm Kiefer, and James 

Turrell, are noted. 

Painting on a Photographic Substrate: Notes Regarding 

Materials and Techniques over the Past 100 Ye ars 

John R. Gayer 

Koivikkotie 3 I 76 

00630 Helsinki 

Finland 

Introduction 

The idea of painting over photographic images has been present since the 

invention of photography. Sources regarding materials and techniques appropriate 

to this process have varied; some information has appeared in scientific 

and photographic publications. Communication with a number of contemporary 

artists, though, has shown that such literature rarely directly influenced 

their work. The principal factors guiding their production tended to consist 

of a willingness to experiment, a general awareness that such media may be 

combined, and the fact that this combination may serve their particular aesthetic 

or intellectual aims. The photographic image has also not only served 

as a replacement fo r the underdrawing, as was noted by Mervyn Ruggles in 

cases of nineteenth-century portraiture, but has been employed for a variety 

of purposes (1). 

The following presents a brief discussion outlining the evolution of materials 

and techniques over the last 100 years. The discussion culminates in a focus 

on contemporary practice since the 1960s, a period in which the painter's 

use of a photographic substrate became more widespread. 

From the late nineteenth century to 1950 

In the late nineteenth century, improvements in photographic technology and 

the availability of commercial products made it far easier for the artist to 

utilize such means in the painting studio. M. L. Winter of Vienna, for example, 

established an operation in 1877 fo r the extensive production of enlarged 

photographs on linen, and proprietary brands of gelatin-silver emulsion-coated 

linen were available from the 1890s (2, 3) . The introduction of faster 

bromide emulsions in the 1880s also greatly reduced the difficulties associated 

with producing enlargements, thus de-emphasizing the need for specialists 

(4). 

Oils were frequently used for painting on photographic images, but other 

materials were also proposed. Instructions published during the 1890s and 

thereafter commonly referred to the coloring of photographs on paper supports 

and were directed at those lacking artistic skills. Details regarding the 

media proposed for this purpose have been included here because this information 

provides a more complete range of the materials that may be encountered, 

as well as indicating the interests and concerns of the time. 

A number of proposals fo r the coloring of photographs appeared in Scientific 

American during the 1890s. An 1894 article, reprinted from Anthony's Bulletin, 

recommended the use of transparent and covering colors. These colors were 

to be made by mixing dry powdered pigments with a medium consisting of 

100 cc filtered albumin, 5 g ammonium carbonate, 3 cc glycerin, 4 cc liquid 

ammonia, and 25 cc water (5). Another article, deriving from Photographisches 

Archiv, noted the use of aniline dyes. These were dissolved in alcohol and 

applied on the reverse (6) . Another article noted the use of oils, watercolors, 

and pastels. These materials were only to be applied over a preparatory layer. 

Gelatin was recommended for oils; shellac was recommended fo r watercolors 

and pastels (7). 

A proposal, reprinted from the British Journal of Photography, for the use of 

wax media appeared in a 1919 issue of Scientific American Supplement (8). Two 

Gayer 191

ecipes were described. The first recipe suggested combining 1 oz. white wax 

(bleached beeswax) , 1 oz. carbon tetrachloride, 1 oz. turpentine, 1 oz. benzene 

(refined naphtha) and 1 dr (118 oz.) 0.88 ammonia. The second medium was 

described as an improved version consisting of 114 oz. white wax (Cera Alba), 

112 oz. spike oil of lavender, 1 dr hard primrose soap, 2 dr gum elemi and 3 

112 oz. turpentine. The author proposed using wax, based on the fact that 

wax, employed since ancient times, had proven its durability. The author's 

rationale was also based on the notion that many of the negative effects 

resulting from the use of oils and varnishes, such as darkening and yellowing, 

would be avoided. 

A 1934 edition of a manual describing the use of various media for coloring 

prints on paper noted that enlargements were to be mounted on beaver board, 

three-ply board, or linen canvas, and then stretched (9). Adhesive residue and 

dirt were to be removed before the application of a preparatory layer. This 

layer was to consist of either a 25% solution of glacial acetic acid, gelatin, and 

Lepage's Liquid Glue in water (1:20), or a mixture consisting of 3 oz. of paste 

(made from 1 oz. pure casein, 180 g powdered borax, 3 oz. water), 3 oz. 

alcohol, S drops glycerin, and 5 drops carbolic acid. The image was then 

ready to be painted with oils, watercolors, or a medium consisting of tempera 

colors and the casein mixture. 

The 1936 edition of a manual on retouching stated that the best results for 

a portrait in oil were to be obtained by painting over a carbon print on 

canvas. A priming, consisting of a starch solution to which some mucilage 

had been added, was to be applied to the print before it could be painted. 

Reference was also made to the "Russian method" of coloring, in which 

layers of transparent oils were rubbed on with cotton (10). Materials recommended 

for this process included Marshall's Transparent Photo Oil Colors, 

Roehrig's Photo Oil Colours, or a combination of artists' oils and megilp. 

The 1930s were also a period in which photographic developments expanded 

into an architectural context. Articles in architecture journals noted that photographic 

murals could be realized through the use of photo-mural paper or 

the spray application of light sensitive emulsions directly onto architectural 

surfaces (11, 12). The notion that a variety of materials could be employed 

as supports fo r photographic images was also reaffirmed by new developments 

in industry. The Glenn L. Martin Company of Baltimore, for example, reduced 

aircraft production time by developing an emulsion that enabled the 

full-scale reproduction of designs on aircraft materials (13). A commercial 

form of this emulsion was marketed following World War II (14). Articles 

outlining the use of such materials frequently noted the need for preparatory 

and protective layers, should the images be colored with oils or other media 

(15, 16). 

Mter 1950 

At the middle of this century, painters' attitudes toward the photographic 

image began to change. Manifestations of this attitude change included the 

Photo-Realist movement and Robert Rauschenberg's and Andy Warhol's use 

of photomechanical processes. The American artist James Couper, who has 

made use of 3M's Scanamural process (a computerized spray painting technique) 

to generate large-scale underdrawings fo r his paintings, has stated that 

he was drawn to this technique after learning how some Photo-Realist painters 

transferred images to canvas through an emulsion process (17). Studies 

published in the United States on the work of such painters, though, have 

contradicted this view. Patton, for example, has held that the intention of 

these painters was to comment on photography, not include it, and interviews 

conducted with a number of these artists have also not confirmed the use of 

any light sensitive coatings (18, 19). 

This change in attitude toward the photographic image was also affected by 

a new interest in early and alternative processes that occurred during the 

192 


1960s in North America. Requiring both a hands-on approach and minimal 

resources, these processes appealed to those who felt that the elements of 

urban life, such as high technology and mass-produced goods, were worthy 

of rejection. Interest in these processes became widespread and resulted in a 

number of publications that provided information on various techniques (20, 

21, 22, 23). 

Commercial products, such as emulsion-coated canvas and ready-to-use 

emulsion, became more widely available at about the same time. Argenta, 

which operated in Munich, Germany, developed an emulsion-coated canvas 

in the late 1950s and early 1960s (24). Designed for theater use, advertising 

purposes, and the reproduction of stitches, this Photoleinen, or photo linen, 

was first marketed in 1962. The product was actually made from a cotton 

fabric prepared with a pigmented gelatin layer. The pigmenting agent comprised 

a mixture of baryta sulfate and titanium oxide. The photographic 

emulsion applied to the gelatin-coated fabric was the same as that used in 

the manufacture of baryta papers of medium gradation. It was also unwashed, 

that is, all superfluous salts were left in the emulsion. A 1991 product list from 

Luminos Photo Corporation of the United States listed the availability of 

sheet and roll forms of a similar product. The sheet form had been impregnated 

with a bromide emulsion, and the roll form with a chlorobromide 

emulsion (25). 

New lines of ready-to-apply photographic emulsions also became, and continue 

to be, available. One version, developed by Argenta of Munich, was 

described as suitable for most surfaces (26). Poor adhesion was to be remedied 

with a preparatory layer of varnish; metallic surfaces were to be precoated 

with gelatin. Print-E-Mulsion, a version developed in the United States, first 

appeared in the mid-1970s (27). At the end of the 1970s, the name was 

changed to Liquid Light. 

Contemporary painters 

Several contemporary painters have employed such materials in their work. 

The American artist Lynton Wells generally used a canvas manufactured by 

Argenta. This was used in the production of paintings and sculptures from 

the late 1960s until about 1983. In some of these paintings, a single image 

spanned multiple panels, where the total length could exceed four meters. To 

create these works, the artist tacked the photo linen to the studio wall and 

exposed the material in situ. To ensure that the image was properly aligned 

when the canvases were stretched, the tacking margin of each section was 

carefully folded under, before the edges of the canvas were put in contact. 

Processing was done in homemade developing trays and the images finished 

with one or more media. Oils, acrylics, aniline dyes, pastels, and charcoal were 

applied in varying densities. Frequently, these additions mimicked elements 

present in the photographic portion of the image. Wells also normally applied 

two to three layers of an acrylic polymer before painting with oils (28). 

Other painters who have made use of photographic bases in their work have 

included Arnulf Rainier of Austria, Shirley Wiitasalo and Kathleen Vaughan 

of Canada, Anselm Kiefer of Germany, and Fariba Hajamadi and James Turrell 

of the United States. From the late 1960s through the 1970s, Arnulf Rainier 

used photographs, photographs mounted onto wood or aluminum, and photo 

linen (29). The range of media employed in these works included oils, oil 

crayon, pencil, and ink. Portions of the photographic image remain visible in 

many of his works, but in some it has been completely negated by the thick 

application of paint. 

Figure 1. Shirley Wiitasalo, Interior, 

1981. Oil and photo emulsion on canvas. 

Photograph courtesy of National Gallery of 

Canada, Ottawa. 

Shirley Wiitasalo created the 1981 painting Interior by applying Liquid Light 

over a stretched cotton canvas prepared with an acrylic gesso (Fig. 1). Exposure 

was carried out using a slide projector. The image was then finished 

with Bellini oils and left unvarnished (30). As one of a series of paintings 

dominated by the television screen and featuring ambiguous and distorted 

Gayer 193

imagery, it was the only work in which a photographic emulsion had been 

employed (31). 

Kathleen Vaughan's painting technique has also involved the use of Liquid 

Light. In some of her work, the emulsion was applied directly to the canvas 

and processed by sponging chemicals over the support. Areas of the canvas 

that the artist intended to appear similar to the photographic portions were 

stained with a distemper medium. These layers were then isolated with a 

coating of an acrylic varnish; texture was applied with acrylic media and the 

final details in oil. Due to allergies, the artist has only used linseed oil to thin 

her oil paints (32). 

Anselm Kiefer has become well known for his robust paintings in which 

photographic enlargements, mounted on canvas, support oil, acrylic, shellac, 

straw, sand, and lead additions. Kiefer's photographic underdrawings have been 

printed on Dokumentenpapier P 90. The harsh or catastrophic appearance 

of these images was created through the manipulation of lighting or processing 

methods (33). Although the photographic image remains visible in 

some of the paintings, a number of them have been completely overpainted. 

In such cases, only the excess of support material on the reverse of the stretcher 

may provide evidence as to the presence of a photographic substrate (34). 

Since the late 1980s, Fariba Hajamadi, an Iranian-born painter living in New 

York, has combined paint and photographic images to produce fictitious interiors 

(35). These paintings have been made by brushing a commercially 

available photographic emulsion directly onto cotton canvas or wood. This 

method was selected to enable the texture or grain of the support to contribute 

to the composition. Following exposure and processing, an airbrush 

was used to apply color. The paint, a commercially available transparent oil 

paint, had been thinned to the appropriate consistency with lacquer thinner 

(36). 

In the planning for his "Roden Crater Project," James Turrell has produced 

studies on frosted drafting mylar in which wax, photographic emulsion, and 

various paint and graphic media have been combined (37). These studies were 

created in the following way: A coating of hot beeswax was first sprayed onto 

the mylar. This wax layer was then coated with a photographic emulsion and 

the desired image of the crater exposed and processed. The image was then 

manipulated and elements may have been removed with an eraser or by 

scraping with a knife. Frequently, wax pastels were used to replace removed 

portions, make additions, or enhance or blur particular details. The colors 

used were carefully chosen so that the additions might stand out or coalesce 

with the existing image. In some cases, a type of sandwich was made by dry 

mounting drafting vellum to the emulsion-coated surface. Further details were 

then added to the front or back of these studies with ink, paint, graphite, or 

wax pastels (38). 

Conclusion 

Painting and photography are techniques that have frequently been used in 

conjunction with each other. For the most part, this relationship has been 

based on visual and intellectual concerns, not the material union of media. 

This paper has attempted to illustrate that although painters may not always 

have utilized photographic materials in their work, information regarding this 

possibility remained available and contributed to related and/ or contemporary 

forms of art production. The genre of traditional portraiture was excluded 

from the discussion; while some portraitists may have utilized such materials 

in the past, recent examples of this were not found. 

Notes 

1. Ruggles, M. 1983. A study paper concerning oil paintings on a photographic 

base. AIC Preprints, 104-12. 

194 


2. Eder, ]. M. 1945. History of Photography. New York: Columbia University Press, 

791, note 4. 

3. Coe, B., and M. Haworth-Booth. 1983. A Guide to Early Photographic Processes. 

London: Hurtwood Press, 19. 

4. Gernsheim, H., and A. Gernsheim. 1969. The History of Photography. London: 

Thames and Hudson, 315. 

5. Joe,]. 1894. Coloring photographs. Scientijic American 71 (11):163. 

6. Scientijic American. 1895. Coloring photographs. Scientijic American 73 (1):27. 

7. Scientijic American. 1899. Coloring bromide prints. Scientijic American 81 (21):323. 

8. Godbold, A. V 1919. A wax medium for the permanent coloring of photographs. 

Scientijic American Supplement 87 (2248):74-75. 

9. Tobias, ]. C. 1934. The art of coloring photographic prints in transparent water-color, 

tempera, opaque and transparent pastel, wax crayons, opaque and transparent oils, chemical 

coloring, and coloring lantern slides. Boston: American Photographic Publishing 

Co., 54, 57-58, 60-61. 

10. Johnson, R. 1936. The art of retouching photographic negatives and photographic enlargements, 

etc. Revised and rewritten by T. S. Bruce and A. Braithwaite, 13th ed. 

(2nd American, revised and enlarged by A. Hammond. Boston: American Photographic 

Publishing Co., 143, 148-150, 152. 

11. Duryea, D. 1936. Notes on murals by photography. Architectural Record 80 (1):57- 

58, 60, 62. 

12. Architectural Review. 1937. The true photo-mural: a new technique of decoration. 

Architectural Review 81 (February) :86-87. 

13. Perry, H. W 1940. Photography speeds up aircraft production. American Photography 

34 (12): 886, 888. 

14. Progressive Architecture. 1946. Photographic emulsion makes any base possible. Progressive 

Architecture 27 (110):110. 

15. Duryea, op. cit., 60. 

16. Hochman, L. 1947. Print your photos anywhere. Popular Science 150 Oune):196. 

17. Corbino, M. 1986. James Couper. American Artist 50 (529):66, 68-69, 84-86. 

18. Patton, P. 1976. Super realism: a critical anthology. Arifo rum 14 (5):52. 

19. Chase, L., N. Foote, and T. McBurnett. 1972. The Photo-Realists: twelve interviews. 

Art in America 60 (6):73-89. 

20. Wade, K. E. 1976. Alternative Photographic Processes. Dobbs Ferry, New York: Morgan 

and Morgan, Inc. 

21. Howell-Koehler, N. 1980. Photo Art Processes. Worcester, Massachusetts: Davis 

Publications, Inc. 

22. Eastman Kodak Co. 1982. Photographic sensitizer for cloth and paper. Photo 

information bulletin PIB-6. December 1982. 

23. Eastman Kodak Co. 1983. Making a photographic emulsion. Photo information 

bulletin PIB-7. February 1983. 

24. Hegenbart, W 1994. Personal communication. Division Manager, Finishing Europe, 

Tetenal Vertriebs GmbH, Norderstedt, Germany. 

25. Luminos Photo Corp. 1991. B & W Products fo r the Professional. Yonkers, New 

York 10705. 

26. Hegenbart, op. cit. 

27. Cone, R. 1993. Personal communication. Rockland Colloid Corp., Piermont, 

New York 10968. 

28. Wells, L. 1991. Personal communication. New York, New York. 

29. Fuchs, R. H. 1989. Arnulf Rainer. Vienna: ARGE Gabriele Wimmer & John 

Sailer, 147-48. 

30. Gagnier, R. 1991. Personal communication. Conservator of Contemporary Art, 

National Gallery of Canada, Ottawa. 

31. Monk, P. 1987. Shirley Wiitasalo. Toronto: Art Gallery of Ontario, 13-14. 

32. Vaughan, K. 1993. Personal communication. Toronto, Ontario. 

33. Winter, P. 1989. Manipulieren, ironisieren, maltratieren: Das Foto unter der Hand 

des Kiinstlers. Kunstwerk 42 (March): 11. 

34. Penn, S. P. 1993. Personal communication. Associate Conservator of Paintings, 

Philadelphia Museum of Art, Philadelphia, Pennsylvania 19101. 

35. Cotter, H. 1990. Fariba Hajamadi at Christine Burgin. Art in America 78 (7):169. 

36. Hajamadi, F. 1991. Personal communication. New York, New York. 

37. Adcock, C. 1990. James Turrell: The Art if Light and Space. Berkeley and Los 

Angeles: University of California Press, 187-96. 

38. Turrell,]. 1993. Personal communication. Flagstaff, Arizona. 

Gayer 195

Abstract 

It is argued here that painters of the 

Baroque adhered to the die-hard tradition 

of loading their palettes with 

a limited number of tints, suitable 

only for painting the passage they 

planned to finish in that stage of the 

work. Support for this proposition 

comes from various directions: written 

sources, studio representations, 

and scientific research generated by 

different methods. An example of a 

specific studio practice is used to 

demonstrate the much discussed interrelation 

of technique and style. 

Reflections on the Relation between Te chnique 

and Style: The Use of the Palette by the 

Seventeenth-Century Painter 

Ernst van de Wetering 

Kunsthistorisch Instituut 

Herengracht 286 

1016 BX Amsterdam 


Introduction 

In the Vitae, as is widely known, Giorgio Vasari attempted to describe the 

development of Italian art from Giotto onward as a process of continuous 

progress culminating in the work of Michelangelo. The idea that art changes 

because better solutions appear was certainly not restricted to Vasari. While 

stylistic developments in Western art were viewed as a matter of progress, it 

was inevitable that the characteristics of earlier styles would be explained in 

relation to problems that had meanwhile been solved. This could apply to 

the invention of perspective and to the invention of oil painting-which, 

according to Vasari, "softens and sweetens the colors and renders them more 

delicate and more easily blended than do the other mediums" (1). Finally, the 

development of style can be applied within oil painting to the development 

of the technique fo r achieving a "'glowing" incarnate, as Karel van Mander 

termed it, by means of an underpainting in vermilion that "glows more 

fleshy" (2). 

Viewed thus, there can be said to exist a clear relation between style and 

technique. During the nineteenth century, however, this was a hotly debated 

issue among art theorists. Some of these, especially writers on architecture 

such as Gottfried Semper (1803-1 879) and Viol1et-le-Duc (1814-1 879), 

aimed to demonstrate a fu ndamental interrelation of style and technique in 

the arts (3). Twentieth-century developments, such as the theories underlying 

the Bauhaus, continued to build on the same ideas. 

Not everyone agrees that style and technique are interrelated. A parallel 

stream of art history adheres to the idea, current since the Romantic period, 

that every fo rm of art had its own fo rmal legitimacy. In this way of thinking, 

it was not necessary to explain styles in terms of technical limitations and 

possibilities. This line of thought culminated in the concept of Kunstwollen, 

originated by Alois Riegl (1858-1 905), which was seen as the manifestation 

of an "urge to fo rm," independent of the restrictive influence of such factors 

as function, materials, and technique. All the same, Riegl did not deny the 

influence of technique. He believed, however, that Kunstwollen overcame the 

technical limitations. Technical frontiers, considered by Semper to play a positive 

part in the creative process, constituted in Riegl's view a "coefficient of 

friction" within the Gesamtprodukt of Kunstwollen (4). 

It could be argued that Riegl's notion of Kunstwollen was partly responsible 

for the fact that "style" has so long remained one of the main domains of art 

historical research, with art historians such as Wollflin and Focillon being 

prominent representatives of this direction. It is worth noting here that the 

stylistics ofWollflin were dominated by an outlook in which the autonomous 

or even abstract qualities of the visual vocabulary took priority over the 

pictorial means employed to achieve a convincing representation of reality. 

Since Wollflin's times, research into artistic techniques has become more and 

more detached from stylistic considerations. Owing to the shift towards sci- 

This article was previously published, in a different form, in Oud Holland 107: 1, 1993, 

137-51; and in KM, vakil'!for111atie voor beeldende Kunstenaars en res fa rato ren, Fall 1994, 

28-31; reprinted, with changes, by permission of the author. 

196 


Figure 1. Jan Baptiste Collaert, Color Olivi, 1566- 1628. After Joharll1es Stradanus. Courtesy of 

the Rij"ksmuseum-Stichting, Amsterdam. 

entific research methods, research into historic techniques has now come to 

play a central part in research on authenticity and workshop practice, as well 

as in conservation and restoration. 

Research into historic techniques has made considerable progress in these 

areas during recent decades, so it now makes sense for current applications 

and future research to turn renewed attention to the interrelation between 

technique and style. This interrelation is examined here through the study of 

developments affecting the painter's palette. 

Palettes 

One of the most fascinating and complete documents concerning the history 

of the art of oil painting is a well-known engraving by Jan Baptist Collaert, 

after Stradanus, which dates from the end of the sixteenth century (Fig. 1). 

This engraving gives a highly detailed picture of an idealized painter's studio: 

the master is working on a history piece, while an assistant is occupied with 

painting a portrait. Two other assistants are grinding and preparing colors. 

The engraving shows countless details that provide valuable hints about dayto-day 

practice in the late-sixteenth-century painter's studio. In the fo reground, 

three boy apprentices can be seen; the smallest is practicing the 

rudiments of drawing, and the more advanced apprentice on the left is drawing 

from plaster casts. 

For our present purposes, we are solely concerned with the apron-clad apprentice 

standing next to the master. He is setting out a palette of small shells, 

presumably containing colors prepared by the assistants, and holds a palette 

similar to that of the master. As it will soon be made clear, the arrangement 

of the master's palette with so few colors is far from arbitrary. Like so many 

other details, this must be regarded as a faithful representation of sixteenthcentury 

practice. The youth in the foreground has applied a limited number 

of colors to the palette in his hand. As on his master's palette, they are spread 

out over the surface of the palette. 

There are generally no written sources to be fo und on the most routine 

activities of the painter's studio through history. Incidental evidence on certain 

aspects of painterly practice can be gleaned from documents, but we have to 

live with the fact that such sources are extremely sporadic in both time and 

place. In the present case, a late-seventeenth-century Italian text can be shown 

van de Wetering 197

to relate to the activities depicted on the previously mentioned engraving by 

Collaert, which originated a hundred years earlier in the Netherlands. Using 

available sources in this way is, of course, only justifiable when the phenomenon 

being investigated is widespread and displays a certain constancy. The 

considerable mobility of painters in the sixteenth and seventeenth centuries 

provided many opportunities for the spreading of painterly techniques and 

procedures. Many young painters traveled across Europe and worked in the 

studios of various masters for shorter or longer periods; a similar situation still 

exists in the international art restorers' world. This form of mobility ensured 

a rapid dissemination of knowledge and experience, leading to a high level 

of international uniformity in knowledge and craft practices. Bearing this in 

mind, one might venture a cautious guess that the question of what the youth 

with the apron is doing in the studio of Stradanus could be answered by a 

passage from the Volpato Manuscript, a seventeenth-century Italian document. 

This text by Giovanni Battista Volpato (born in 1633) must date from 

somewhere around 1680, and was written in the fo rm of a series of dialogues. 

It contains the following exchange between F., an older painter's apprentice, 

and Silvio, a younger apprentice (5): 

Silvio: Tell me, if you will, whether you set your master's palette. 

F: Surely . ... It suffices Jor him to tell me what he intends to paint, Jor 

I then know which colors I must place on the palette. 

The engraving after Stradanus and this late-seventeenth-century text provide 

two of the few hints-which until now have not been given any consideration 

in the literature on art history and painting techniques-that painters 

formerly used palettes which were set with groups of colors specifically for 

certain parts of the painting, and which thus did not include all the available 

pigments. This understanding of the situation has a far-reaching consequence: 

one must then see the seventeenth-century (but also earlier or later) painting 

as a composite image made up of interlocking passages, comparable to the 

giornate, the successively executed "daily portions," of fresco painting, although 

in the case of oil painting, a number of passages would have generally been 

completed on a given day (6). 

In using such a method, the seventeenth-century way of painting differed 

fu ndamentally from the approach of late nineteenth- and twentieth-century 

artists, namely in developing the painting as a tonal entity. For example, the 

Hague School painter Jozef IsraeIs used a palette with a full range of colors 

and with a mixing area covered with patches of mixed paint, of which the 

tone and color could be further modified. A palette of this kind enabled the 

artist to continue working over the whole area of the painting simultaneously, 

with an eye to controlling the tonal consistency of the painting. 

In the first instance, the idea of earlier artists working in giornate as described 

above may sound highly exaggerated. After all, we know that the great majority 

of seventeenth-century painters did in fact conceive their painting as 

a tonal unity, as is evidenced by the practice of starting with a largely monochromatic 

or "dead color" underpainting (7) . The point of importance here 

is that after laying in the underpainting, the artist developed the composition 

further by successively adding islands of modulated local color. Once this is 

understood, it becomes clear that a painting such as Rembrandt's Jewish Bride 

has far more in common with work of predecessors such as Raphael than 

with the paintings of Jozef IsraeJs and his contemporaries, however much 

inspiration IsraeIs drew from that painting. 

A method such as that described is intimately connected with the material, 

technical, and economic constraints that were inherent to oil painting, constraints 

that only vanished (and were subsequently forgotten) with the introduction 

of ready-to-use, "mutually compatible" tube colors (8). 

The early palettes were small. Only in the course of the nineteenth century 

did they grow to the size of small tabletops. It will be clear that this relatively 

198 


sudden increase in the dimensions of the palette is part of the argument put 

forth here (9). 

When one studies the distribution of various tints on the countless palettes 

that appear in paintings, it is impossible to avoid the impression that the 

painters adhered strictly to a certain set of rules. This is certainly true for the 

period after about 1600. Prior to that time, the palette was ostensibly set with 

relative freedom. But it is precisely in those earliest representations of palettes 

that it is sometimes obvious that the palettes were set up specifically for each 

passage to be painted. 

Niklaus Deutsch's painting St. Luke Painting the Madonna (1515) is typical of 

a group of paintings completed in the fifteenth and sixteenth centuries, in 

which St. Luke is in the process of painting the Madonna's robe. The depicted 

artists' palettes show a limited number of patches of paint, with various shades 

of blue as well as black and a little white. These are exactly the colors needed 

to render the modeling in the drapery of the blue robe-distributed apparently 

at random over the surface of the palette (10). In paintings of the same 

period in which St. Luke is painting the naked Christ Child or the face of 

Maria, the palette carries the range of colors needed to mix the various tints 

of the flesh: white, yellow ochre, vermilion, red lake, various browns, black, 

and sometimes terre verde (11). 

As explained in the following discussion, the flesh tint, like blue, had an 

important status. The colors for painting the human skin were not yet systematically 

arranged on the palette before 1600, but appear to be distributed 

at random. For example, the position of the white paint differs from one case 

to another on earlier palettes. From 1600 onward, studio scenes and self portraits 

depict palettes with a row of lumps of paint spaced evenly along the 

top edge. The range of colors depicted normally runs from a somewhat larger 

portion of white near the thumb, to yellow ochre, vermilion, red lake, and 

then through a series of progressively darker browns to black. Alternately, the 

vermilion is sometimes placed between the white and the thumb. This arrangement 

agrees with a passage in the Mayerne Manuscript: "It is to be 

observed, moreover, that in setting the palette, the lightest tints must always, 

without exception, be placed at the top and the darker tints lower down" 

(12). It is striking, and very significant in the context of this article, that the 

author has found no instance of the depicted standard palettes from either 

before 1600 or afterward that includes any intense green, bright yellow, or 

blue paint. These are what are called, in the Mayerne Manuscript of 1630, 

"the strong colors" (13). The reason for the absence of these colors is clarified 

by the following. 

In the Volpato Manuscript, the older apprentice tells the younger that his 

master merely has to indicate what passage is to be painted in order to lay 

out an appropriate palette. This statement implies the existence of fixed recipes 

for reproducing the various elements of nature. Around the same time, 

the Dutch painter Willem Beurs wrote down such recipes for the benefit of 

both "students of the Noble Painterly Art" and interested "amateurs" (14). 

As an example, a recipe for painting a white horse follows (15): 

One paints the illuminated side using white, light ochre and black, with 

pure white for the highlights; light ochre is recommended for the intermediate 

color, and it is advisable there to be rather sparing with white. For the 

shadow, black and light ochre must be mixed together with a little white; 

the niflection under the belly should be mostly light ochre, with sparing use 

of black and white. The hooves can sometimes be painted with black, white 

and light ochre, with a touch of vermilion; and sometimes with black, white, 

and umber. The color of the nose is the same as that of the hoofs. But as 

fo r the eyes: the pupil should be painted with bone black, and the rest with 

umber, black and white. 

It is clear that if an artist plans to paint a white horse, the only pigments 

needed on the palette are white lead, yellow ochre, vermilion, umber, and 

van de Wetering 199

one black; and if the painter expects painting the horse to be a day's work, 

no other colors need be prepared for that day, apart from the five pigments 

just mentioned. 

It is notable that throughout hundreds of recipes of this kind, only a severely 

limited range of pigments is prescribed. This provides a piece of information 

that is of critical importance to the picture of the artists' practice being 

sketched here: paint was only ground and prepared when it was needed. It is 

precisely because painters wished to keep working without unnecessary delays 

and therefore found it advantageous to use paint that dried quickly (thus 

containing strong drying oils and other drying agents), that paint could not 

be kept for long. If the complete range of pigments, each already ground 

with oil, had to be available, this would mean that much paint would have 

to be thrown away unused. Hence the economic reasons for the method of 

working with restricted, specific palettes are clear. The technical background 

of this method will be discussed briefly in this paper. 

It is significant that the series of recipes recorded by Willem Beurs concludes 

with flesh colors (16). Beurs writes as follows: "Just as we humans consider 

ourselves the foremost among animals; so, too, are we the foremost subject of 

the art of painting, and it is in painting human flesh that its highest achievements 

are to be seen" (17). The palette Beurs gives fo r painting human flesh 

comprised mixtures of the pigments lead white, light ochre, schijtgeel (an organic 

yellow), vermilion, red lake, tawny ochre, terre verde, umber, and "coal 

black" (probably ground charcoal which gives a bluish black). It was the 

palette for what was considered to be the summit of creation, and the most 

difficult subject of all to paint, the human figure. It is almost invariably the 

palette for flesh colors that is depicted in self portraits and studio scenes after 

1600, paintings that are generally intended to represent the art of painting at 

its noblest. 

It will come as no surprise to anyone who has examined fifteenth- or sixteenth-century 

paintings to any depth as material objects, that every passage 

was executed as a separate entity. The additive character of the painting as a 

whole is generally plain to see, despite the psychological compulsion perceptually 

acting on the viewer to transform the painting into a Gestalt. 

Even though revolutionary developments during the seventeenth century 

brought the pursuit of pictorial unity to an unprecedented level, the sources 

quoted above suggest that no change had taken place in the tradition of using 

recipes for various components of the painting. The economic rationale for 

this approach has already been mentioned above. Originally, however, technical 

reasons may have provided an even stronger motivation for executing a 

painting as a series of successive passages. 

For the twentieth-century painter, who normally regards paint as a pasty 

substance of a certain color that can be squeezed out of a tube, it is hard to 

imagine that to artists of not only the fifteenth and sixteenth centuries, but 

also through the first half of the nineteenth century, each pigment presented 

its own inherent possibilities and constraints (18). Some pigments could not 

be worked up with oil; some pigments could only safely be mixed with one 

or two other pigments; some pigments could only be used transparently and 

yet others only opaquely. Other properties, too, such as color permanence, 

workability, drying qualities, and so on, could differ so strongly from one 

pigment to another that it was normal to use a given pigment either in pure 

form or mixed with any of a limited number of other pigments in order to 

somewhat modifY the tone and color (19). This helps explain why, in the 

work of artists such as van Eyck or Lucas van Leyden, the colors unmistakably 

interlock like the pieces of a jigsaw puzzle, and each color has an individual 

character, especially as to transparency, surface texture, and thickness of the 

paint layer. The most easily workable pigments were the earth colors, which 

ranged from yellow ochre through red ochres to the darkest brown tints and 

were varied in tone by mixing together and by the addition of white or 

200 


lacks. Passages painted with mixtures of these pigments are subtle in their 

tonal values and graduations. Whenever it was necessary to achieve strong, 

bright colors (i.e., red, yellow, and blue robes), it is clear that the passage 

concerned was executed within carefully delineated contours in accordance 

with a fixed recipe, involving a specific layering or a fixed type of underpainting. 

This also explains why these colors are usually absent from fleshcolor 

palettes depicted in studio scenes and self-portraits. 

Figure 2. Neutronactivation autoradiographic 

image of Rembrandt's Bellona, 1633. The 

Metropolitan Museum of Art, New York. In 

this image the activated phosphor atoms (from 

bone black) and mercury atoms (from vermilion, 

used on the figure'S mouth) have mainly 

blackened the film. 

Figure 3. The image of the autoradiograph 

in Fig. 2 is mainly determined by the radiation 

from the copper atoms (from the bluegreen 

parts in the painting) and again the 

mercury of the vermilion. 

That artists worked in this way during the fifteenth and sixteenth centuries 

can easily be discerned from the paintings themselves. But did this practice 

continue into the Baroque period? In the case of Rembrandt, this is by no 

means self-evident. There were already indications, however, that Rembrandt 

must have used this method; it has been observed that he completed his 

paintings systematically from background to foreground, passage by passage, 

on the basis of a monochromatic underpainting (20). This does not necessarily 

imply, however, that he did this using a restricted palette. An important piece 

of evidence for that practice is supplied by scientific studies. It was the failure 

of a reconstruction project that, in fact, formed the seed of the study presented 

here. The preparation of a dummy Rembrandt using all the procedures of 

Rembrandt's studio that were known at the time was unsuccessful because 

the painting was executed using a complete palette. All the pigments that 

were in use during the seventeenth century were present on the palette and 

they were mixed on the same mixing surface, which was cleaned at intervals. 

Autoradiographic investigation of the dummy made it clear that the procedure 

used resulted in all the pigments being present to a greater or lesser 

degree all over the painting. By contrast, autoradiography of Rembrandt's 

actual paintings resulted in surprisingly "clean" images (Figs. 2, 3) (21). 

In Rembrandt's painting, certain pigments occur only within clearly demarcated 

zones, not in the rest of the painting. This observation is additionally 

confirmed by studies of paint samples from Rembrandt's paintings, which 

time and again reveal that only a limited number of pigments were used in 

a given passage. The mixtures found usually consist of two to four different 

pigments; mixtures of five-or, in very exceptional cases, six-pigments are 

found only incidentally (e.g., in flesh passages). To achieve the typically Baroque 

tonal unity using such methods implies a highly developed level of 

"management" in the use of colors and tones. It is therefore no coincidence 

that artists began theorizing about the nature of this "management" during 

the seventeenth century. This was made clear by Paul Taylor's study of the 

term houding (roughly, "disposition") that appears regularly in seventeenthcentury 

sources (22). Until some time into the sixteenth century, the picture 

space was still structured simply as a foreground with a background or campo, 

a concept on which Jeroen Stumpel performed an important study (23). The 

idea denoted by the term houding was a far more complex one, however. 

Seventeenth-century writers used houding to denote the spatial coherence 

created in the painting by the disposition of tones and colors. The viewer 

could "walk" through this space in his imagination. In a formulation by Willem 

Goeree (1668), houding is (24): 

. . . that which binds everything together in a Drawing or Painting, which 

makes things move to the front or back, and which causes everything from 

the foreground to the middle ground and thence to the background to stand 

in its proper place without appearing fu rther away or closer, and without 

seeming lighter or darker, than its distance warrants; so that everything 

stands out, without confusion, from the things that adjoin and surround 

it, and has an unambiguous position through the proper use of size and 

color, and light and shadow; and so that the eye can naturally perceive the 

intervening space, that distance between the bodies which is lift open and 

empty, both near and far, as though one might go there on foot, and 

everything stands in its proper place therein. 

When one considers how a painter would have been intensively concerned 

with obtaining a good houding while at the same time working with selective 

van de Wetering 201

palettes as described above, it is clear that the level of accomplishment required 

was of a quite different nature to the methods of painters from the 

later nineteenth and twentieth centuries, who, thanks to the full palette, could 

now work on the whole painting at once. This fact clearly explains the somewhat 

"muddy" effect, compared to earlier paintings, characteristic of Jozef 

Israels and many of his contemporaries and successors. 

Research into the borderline region between style and technique throws light 

on unexpected aspects of seventeenth-century studio practice. It is clear, for 

instance, that the view on the genesis of paintings presented here must have 

implications for scientific research into Baroque paintings. Analytical studies 

of the binding media used in the seventeenth century (e.g., by Rembrandt) 

suggest that a different blend of media is likely to be found in every giornate 

of the painting (25). 

This study has produced a picture that may be relevant to the discussion of 

the interrelation of style and technique, namely that of a development of 

Baroque painting in which artists strove for a unified tonal image that did 

not appear to be composed by additive methods. The fact that these additive 

techniques were used gives a different picture of the artist at work, one in 

which technical limitations actually constituted the "coefficient of friction" 

(to use Riegl's term) that thwarted the artists in their pursuit of a new style. 

It was even necessary for the seventeenth-century painters to introduce a new 

art-theoretical concept, namely houding, to bring the discussion of these efforts 

to a new level of abstraction. Only in the nineteenth century did the 

technical materials and means arise-namely, the full color palette, paint in 

tubes, and the associated large palette sizes-that made it possible to apply a 

technique appropriate to the painter's stylistic aspirations (as exemplified by 

Cezanne's remark, "You must understand that I handle the whole painting at 

once, in its totality") (26). Semper appears to be borne out, on the other 

hand, if one observes that a late Rembrandt, placed between a Raphael and 

a nineteenth-century piece of "Rembrandtism" such as a late work by Jozef 

Israels, shows closer kinship to the Raphael than to the Israels. Indeed, the 

technique of additive painting has an unmistakable determinative effect on 

the style of a painting. 

Aclrnowledgll1ents 

Thanks are hereby expressed to Jos Koldeweij , Hayo Menso de Boer, Paul Broeckhof, 

Dieuwertje Dekkers, Karin Groen, and Charlie Srnid. Much of the research underlying 

this study was carried out within the framework of the Rembrandt Research 

Project, financed by the Netherlands Organization fo r Scientific Research, and also 

received valuable support from the Director and the Head of the Art History Department 

of the Central Laboratory for Research on Objects of Art and Science, 

Amsterdam. 

Notes 

1. Vasari on technique. 1960. Ed. B. Brown. New York, 230. 

2. van Mander, K. 1604. Den grondt der ede! vry Schilderconst. Haarlem, fol. 64r. P 

Taylor of the Warburg Institute, London, is carrying out research on the concept 

"glowing" in connection with the painting of flesh in the early seventeenth 

century. 

3. Semper, G. 1860-1863. Der Stil in den technischen und tektonischen KiJnsten. Frankfurt 

am Main. See also Viollet-Ie-Duc, E. E. 1863-1872. L'entretien sur I'architecture. 

Paris. 

4. Riegl, A. 1893. Stilfragen. Berlin, v-xix, 20, 24. See also Kultermann, U. 1966. 

Geschichte der Kunstgeschichte. Vienna, 288. 

5. Giovanni Battista Volpato: Modo da tener nel dipingere. 1967. In Original Treatises 

on the Arts oj Painting. Ed. M. P Merrifield. New York, 726-55, especially 

746-48. 

6. Mora, P, L. Mora, and P Philippot. 1984. Conservation oj Wall Paintings. London, 

138-64. 

7. van de Wetering, E. 1977. De jonge Rembrandt aan het werk. In Qud Holland 

(91):7-65, especially 20-24. 

202 

Historical Painting Techniques, Maten:als, and Studio Practice

8. It is not clear when people first started keeping prepared paint in small pouches 

made of pig's bladder. The pouches are occasionally illustrated in eighteenthcentury 

sources. According to R. G. Gettens and G. L. Stout (1942. Painting 

Materials: A Short Encyclopedia. New York, 318), the tube came into widespread 

use around 1840. 

9. A description of the evolution of the palette can be fo und in Gettens and Stout, 

op. cit. (note 8), 299-304. 

10. Colyn de Coter, St. Luke Painting the Madonna, Vievre, Alliers (near Moulins), 

ca. 1490; de Meester van het Augustiner Altaar, St. Luke Painting the Madonna, 

Germanisches Nationalmuseum Niirnberg, 1487; Nicolaus Manuel Deutsch, St. 

Luke Painting the Madonna, Bern Kunstmuseum, 1514-1515; Unknown Flemish 

Master, St. Luke Painting the Madonna, the Grimani Brevarium, fol. 781 v, Bibliotheca 

Marciana, Venice. 

11. Hinrik Bornemann, St. Luke Painting the Madonna, Hamburg, Jakobikerk, 1499; 

School of Quinten Massys, St. Luke Painting the Madonna, National Gallery, London; 

Derick Baegert, St. Luke Painting the Madonna, Landesmuseum, Munster, 

1490. Belonging to the same category of palette are: Jacob Cornelisz van Oostzanen, 

Self Portrait of the Artist Working on a Portrait of his Wife, Museum of Art, 

Toledo, ca. 1530; Catharina van Hemessen, Self Portrait, Offentliche Kunstsammlung 

Basel, ca. 1548. 

12. Pierre Lebrun, Recueil des essaies des merveilles de la Peinture. In Original Treatises 

on the Art of Painting, Vol. II. 1967. Ed. M. P. Merrifield. New York, 770- 

71. 

13. Quellen fur Maltechnik wahrend der Renaissance und deren Folgezeit (XVI-XVII Jahrhundert), 

Vol. IV 1901. Ed. E. Berger. Munich, 288-89. 

14. Beurs, W 1698. De Groote Waereld in het Kleen geschildert. Amsterdam. 

15. Beurs, op. cit., 165-66. 

16. Beurs, op. cit., 184-87. 

17. Beurs, op. cit., 186. 

18. This standardization of the working properties of various colors of paint is 

achieved by the addition of fillers such as bentonites or aluminium hydrates, or 

by the use of alternative pigments. Naples yellow, for example, is substituted by 

a mixture of titanium white and cadmium yellow. 

19. See for example the specifications given for the various pigments in a section 

about oil paint in De volmaakte Verwer, 1770. Amsterdam, appendix. 

20. van de Wetering, E. 1977. De jonge Rembrandt aan het werk. In Oud Holland 

91:7-65, especially 20-2. See also van de Wetering, E. 1982. In A corpus of Rembrandt 

Paintings, Vol. I. J. Bruyn, B. Haak, S. H. Levie, P .J. J. van Thiel, E. van de 

Wetering. Den Haag, 25-31. 

21. Ainsworth, M. W et al. 1982. Art and Autoradiography: Insights into the Genesis if 

Paintings by Rembrandt, Van Dijck and Vermeer, New York. For activities in the 

autoradiographic area by the Gemalde-Galerie and the Hahn-Meitner-Instituut 

fUr Kernforschung in Berlin, see MaltechniklRestauro I (1987):21 et seq. 

22. Taylor, P. 1992. Harmony and Illusion in Golden Age Dutch Art Theory: The 

Concept of Houding. In Journal if Warburg and Courtauld Institutes. (LV) :210-32. 

23. Stumpel, J. 1990. The Province of Painting. Theories if Italian Renaissance Art. 

Utrecht, 131-174 (also published in: Simiolus 1989, 19 [3/4]:219-43). 

24. Goeree, W 1670. Inleyding tot d'Algemeene Teykenkonst. (First printing, 1668.) 

Middelburg, 219-43. 

25. c.f. the article by C. M. Groen in the still unpublished publication in connection 

with the restoration of the late Rembrandts in the Rijksmuseum, Amsterdam. 

26. "Je mene comprenez un peu, toute ma toile a la fois, d'ensemble." Gasquet, J. 

1926. Cezanne. Paris, 89. 

van de Wetering 203

Index of Contributors 

J. R. J. van Asperen de Boer, 135 

Sylvana Barrett, 6 

Ulrich Birkmaier, 117 

Barbara H. Berrie, 127 

Leslie A. Carlyle, 1 

Dace Choldere, 155 

Josephine A. Darrah, 70 

Kate I. DufiY, 78 

Jacki A. Elgar, 78 

Molly Faries, 135 

Beate Federspiel, 58 

John R. Gayer, 191 

E. Melanie Gifford, 140 

Helen Glanville, 12 

Stephen Hackney, 186 

Erma Hermens, 48 

Helen C. Howard, 91 

Melissa R. Katz, 158 

Jo Kirby, 166 

Ann Massing, 20 

Catherine A. Metzger, 127 

Ilze Poriete, 155 

Andrea Rothe, 111, 117 

Ashok Roy, 166 

Zuzana Skalova, 85 

Eddie Sinclair, 105 

Christa Steinbiichel, 135 

Dusan C. Stulik, 6 

Joyce H. Townsend, 176 

J0rgen adum, 148 

Arie allert, 38, 117 

Ernst van de etering, 196 

Sally A. 00dcock, 30 

Renate oudhuysen-Keller, 65 

204 

Index oj Contributors

Errata for Historical Painting Techniques PDF 

COVER ILLUSTRATION and PLATE 8 

Gherardo Cibo, “Colchico,” folio 17r of Herbarium, ca. 1570. © British Library Board (ADD MS 22333). 

PLATES 7a, 7b and PAGE 50, Figure 1 

Valerio Mariani de Pesaro, Battaglia di San Fabiano, 1618-1620, GDSU. Images provided by permission 

of the Ministero per i Beni e le Attivita Culturali, Florence, Italy. 

PLATE 12 

Plate printed upside down in relation to Plate 11. 

PLATE 17: (c. 50 x 246.5 x 2.5 cm) 

PLATE 18: (c. 50 x 212 x 2.5 cm) 

PLATE 29e 

Plate printed upside down in relation to Plates 29a-d,f. 

PAGE 52 

Figure 3. Gherardo Cibo, “Hemionite,” folio 143r of Herbarium, ca. 1570. © British Library Board (ADD 

MS 22333). 

PAGE 53 

Figure 4. Gherardo Cibo, “Fusaina [. . .] nocella qui chiamata a Roccha C[ontrada],” folio 183v of 

Herbarium, ca. 1570. © British Library Board (ADD MS 22332). 

PAGE 53, CONT.

Figure 5. Gherardo Cibo, several proofs of colors and “Fusaina” flowers, from folio 184v of Herbarium, 

ca. 1570. © British Library Board (ADD MS 22332). 

PAGE 85 

Paragraph 2. Alfred Butler assessed the beams in 1880, not the 1870s. 

PAGE 86 

Paragraph 4. The first icon, The Virgin Mary and Child enthroned between Archangels, Prophets and 

Egyptian, Greek and Syrian Holy Bishops and Monks, measures c. 50 x 246.5 x 2.5 cm, and is shown in 

Plate 17 and Figures 3 (left) and 4a, b. The second icon, Six Equestrian Saints (originally ten saints), 

measures c. 50 x 212 x 2.5 cm, is shown in Plate 18 and in Figures 2a,b; 3 (right); and 4c-e. 

PAGE 89 

Acknowledgments. The author…has been working since 1989 as the Field Director of the joint Egyptian- 

Netherlands ‘Coptic Icons Conservation Project’ based in the Coptic Museum in Cairo, not as the Field 

Director of the Coptic Museum in Cairo as originally published. 

Note 3. The correct reference: Butler, A. J. 1884. 

PAGE 90 

Note 13. The correct reference: MacCoull, L.S.B. 1993. The Apa Apollos Monastery at Pharoou. In Le 

Muséon (106), note 11. 

Note 20. Abu al-Makarim, a Coptic priest from the church of the Holy Virgin at Harat Zuwayla, Cairo, 

left folios in a manuscript dating from around 1200 known as History of the Churches and Monasteries in 

Lower Egypt in the 13th Century. See Père Samuel du Monastère des Syriens. 1990. Icônes et 

iconographie en Égypte au XIIe siècle d’après le manuscrit d’Abū- El-Makārim, publié en Arabe au Caire 

en 1984. Le Monde Copte, No. 18. Limoges, 78. 

PAGE 115 

Figure 3. Mantegna, Andrea (1431-1506). The Presentation in the Temple. Original strainer with panel 

inserts seen from front. 68.9 x 86.3 cm. Gemaeldegalerie, Staatliche Museen, Berlin, Germany. Photo 

credit: bpk, Berlin/Art Resource, NY. 

PAGE 135 

Lines 9,12, and 14. “Mander” should read “van Mander.” 

PAGE 137 

Line 2. “No ground is used” should read “no ground appears in this sample.” 

PAGE 158 

Figure 1. Image source: Period photo reproduced in O. von Schleinitz, William Holman Hunt (Bielefeld 

and Leipzig: Verlag von Velhagen & Klasing, 1907).

PAGE 159 

Figure 2. Image source: Period photo reproduced in O. von Schleinitz, William Holman Hunt (Bielefeld 

and Leipzig: Verlag von Velhagen & Klasing, 1907). 

PAGE 160 

Figure 5. William Holman Hunt, Autoritratto, Galleria degli Uffizi. Image provided by permission of the 

Ministero per i Beni e le Attivita Culturali, Florence, Italy. 

Figure 6. Unknown Artist. Paint Bladder, 19th century. Oil paint contained in an animal bladder with 

three ivory plugs; 4.45 x 2.54 cm (1 3/4 x 1 in.) Harvard Art Museums/Fogg Museum, Straus Center for 

Conservation. Gift of C. Roberson and Co., London, FINV2155. 

Photo: Imaging Department © President and Fellows of Harvard College 

Brass paint tube, (not assigned). Harvard Art Museums/Fogg Museum, Straus.698 


Empty paint tube, (not assigned). Harvard Art Museums/Fogg Museum, Straus.696 


Figure 8. William Holman Hunt. The Miracle of the Sacred Fire, Church of the Holy Sepulchre, 1892- 

1899. Mixture of oil and resin on canvas; 92.1 x 125.7 cm (36 1/4 x 49 1/2 in.) Framed: 114.5 x 148.3 x 7 

cm (45 1/16 x 58 3/8 x 2 3/4 in.) Harvard Art Museums/Fogg Museum, Gift of Grenville L. Winthrop, 

Class of 1886, 1942.198. 


PAGE 169 

Figure 3. The Execution of Lady Jane Grey. Pencil drawing for the central group, 18 X 16.5 cm. © The 

Trustees of the British Museum. 

PAGE 178 

Figure 2. J. M. W. Turner, Dolbadern Castle, North Wales, 1802. Oil on canvas, 1190 x 889 mm. 

Courtesy of the Royal Academy of Arts, London, www.racollection.org.uk. 

PAGE 182 

Figure 6. Detail of Anna's veil, painted over clouds and sky, from Joshua Reynolds's The Death of 

Dido, ca. 1781. Oil on canvas, 1473 X 2407. Supplied by Royal Collection Trust / © HM Queen 

Elizabeth II 2012. 

PAGE 187 

Figure 3. Brushes, paints, palettes, and other painting materials. © The Hunterian, University of Glasgow 

2012.

Historical Painting Techniques, Materials, and Studio Practice

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?