By using this website you agree to our Cookie policy

March 2018

Vol. 160 / No. 1380

The Modigliani Technical Research Study. An introduction to Modigliani’s materials and techniques

By Annette King, Nancy Ireson, Simonetta Fraquelli and Joyce H. Townsend

Over the next three months, the Burlington will be publishing a series of articles on the Modigliani Technical Research Study (2017–18), which was initiated by Tate during the preparation of its current exhibition on the artist. An exceptional example of international collaboration, more than thirteen institutions have contributed to the research. This introductory article explores the aims of the project and brings together some of its key findings.


WORKING ACROSS SEVERAL media in the course of a short but intense career, Amedeo Modigliani became an accomplished draughtsman, sculptor and painter. From his arrival in Paris in 1906, he was an active participant in the art world, engaging with a broad range of ideas and methods as he developed a style of his own. A variety of influences challenged and built upon his traditional Italian artistic training: the influential Paul Cézanne retrospective at the Salon d’Automne (1907); encounters with ancient Egyptian art in the Louvre and with the Ivory Coast carvings in the collection of his first major dealer, Paul Guillaume; and sustained contact with peers from many different countries, including Pablo Picasso and Constantin Brancusi. He toyed with various stylistic tendencies – from Symbolism to Cubism – before devoting himself primarily to a highly distinctive form of portraiture.

Although accounts of Modigliani’s bohemian lifestyle have often overshadowed his achievements, in the current exhibition at Tate Modern (to 2nd April),1 the artist’s work takes precedence. It includes one hundred works in various media, including seventy-six paintings, which appear in a largely chronological hang, and nine stone heads. The exhibition aligns the evolution of the artist’s mature style with changes in his environment, from the bright colours of his arrival in Paris to his adoption of a Mediterranean palette on his move to southern France. By uniting so many of his paintings the viewer is able to experience their diverse surfaces (so often lost in reproduction), from the more textured images of the war years and the smooth flesh of the nudes to the translucent surfaces of his south of France works. It argues that contemporary culture – including the subjects tackled by early cinema and the changing status of women – shaped the ways in which he represented his sitters.

As part of this commitment to looking at the artist in new ways, and alongside the curatorial development of the exhibition, Tate began technical examination of paintings by Modigliani in its permanent collection. Cognizant of how rarely the artist’s works appear together in such quality and strength, Tate also invited a number of lending institutions to undertake their own investigations into his work. Additional participants were the National Gallery of Art, Washington, and the Barnes Foundation, Philadelphia, where the constraints of donors’ bequests prevent their works by Modigliani from travelling. Our ambition was modest – to develop and encourage scholarship on Modigliani’s materials and methods – with each institution examining its paintings with the techniques at its disposal. These included X-radiography, infra-red reflectography, ultra- violet and visible light photography, surface examination and light microscopy. Where possible, participants undertook paint analysis in order to characterise Modigliani’s pigments and media. The non-destructive analytical technique of X-ray fluorescence (XRF) allowed chemical analysis of pigments in areas that are not suitable for samples to be taken, yielding some unexpected results. In a few cases medium analysis was carried out; this proved particularly interesting when applied to works that belong to different moments of Modigliani’s career. The strength of any such study lies in the number of works examined and the collaborative comparison of results and discussion of inferences. An overview of the project’s findings is outlined below. Four case studies – presented in summary form as articles – will appear in two subsequent issues of this Magazine.

The need for a better understanding of Modigliani as a practitioner is clear, not only because of the controversy that surrounds the question of autograph works,2 but also (and most importantly) because of the lack of technical research into his accepted oeuvre. When work began on the project, the only major scientific study of Modigliani’s painting available, produced by the Laboratoires de Recherche des Musées de France, Paris, was already over thirty-five years old.3 This encompassed the technical study of works in French museums and a few from private collections, and outlined some of the challenges still presented by technical research into Modigliani’s work.4 Although it provided valuable insights and employed techniques that form the basis of our own study, technical imaging has moved on apace since its completion. Tate’s project presents an opportunity to bring these new advances to the subject. This year, in a French state-funded project, the Centre for Research and Restoration of Museums of France (C2RMF) and the Lille Métropole Museum of Modern, Contemporary and Outsider Art (LaM) will also begin technical examinations of Modigliani’s paintings in the national collection, with a view to building on the findings of 1981.5 It is hoped that the results of our respective efforts, when combined, will help to further a clearer understanding of the artist’s process.

Although many collections involved in the present study – including Tate’s own – have works attributed to Modigliani that do not feature in Ambrogio Ceroni’s widely accepted catalogue raisonné of 1970 (see the related article in this issue, pp.189–95), because of limited analysis of the accepted oeuvre, we chose to focus only on works that are listed therein.6 Our project makes no attempt to expand the canon. There is a lack of technical examination of the artist’s work in other media – and a keen need, in particular, for further work on Modigliani’s drawings – but we chose to focus our efforts on paintings.  

To encompass a range of Modigliani’s work – and to facilitate conversation between colleagues – we highlighted four groups of paintings for comparison. They bridge a significant portion of his career in France and follow the chronology established by Ceroni. Notably, we chose not to focus on works made immediately following the artist’s arrival in France, because variations in style and format render their comparison more challenging. Before 1910 Modigliani experimented widely, drawing inspiration from a diverse range of artists including Cézanne, Henri de Toulouse-Lautrec and Kees van Dongen, and changed his technique frequently. Erratic financial support, which also caused him to move studios often, may have led to inconsistency in his choice of art supplies. In the hope of finding patterns in his use of materials, we looked instead to works that are similar in theme and which date from the artist’s more established years.

The first group of works comprises portraits Modigliani made during the First World War, when he lived and socialised mostly in Montparnasse (Fig.1). These tend to be darker in palette than works from other stages of his career. They also engage with cubist tendencies. The second study focuses on the controversial nudes made around 1916–17 (Fig.2), a time when Modigliani enjoyed the support of the Polish dealer Léopold Zborowski, who supplied him with models and materials.7 The third group of paintings dates from the last months of the First World War when, fearful for Modigliani’s safety and health, Zborowski paid for the artist to live and work in the south of France, between Cagnes and Nice. This period saw distinct changes in Modigliani’s work: he adopted brighter colours and applied paint thinly (Fig.3). The final group comprises paintings made in the two years preceding his death. Despite his worsening health, Modigliani continued to work until around a month before he died in January 1920 in Paris.8 His late paintings demonstrate a striking degree of technical confidence (Fig.4). Some may have been made in his final studio, at 8 rue de la Grande-Chaumière, where the artist Nina Hamnett recalled that the walls were painted with blocks of colour to provide backgrounds for portraits.9 Although separate articles on these themes will feature in forthcoming issues of the Burlington Magazine, this introduction outlines some of the overarching discoveries that have emerged from the project and, more generally, from the experience of seeing the wider range of works in the exhibition.

Exceptional institutional collaboration has allowed contributors to compare works unframed. To look at the reverse of a painting in this way can yield valuable information, including whether the strainer dates from the creation of the work and whether the attachment of the canvas is original to the strainer. Sometimes an original support yielded the stamp of a supplier, occasionally with a name or address – helpful insights into where Modigliani sourced his materials.

Although many of the works in this study have been lined (the attachment of a new canvas at some stage to the reverse of the original canvas support, using either wax, glue paste or, more recently, synthetic heat-seal adhesive), a good number remain unlined. Some of these unlined paintings also retain their original stretchers or strainers.10 Of the original secondary supports under examination, most were rudimentary fixed-corner strainers, typical of the time. These simple lap-jointed strainers are held together with three or four nails at the corners and would have been relatively cheap to purchase. French canvas sizes were standardised in the nineteenth century and in the early twentieth were still sold in distinct formats. Suppliers differentiated sizes by subject: a portrait canvas (figure) is longer on the vertical edge than on the horizontal, while a landscape (paysage) is longer on the horizontal edge. A seascape (marine) proportion to one another on the diagonal, so that a size figure 0 is the same height as a paysage 1 or a marine 2. Many of the strainers made by artists’ suppliers would bear a stamp such as 20F (size 20, figure).

Looking at the dimensions of all the paintings in the exhibition that used strainers (so excluding those on board), interesting patterns of use emerged. Of seventy-six paintings, twenty-one were non-standard sizes, eleven were figure format, nine were paysage and thirty-five were marine. More remarkable was the fact that Modigliani used these – mostly standard sized – strainers and stretchers in a highly unconventional manner. Often, he turned paysage and marine strainers on their shorter edges, in order to use them for portraits. Perhaps Modigliani liked the extended height provided when he turned a marine format in this way; it certainly suits the elongated form of many of his portraits. Although there are examples from as early as 1908, Modigliani used marine canvases most often during his year in the south of France, which may mean that, as financial support for his work grew, he was better able to exercise a personal preference. 11 Of the thirty-five marine paintings in the exhibition, only five were used in the horizontal format, on each occasion as the support for a reclining nude.

It is likely that Modigliani acquired materials through his patrons and dealers, Paul Alexandre, Paul Guillaume and Zborowski. Little is known about the exact arrangements between them and the artist and Modigliani has not previously been linked to specific merchants of artists’ materials. This renders the suppliers’ marks on original strainers, or on the backs of unlined canvases, particularly valuable. It is likely that materials came from a variety of sources, probably quite pragmatically, as his workplace changed often. A number of names emerged: ‘A. Moreaux, 106 Bd Montparnasse’, was found on the reverse of the canvas of a caryatid painting from 1913 (private collection; C37), an address near Cité Falguière, where Modigliani was known to have had a studio from 1909 to 1912;12 ‘Ossocoz, 12 rue de Douai’ was found on the reverse of the original strainer on Portrait of a girl (c.1917; Tate, London; C118). Modigliani had probably stayed on the rue de Douai four or five years earlier, in 1911–12. An X-radiograph of this painting revealed an earlier image on this canvas: the discovery that he might have acquired the canvas in 1911–12 could suggest that he made the underlying painting at that time. Modigliani is listed as one of the clients of Lefebvre-Foinet on the corner of 19, rue Vavin, and 2, rue Bréa, which was close to his final studio on the rue de la Grande-Chaumière, although no mark by them on any of his works has yet been discovered;13 Guide Labreuche also lists Modigliani as a client of ‘Sennelier’,14 a supplier whose main address was 3, quai Voltaire, but whose second address – 4, rue de la Grande-Chaumière – was also seconds away from the artist’s last studio.15 There is also a partial canvas stamp on the reverse of Bust of a young girl (1908; private collection; C10) in the shape of a large artist’s palette, but unfortunately it is obscured by the strainer and remains unidentified.

The canvases examined in the study seem to have been commercially prepared with an oil-based priming and could have come from any of the listed suppliers. The weight of the canvas varies throughout the sample of paintings examined, from finely woven to coarse linen and, later, when linen was in short supply after the war, on cotton supports. The contributors sent digital X-radiograph images to Don Johnson at Rice University, Houston, who submitted them to the groundbreaking Thread Count Automation Project.16 By scanning digital X-radiographs, accurate measurements of its thread density in both warp and weft direction can be provided, even if a canvas is lined. In addition to measuring the thread count, the project’s software can record anomalies in weave regularity; it records specific areas of distortion in canvas weaves and compares the lines of tension with those of other canvases to find ‘cliques’ with identical thread counts and areas of disparity. The identification of these clusters depends on comparative data and, where canvases are identified as an exact match, they may well come from the same roll of canvas. Excitingly, within our small sample group, exact matches did occur. We cannot be sure that Modigliani obtained a roll of canvas from his dealers to stretch up himself, rather than buying ready-made supports from a colour merchant, but it now seems likely. Some of the paintings in our study have pinholes in all four corners which suggests that, having cut pieces from a large roll, Modigliani could have painted these canvases pinned to the wall of the studio. If this were the case, only on completion would a work have been attached to a strainer.

The study also revealed that a large number of canvases had painted borders and, in one case, pencil lines beneath the paint that define the limits of the image.17 If Modigliani did cut his canvases from a large roll and stretch them at a later stage, it is possible that he drew the size of the strainer as a guide and painted within the lines. Painted borders over the paint film are not a consistent feature of his works; but they appear often enough to be noteworthy. Several paintings have black outlines, like a painted frame (see Fig.2). This feature spans periods; it appears in certain nudes and in some of the works he made in the south of France. Some outlines are brushy and indistinct, some take the form of solid lines, while others have just one or two edges. Some of the works have an unfinished, uneven lower edge, which leaves an area of ground exposed, suggesting that the artist made them on an easel with a lipped shelf (Fig.3, for example).

This study deployed a range of technical imaging to assess Modigliani’s technique and use of colour. The predominant feature revealed by X-radiography, for example, is the use of lead white,18 which appears dense and white, in the X-radiograph; zinc white, found frequently in early twentieth century paintings, appears as more transparent. This allows the differentiation of passages of white that may appear uniform to the naked eye. Most grounds of the Modigliani paintings examined in the current study contain lead white but in some interesting instances it appears that the artist or colour merchant applied an additional layer of zinc white as priming. X-radiograph images have also proved crucial to showing the characteristics of Modigliani’s handling of paint in our sample group: distinctive stabbing brushstrokes, made with large round brushes; overlapping brushstrokes with square-tipped stiff brushes; textured backgrounds; and paint applied in distinctive zigzig patterns.

X-radiographs also capture an important feature of Modigliani’s method. In the initial stages of making a painting, it seems that after he had drawn the model, rather than paint over the lines he had made, he painted around them. In many instances this has left a thin border of unpainted ground around the figure, which he might later reinforce with thin black painted lines. Alternatively, sometimes Modigliani reiterated his initial lines by scraping through wet paint, using the rounded end of the brush. This means that a characteristic dark halo, where the paint application is minimal, appears around his figures in X-radiograph imagery. Up to and around those dark outlines there are often thick ridges of paint. Exceptions to this rule may have been the result of revisions, where the artist was dissatisfied with a composition and chose to rework the contours.

In other works from our sample (particularly later works), paint application is generally very thin and has little leadcontaining pigment. X-radiographs of these paintings appear largely dark, with meagre density that appears only in areas that contain vermilion or lead-containing pigment mixtures (often in the flesh of the faces).

Modigliani often reused canvases and, in our study, X-radiographs have shown changes that occurred during the painting process. In one instance, Tate’s Portrait of a girl (C118; Fig.5), they revealed the superimposition of a later painting over an earlier figure (Fig.6). The underlying images are at times directly beneath the upper composition, sometimes off to one side, or occasionally inverted. One of the questions that arose from this study is whether Modigliani applied a masking layer before repainting the canvas with a new image, or whether he painted directly over the previous work.19 This remains unresolved, but given his apparent penchant for purity of line, it is possible that he preferred to work on a clean surface. In at least two of the paintings under scrutiny, however, this is clearly not the case. The head of an underlying figure is clearly visible in Bust of a girl (1908; private collection; C10), as is one at the bottom right of Nude study (C8; Fig.7), which is upside down, so that the canvas must have been reorientated. It may be that the paint that concealed the head in the lower half of each of these images has become more transparent with age or that, at a stage in his career in which Modigliani was yet to settle on a distinctive way of working, he did not bother to conceal his earlier composition.

Infra-red reflectography (IRR) is a valuable technique for the study of underdrawing and the initial laying out of the image before paint application. IR radiation penetrates paint and – where a white ground is present – is reflected. Where carbon black is present, in either drawing or painted layers, it is absorbed. There are limitations to the technique as it is sensitive only to the contrast between materials that reflect IR radiation (white and red) and those that absorb it (carbon black).20 Coloured grounds, where the contrast is diminished, can render underdrawing difficult to read. Not all the paintings in the study were subjected to IRR and, in those examined, variations of technique emerged. While careful preparatory drawing could be seen in all the paintings in the study, this varied from clean graphite lines to ones made with thin brushes in paint of varying colours.

Where possible, analysis of pigments and media took place, the results of which will feature in the forthcoming essays. Because most of the paintings in the study are in extremely good condition, opportunities to take microscopic samples for analy-sis were largely confined to the foldover edges, but this did allow for medium analysis, which provided a good indication of ground materials and the pigments present in certain colours. Some investigations also used non-destructive XRF analysis to identify elements within the paint film. The French study of 1981 suggested that Modigliani used a reduced palette to create all his colours,21 but the more advanced analysis of our small sample of paintings indicates that he used a more extensive – if still restricted – range. Here, once again, Modigliani’s conventional training is apparent: he used complementary colours to great effect.

Surface examination with a light microscope proved essential to understanding the finer features of the painting and helped to interpret findings from imaging processes such as IRR. Often, if lines of preparatory drawing are visible on the surface of a work, close looking reveals whether they were made with pencil or paint. This is particularly evident in the more thinly painted works of the later period. It is also possible (if no carbon black drawing is visible in IRR) that lines were applied in a colour that infra-red imaging cannot detect. Viewing paint under magnification can reveal the order of its application: overlapping layers of paint, or wet-on-wet paint application, which can produce a marbling effect on a microscopic level. Microscopic examination of Modigliani’s technique is extremely rewarding; it reveals a delicacy of handling not readily visible to the naked eye. Although the artist is known for his broad, energetic brushwork, microscopy reveals fine, deftly applied brushstrokes. In many of the works examined, the final outlines were applied on top of the paint layers, using a very fine brush. In the eyes of some of Modigliani’s painted sitters, fine, feathered outlines and careful, minimal highlights are discernible (Figs.8 and 9). It is perhaps at this level of examination that the artist’s technical mastery is most evident. The images captured in the process, in relation to their respective paintings, are in a sense analogous to the project and its contribution to Modigliani studies. They provide rich and fascinating detail, but they are only part of a far bigger picture.

Most importantly, despite a critical tendency to emphasise the artist’s unruly and tragic personal life, this study has revealed that Modigliani was anything but chaotic in his painting methods. He developed a highly skilled, unique and consistent technique. The Burlington Magazine is a particularly fitting place for its exploration; it was here that almost one hundred years ago, Roger Fry’s essay on Modigliani flagged the artist’s importance to a community of art connoisseurs.22 Using new methods of examination, with a quality of imaging that would have been unthinkable at that time, we hope these articles will reinvigorate discussions of Modigliani’s contribution to twentieth-century art.



The Modigliani project began as the brainchild of Nancy Ireson and Simonetta  Fraquelli, who approached Tate Conservation about examining Tate paintings in  preparation for the exhibition Modigliani at Tate Modern, London (23rd November  2017–2nd April 2018). Tate would like to thank the Bank of America Merrill Lynch Art Conservation Project, whose support made the technical research at Tate possible and supported Annette King’s and Joyce H. Townsend’s participation in the wider Modigliani research. Tate approached other institutional lenders with the facilities to carry out technical examinations on their own works: those who have collaborated on the project are listed in bold below. The results are thanks to a great many people who do not appear as co-authors. The authors  would like to gratefully acknowledge Francesca Casadio, Kim Muir and Frank  Zuccari of the Art Institute of Chicago; Robin Craren, Judith Dolkart (formerly),  Cindy Kang and Anya Shutov of the Barnes Foundation, Philadelphia;  Silvia Amato, Aviva Burnstock and Barnaby Wright from the Courtauld Institute  of Art, London; Marie-Christine Decroocq and Marc Restellini of Institut  Restellini, Paris; Elsje Janssen, Lizet Klaasen and Lies Vanbiervliet at Koninklijk  Museum voor Schone Kunsten, Antwerp (KMSKA); Elma O’Donoghue,  Charlotte Eng, Yosi Pozeilov and Virginia Rasmussen at the Los Angeles County Museum of Art; Sylphide de Daranyi of Musée de l’Orangerie, Paris;  Silvia A. Centeno at the Metropolitan Museum of Art, New York; Anny Aviram,  Diana Hartman, Kate Lewis, Chris McGlinchey and Ana Martins of the  Museum of Modern Art, New York; Ann Hoenigswald from the National  Gallery of Art, Washington, for her generous sharing of information on the  Chester Dale Collection; Don H. Johnson of the Thread Count Automation Project  at Rice University, Houston; Adam Finnefrock, Alyssa Hull and Jennifer Mass at  Scientific Analysis of Fine Art (SAFA); Julie Barten, Corey D’Augustine, Susan  Davidson, Lauren Engel, Jennifer Hickey and Hillary Torrence at the Solomon  R. Guggenheim Museum, New York; Marian Cousijn, Mark Heathcote, Joe  Humphries, Emma Jones, Marcella Leith and Emma Lewis of Tate, London;  Brian Singer at the University of Northumbria, Newcastle upon Tyne; Pedro  H.O.V. Campos and Márcia Rizzutto, Institute of Physics at the University of  São Paulo (IF USP); Anikó Bezúr and Pablo Londero at the Institute for the  Preservation of Cultural Heritage, Yale University Art Gallery, New Haven;  and the independent art historians Marianne Le Morvan and Kenneth Wayne. 

1 N. Ireson and S. Fraquelli, eds.: exh. cat. Modigliani, London (Tate Modern) 2017–18; it was reviewed by Sarah Whitfield in the February issue of this Magazine, pp.152–53. 

2 See M. Esterow: ‘The art market’s Modigliani forgery epidemic’, Vanity Fair (May 2017),, accessed 2nd February 2018. 

3 See S. Delbourgo and L. Faillant-Dumas: ‘L’Etude au Laboratoires de Recherche des Musées de France’, in B. Contensou and D. Marchesseau: exh. cat. Amedeo Modigliani, Paris (Musée d’Art Moderne de la Ville) 1981, pp.20–47. 

4 Issues identified included the fact that Modigliani did not date his works and was known to have destroyed a great many of them. He was very reticent about his art and rarely spoke or wrote about his motivations or methods. There is also no open access archive to which one can refer for the artist’s letters, memoirs, receipts detailing materials or artists’ suppliers. 

5 ‘Amedeo Modigliani and his Secrets’ is a research project that aims to study every painting and sculpture by Modigliani in French public collections. It is coordinated by LaM in association with the C2RMF and the CNRS (French National Centre for Scientific Research) Miniaturization for Synthesis, Analysis and Proteomics Laboratory (MSAP in Lille 1 University). 

6 A. Ceroni: I dipinti di Modigliani (Classici dell’arte 40), intr. L. Piccioni, Milan 1970. The one work by Modigliani dealt with in the study that is not listed in Ceroni is Madame Zborowska, 1918, canvas, signed top right, 64.5 by 46 cm. (Tate,  London, T03569). Throughout this series Ceroni’s catalogue numbers will be used. 

7 Modigliani’s exhibition at the Galerie Berthe Weill, Paris, 3rd–30th December  1917, caused a scandal and the gallery was forced to take down the nudes, see E.  Braun, S. Fraquelli and N. Rosenthal: Modigliani and his Models, London (Royal Academy of Arts) 2006, p.24; and P. Sichel: A Biography of Amedeo Modigliani, New York 1967, pp.395–97. 

8 See A. Zborowska: Modigliani et Zborowski, Paris 2015, p.41. 

9 N. Hamnett: The Laughing Torso, London 1932, p.134. 

10 The difference between a stretcher and a strainer is that a stretcher has expandable  corner joints, so it can be adjusted by means of wooden ‘keys’ to provide increased  tension in the canvas. A strainer has fixed joints and tension cannot be adjusted. 

11 Portraits by Modigliani made on marine canvases in size 40M include The little peasant (c.1918; Tate, London; C257), The Italian woman (c.1918; Metropolitan Museum of Art, New York; C268) and La belle Epicière (1918; Nahmad Collection,  Monaco; C256). 

12 Interestingly, this strainer was not a standard size; for Modigliani at the Cité  Falguière, see A. de Stefani: ‘Modigliani alla Cité Falguière: La Prima fase della  scultura nel suo contest immediate’, Studiolo 12 (2015), pp.242–67. 

13 Modigliani is listed alongside artists such as Brancusi, Picasso, Rousseau, Léger, Kandinsky and Miró as customers of Lefebvre-Foinet in ‘Notes from the underground:  The Lefebvre-Foinet collection’,, accessed 25th January 2018. 

14 P. Labreuche: Guide historique des fournisseurs de matériel pour Artistes à Paris, 1790-1960 (hereafter Guide Labreuche),, accessed 1st January 2018.   

15 Guide Labreuche lists the date of the rue de la Grande-Chaumière shop as 1936, so it would have been after Modigliani’s time if this is the date it opened for business,, accessed 1st January 2018. 

16 D.H. Johnson: ‘Thread counting: its role in technical painting examination  and how to obtain thread count results’, Thread Count Automation Project, http://, accessed 1st January 2018. 

17 The pencil lines were identified in Portrait of Henri Laurens, by Amedeo Modigliani. 1915 (Private collection; C53). 

18 A. Burnstock: ‘Interpreting the results of technical analysis’, in idem, T. Cooper, M. Howard and E. Town, eds.: Painting in Britain 1500–1630, Production, Influences, and Patronage, Oxford 2015, p.97. 

19 This will be discussed with regard to the portrait of Jacques and Berthe Lipchitz (1916; Art Institute of Chicago; CR161) in a forthcoming article in this series. 

20 Burnstock, op. cit. (note 18), p.100.

21 Delbourgo and Faillant-Dumas, op. cit. (note 3), pp.44–45. The pigments listed are chrome yellow or cadmium yellow, ochres, vermilion red, chrome green, Prussian blue and some organic colours, lead white and zinc white. 

22 R. Fry: ‘Line as a means of expression in modern art’, the burlington magazine 34 (1919), pp.62–63, 66–67 and 69.