[ Effortless Virtuosity ]

Looking at the later portraits of Sargent, the viewer is overwhelmed by an impression of effortless virtuosity. Indeed, the observation has been used both to laud and to criticize his ability.(1) But despite this impression, close scrutiny reveals more often than not contrary evidence for extensive development and range by the artist. This became increasingly evident during the treatment of a group of his paintings at the Tate Gallery prior to the John Singer Sargent exhibition due to open at the museum in October [1998], and prompted a closer study of his working method and his choice of materials.[p]

These are illustrated here by Sargent’s portraits at the Tate, notably the nine of Asher Wertheimer and his large family[pic], the unfinished Study for Madame Gautreau[pic], and the portrait of Vernon Lee[pic]. Small samples were removed from some of these paintings for analysis of the pigments and media, which were undergoing varnish removal in the conservation studio. The authors have also examined a number of other portraits by Sargent in collections in the United Kingdom and the United States, and have been given access to conservation records and x-radiographs held by other institutions(2).

[ Limited Working-out of Ideas ]

Sargent’s clear ability to draw, apparent today in the drawings visible beneath his watercolours,[pic] and his pencil drawing and sketches for subject’s other than portraits,[pic] surely suggest that drawing would be an integral part of his process. There are preparatory drawings for some paintings. Loose sketches, in very soft pencil on paper (Fig.1) have been identified for Asher Wertheimer’s portrait (Fig.2), and for the original companion portrait of Mrs. Wertheimer (Fig.4).(3) For the latter, Sargent executed several sketches to play with the details [pic]. The drawings are not minutely detailed but provided the essence and feel of the chosen pose seen in the finished painted portraits. The backgrounds were superimposed during painting. There are many pencil sketches for Madame Gautreau[pic] and also for the lilies and the head of one of the children in Carnation, lily, lily, rose [pic] which apparently serve the same function.(4)

There is little evidence to support the idea that Sargent habitually drew or painted an outline directly onto the canvas, though he did do so on occasion. Suggestions of drawing lines have been revealed in Oyster gatherers at Cancale(5)[pic] by the use of infrared photography. For the informal oil sketch, Vernon Lee (Fig.3), the sitter noted in a letter that Sargent ‘did that sketch of me in a sitting of about an hour, with scarcely more preparation than the pencil outline still visible along the jaw'.[pic](6) In contrast another sitter noted that no drawings were made for her portrait.(7) In the unfinished Tate work Study For Madame Gautreau (Fig.5), rightly regarded as a replica of the better known and fully-finished portrait at the Metropolitan Museum of Art, New York[pic],(8) there are very finely drawn reddish brown lines delineating the skirt,[pic] and to indicate the presence of the table seen only in the finished work. In addition the broad outline of her hemline is indicated with a sweeping black brushstroke. This is in stark contrast to the unfinished Edward Wertheimer (Fig.6),(9) where there is no drawing and the solid form is generated by the coming together of broad planes of colour. Madame Gautreau's unfinished state should, however, be seen as a copy of a much altered yet already finished image: the outline in pencil of a fully resolved painted design is not surprising.(10) Carrying out studies in oil on canvas may also have provided Sargent with the opportunity to gain the necessary understanding of the sitter's character to facilitate a painting. 

If it were the broad feel for the painted image that was captured at the drawing stage, rather than the finished outline, this goes some way to explaining the need for further development of the image during painting. Because it is unfinished, Edward Wertheimer provides us with the opportunity to see the early paint applications before changes were introduced.[pic] On the cool grey priming, the face is the most developed area; it is put in with broad strokes of colour in a technique which mirrors closely the description, by a fellow student of Sargent's, of his teacher Carolus-Duran's method:

No preparation in colour or monochrome was allowed, but the main planes of the face must be laid directly on unprepared canvas with a broad brush. These few surfaces -- three or four in the forehead, as many in the nose, and so forth -- must be studied in shape and place . . .(11)

Sargent did not, however, as was also cited, begin by 'drawing it [a head] on the canvas in charcoal and fixing it'.(12) In contrast, though he certainly did carry through his teacher's recommendations "to capture the envelope of a figure" and the dynamic relationship between the model's contours and what surrounds them'.(13) The background of Edward Wertheimer was blocked in, leaving space for the figure, whose bulk was then suggested with broad strokes and planes of colour which nudged the background back into place. A similarly worked but much smaller sketch [pic] has been published for Sargent's portrait of Carolus-Duran himself, executed in 1879.[pic](14)Study for Madame Gautreau[pic] again provides an interesting contrast. With development unnecessary we still see the envelope of the figure captured, but it was not achieved by those nudging broad planes of colour seen in the portrait of Edward Wertheimer, which would be continually modified in finished works, but was simply laid on with the background tones painted round an outline. With Study for Madame Gautreau, Sargent's method did not undergo any major development, though it permits us to view his immense skill in handling flesh tones and shape once an idea had been determined. Sargent's approach was captured by a fellow-artist's description of him at work:

Sargent, when he painted the size of life, placed his canvas on a level with the model, walked back until canvas and sitter were equal before his eye, and drew with his brush, beginning with the shadows, and gradually evolving his figure from the background by means of large, loose  volumes of shadow, half tones and light, regardless of features or refinements of form. He painted with large brushes and a full palette, using oil and turpentine freely as a medium. When he repainted, he would smudge and efface the part he wished to reconstruct, and begin again from a shapeless mass.(15)

[ Application of Paint ]

Sargent's liberal use of oil and turpentine to thin his darker paint is certainly confirmed by the dribbles seen on tacking margins. The subtlety of the almost bare section of canvas, the reserve, kept for Edward Wertheimer's right hand should not, however, be overlooked.[pic] In the midst of the rapid and broad paint applications the basic form of the hand is simply captured by the absence of paint. In his preparation for his large and complex schemes, such as the murals for the Boston Public Library[pic], he used small oil sketches scaled up by a complex process involving photography.[pic](16) It is of note that the broad brushstrokes, bold delineation of light and shade, lack of clear outlines and contours,[pic] and generally muted colours of the Library sketches in comparison to the final works,[pic] are all echoed in the unfinished portraits which the authors have examined. Perhaps these works may also illustrate his initial painting stage for portraits.

Having achieved the fundamental laying-in of the design, Sargent continued to develop and modify the image, by shifting the boundary between the figure and the background, to achieve a deliberately sought-after fine balance of shape and colour. A sitter wrote that Sargent would 

...slowly and deliberately recede about a dozen steps from the easel and suddenly,[pic] the brush lifted ready for action and without ever taking his eyes off of me, make a dash for the canvas on which he then recorded his impression, generally accompanying the act by contentedly humming a little tune.(17)

The rubbing-out and the charge at the canvas are somewhat reminiscent of Whistler's methods for portraits,(18) although Whistler made a more thorough job of effacing the work of his many sittings than Sargent appears to have done. Whistler also frequently required many more sittings than the two to ten which Sargent tended to work with (though with notable exceptions).(19) Sargent's struggles with composition and balance are seen in the group portrait, Hylda, Almina and Conway Wertheimer (Fig.7), where close examination reveals the alterations in the position of the dogs (reluctant sitters, presumably, who kept suggesting new poses to Sargent as they wriggled into more appealing positions), of Hylda's hat and her arm supporting the terrier, and the removal of a pipe which Conway originally smoked, though the most extensive change was to Hylda's skirt. Originally almost white, with a purple and pink pattern, the thin coat of varnish applied over this skirt (seen in paint cross-section) indicates that this version was considered to be complete by the artist, at least initially. Sargent, however, entirely repainted it in brown and added the pattern with very free brushwork, working wet-in-wet.[p]

This change effectively disrupted the tonal balance of the composition, but by modifying Conway's grey jacket with a brown resinous wash, and lightening the sky considerably on the right, he re-established the diagonal tonal opposites in the painting. Heavily-worked areas such as the sky at the upper right were worked wet with thick and opaque paint, but the background looks yet to be resolved, and despite the extent of the reworking, the mid-grey priming remains visible in places. A contemporary observation of changes to Ena and Betty Wertheimer (Fig.8) when seen on exhibition records that: 'Sargent's Wertheimer girls are changed, both appear to have been all painted over since I saw them in the studio. He has painted out a marble slab, put in a large vase and lightened parts here and there'.(20) Betty's slipped shoulder strap was also scraped out and repainted in the 'correct position: the change is visible on close examination.(21)

The arm of Mrs Wertheimer's chair (Fig.10) was rotated from the central and straight position across the picture plane to off-center diagonal which balances the ornament on the table, the chair back and the highlight on the right. In Asher Wertheimer's portrait,[pic] the balance is pushed to the limit, with the brilliant pink tongue of the poodle, Noble, almost draping over the edge of the gilded frame, leading a diagonal from the bottom left towards the background in the top right. Sargent also repositioned the feet of Graham Robertson[pic](22) to ensure he stood firmly on the ground, whilst the extreme verticality of the pose was exaggerated by redefining and narrowing the contours of his coat.

But this was not the end of it. A characteristic of Sargent's supports is an excess of primed canvas, in some cases several inches, lapped round the back of the stretcher. This excess of canvas provided yet further opportunity for Sargent to make modifications. For the portrait of Lord Ribblesdale,[pic](23) an originally folded-over edge, tack holes included, has been incorporated into the final picture. Almina Werrtheimer's portrait,[pic] by contrast, was reduced in size by at least four inches by tacking it to a smaller stretcher, with numerous early paint strokes hidden from view (Fig.9). No other size changes were made to the rest of the Wertheimer series but Ormond and Kilmurray describe three other instances among the early portraits of a reduction in size, and one further example where Sargent incorporated the tacking margin.(24)Carnation, lily, lily, rose [pic] has also been also been made smaller at a late stage in painting, on three of its sides. The informal portrait of Vernon Lee,[pic] however, has no such size changes. Adjustments of little more than an inch in some cases suggest how critical the final positioning of the figures within the framed area could be to the artist.

The process of developing the image continued even once the painting was framed. All of the Wertheimer portraits show signs of further paint being added after the frame was in place and this was not just the occasional stroke or final touch. In Essie, Ruby and Ferdinand Wertheimer,[pic](25) the majority of the eye-catching bold red strokes of the dress stopped markedly short of the frame edge, with only the laying-in stage of the fabric extending to the canvas edge. It is possible that these additions were made with the paintings hanging, since they were painted for a specific location. With Carnation, lily, lily, rose,[pic] the two upper-most lanterns on each side were added, along with other details and significant heightening of tones in places. The final varnish for each of the Wertheimer portraits was also applied once the paintings were in their frames.

Quite a number of photographs exist of Sargent in his studio, but since they have the air of a formal pose, they illustrate few painting implements and shed little light on his materials.(26) A study of Sargent's writings similarly reveals little about his suppliers of materials, and even the surviving records of his sporadic dealings over forty years with the colourman Roberson tell more about the accessories he bought for his easels (some ordered from abroad) than on the materials he used for painting.(27) The paintings themselves offer the only clues.

[ Canvases and Priming ] 

His portrait canvases all have a plain weave' (that is, with one warp and one weft thread), generally very fine, though a few are of medium weight. With an almost complete absence of canvas stamps or stretcher labels on both the Tate's paintings and those in collections in Boston,(28) there is little evidence for the supplier. A letter dated 1926 from Major George Roller, the restorer, concerning Sargent, stated that he had: 'an affection for a particular sort of French canvas. A canvas that probably was not often used by other painters, therefore not primed and kept in stock. Only got ready for him when he wanted it . . .'(29) The grey primed canvas for The Misses Vickers [pic] has a French stamp for Hardy Alan, Paris, with the stretcher seemingly supplied locally from Hibbert Brothers, Sheffield.(30) Edward Wertheimer's portrait has a stretcher label from the French colourman Meunier, with a handwritten note that it was ordered by Asher Wertheimer. However neither can be confirmed as Sargent's regular source of supply.

In every case, the generously stretched canvas is primed right to the edge, demonstrating that it was cut from a larger length. This gives a possible clue to the absence of canvas stamps: there might be only a few on large primed lengths, which could end up on discarded strips of material, or wrapped round a stretcher bar. The colour of the priming for nearly all of the Tate's portraits is grey, as it is in the majority of the later Sargent portraits seen by the authors where the priming was visible: the rest are white. The grey primings contribute to the overall cool appearance of these works, and serves to provide a mid-tone which also intensified the colour of the brown, thinly-painted backgrounds which Sargent frequently used in portraits.[p]

The primings on the Tate paintings are typical commercially-applied ones, and show a subtle yet significant variation in shade. This could have been Sargent's deliberate choice, but equally could reflect batch-to-batch variation in primed canvases ordered over a period: the Wertheimer portraits were painted over ten years, and have primings which vary in tone from light grey to mid grey. All have a cool tone made by mixing ivory black and lead white in linseed oil,(31) and several have a thick layer of glue size applied over the canvas, a feature not often seen in British primed canvases of this date, which suggests purchase abroad. Fine, linear, brittle cracking, often seen in Sargent's paint layers, suggests that the pre-primed canvas was rolled early in its lifetime, since cracks in the priming eventually impose a similar pattern of cracking in the paint. The canvas was possibly rolled for delivery to Sargent's studio, then cut to fit his stretchers, leaving the margin of fabric already noted. This would have been a practical method for a successful - and therefore busy - artist who did not use consistent stretcher dimensions or formats for his paintings. Because no correspondence with colourmen has been located, we cannot tell whether Sargent ordered a bespoke stretcher for each portrait and then a second one when he wanted to change the format slightly.

[ The Painter's Palette ]

As for his suppliers of paint it seems likely he used materials easily available at the time. Sargent used tube and block watercolours from Winsor and Newton, as well as a wider range of tube colours from American suppliers,(32) and some Meissonier watercolour brushes,(33) when in America. The same commonly available range of pigments is seen in virtually all the Tate's later portraits, and has been found in the Boston Public Library murals,(34) in examination of El Jaleo,[pic](35) and on his palettes in the Fogg Art Museum, Harvard University,(36) and the Royal Academy of Arts, London (Fig.12).(37) The range is quite wide but does not include every pigment available at that time. He regularly used Mars yellow (a synthetic iron oxide) and cadmium yellow; viridian and emerald green, sometimes mixed; vermilion and Mars red, both alone and mixed; madder; synthetic ultramarine or cobalt blue; and ivory black, sienna, and Mars brown. The dark backgrounds of many portraits include a mixture of ivory black, Mars brown, and a generous quantity of paint medium, which gives a colour similar to the traditional Vandyck brown.  A pale shade of chrome yellow, cadmium red, and cobalt violet were found on occasion, but not in every portrait examined. There is a more limited selection of blue and yellow pigments in the later portraits than in the earlier ones, which is once more reminiscent of Whistler's deliberate limitation of his palette to create a colour harmony, and to fix a cool or warm overall tone to each painting.(38) Like other artists in Britain in the late nineteenth century, Sargent's tube paints include lead white, and easily detectable quantities of extenders such as kaolin, chalk, talc and barytes, but very rarely any zinc white. Where zinc white has been found in a portrait, it is generally associated with a single colour, and may not have been selected deliberately by the artist. In the twentieth century, it would become more common in oil paintings.

Analysis of a small number of paint samples from the Wertheimer portraits(39) revealed that Sargent's paint medium consisted of linseed oil in the darker colours and poppyseed oil in white paint: this is typical of manufactured tube paints at this time. Poppyseed oil turns yellow more slowly than linseed oil, and was preferred for grinding colours whose yellowing should be minimized, such as whites and pale blues.

Sargent mixed lighter colours such as flesh tones by adding to lead white, vermilion and a selection of other pigments including bone black, on occasion rose madder and even green viridian, and mixing them together roughly on the palette, then working them into and onto adjacent brushstrokes on the canvas, to give more subtle variations in tone, apparent on highlights. This kind of mixture can be seen on the palette which Sargent presented to the Royal Academy of Arts, London (Fig.12). The paint was not encrusted on the palette, but was essentially one layer thick, as though the palette were cleaned before each use for a particular painting. The finishing touches to faces were applied very much wet-in-wet, as the cross-section of paint from Ena Wertheimer's lips makes clear (Fig. 11). Fabrics which pick up reflections, for example Betty Wertheimer's white evening gown,[pic] can be seen under magnification to include a great number of colours, each applied as a separate brushstroke to paint which was wet. All the pigments used elsewhere on this painting can be identified in the paint of the gown, albeit in tiny amounts. The vigorous way Sargent applied paint to the canvas is clearly indicated by the numerous hog-hair bristles found trapped deep in the paint. The width (If his brushes varied considerably, with fine points being used for subtle details of faces, in contrast to the sweeping strokes up to an inch in diameter which he used to capture folds of fabric, in the later stages of painting. Unfinished portraits show that initial paint-layers have brushstrokes from quarter-inch and half-inch brushes: the boldest, broadest strokes were used for finishing.

There is little evidence to suggest that Sargent applied varnish at an intermediate stage, and since he used paints straight from the tube, or else thinned them, there would not have been much likelihood that the paint would have sunk badly enough to require an intermediate varnish. (Very lean paint tends to sink and become dull in appearance, which makes naturally dark passages difficult to interpret.) The Wertheimer portraits are unusual in that the original varnish, much discoloured, had been preserved, under several layers of later varnishes. It is typical of his era: a mastic-type spirit varnish(40) which yellows within a generation if the paintings are hung in daylight. A photograph exists of almost all of this group hanging together in the Tate c. 1931 in a daylit gallery, and it is known that they were displayed from the 1920s until then, so the  extreme yellowing of the varnish was not surprising. The recent cleaning of this series has brought about dramatic changes. Ena's seemingly flat yellow gown was revealed as white satin with lively reflections of her sister's burgundy velvet gown and the gold of the vase, surprising even conservators who are accustomed to visualize the suppression of form as well as colour that a very discoloured varnish imposes. Asher Wertheimer's portrait, with the tongue of his dog now a bright pink, has also regained vitality through removal of the yellowed varnishes. Yellowed varnish is particularly disturbing when present over paintings that have been painted with a cool tonality.

Sargent's extensive changes in composition to achieve the required fine balance in both shape and colour are now often visible through the development of a fine network of drying cracks, or as areas of wrinkled paint. However, a photograph showing Hylda, Almina and Conway Wertheimer [pic] on display at the Tate Gallery in the 1920s, more than fifteen years after its completion, does not show the severe cracks seen during recent conservation, which suggests that they did not develop rapidly due to a fundamental flaw in Sargent's method. A letter from MajorGeorge Roller in 1926 concerning the deterioration of Sargent's works stated, 'Some of them are certainly cracking. but not worse than the paintings of many other modern artists.. .', a painting by 'one of our greatest portrait painters, an R.A.. .', he stated, 'has cracked like a tessellated pavement, worse than any Sargent that I have ever had to treat'.(41) The condition of Sargent's painting was perhaps becoming an issue towards the end of his life but unlike Whistler, Sargent does not appear to have felt the need to modify his technique to overcome  potential drying problems.(42)

[ Conclusion ]

The limited working-out of preliminary ideas goes some way towards explaining the reason for the alterations and modifications found so regularly in his portraits. As has been noted recently: '...although Sargent was a gifted draughtsman and painter, he laboured over his work. He plotted his major canvases at length, often scraped away days if not weeks of effort, sometimes made second versions of a finished composition when the traces of the campaign appeared too evident. He strove hard, successfully, to make the result seem effortless'.(43) With his water colours  he seems to have simply discarded the sheet and restarted or simply restricted access to those he considered inadequate.(44) Carolus-Duran had taught that: 'Objects in nature relieve one against each other by the relative values of light and shade which accompany and are a part of each local colour. An outline, a contour, is, as we all know, a pure convention... an accepted convention...'(45)This description can be effectively applied to Sargent's portraits where the shapes and forms are never defined but implied by the positioning of planes of colour and seemingly random brush strokes. X-radiographs of finished portraits are dominated by these bold planes of colour, to such an extent that the subject has an almost complete absence of actual contours within the image. The thickness of paint varies widely over the canvas in every portrait. Heavy impasto was used to sculpt jeweled details and hard highlights. The priming colour, simply washed over a warm tone, was used to provide the half-tones in Mrs Wetheimer's dress. The merest glimpse of underlying paint plays a vital role in constructing the sensation of solid shape. Sargent's portraits, with their seemingly unconnected mass of brushstrokes when viewed up-close, spring into shape at normal viewing distance. Close examination has revealed some common features in his portraits both in the method and choice of materials, but it is the lack of pedestrian methodology that in fact provides the common thread.
 

Notes:
The authors are very grateful to David Fraser Jenkins, Stephen Hackney, and Annette King for making useful comments on the text; Annette Southall who initiated research into Sargent's techniques when she began the conservation treatment of the Wertheimer portraits, continued later by Lesley Stevenson; Kate Oliver, who made information available from the Harvard University Art Museums; and Lydia Vagts (Museum of Fine Arts Boston) who has very generously shared her unpublished research with us.

Back to Intro Page

[p]  Paragraph breaks added with [P]  to help readability of large blocks of text on the net only

1) Marc Simpson, 'Sargent and His Critics’ in Uncanny Spectacle: the Public Career if the Young John Singer Sargent, Marc Simpson with Richard Ormond and H. Barbara Weinberg, exh. cat., New Haven and London, 1997

2) Lydia Vagts, John Singer Sargent: The preparatory sketches for his murals ‘The Triumph of Religion’ at the Boston Public Library, and their technical examination Harvard University Art Museums, 1993. The Alan Burroughs collection of x-radiographs and the conservation records, both held at Harvard University Art Museums

3) Mrs. Wertheimer, 1898, Oil on canvas, 147 x 95 cm. New Orleans Museum of Art.[pic] Painted as a companion piece to Asher Wertheimer, it was not considered as successful, and commissioned a second portrait of his wife in 1904, now in the Tate Gallery, London [pic]

4)  Richard Ormond and Elaine Kilmurray, John Singer Sargent: The Early Portraits, Complete Paintings, vol 1, New Haven and London, 1998, pp. 113-16.

5) Vagts, op. cit., for Oyster gatherers at Cancale, 1878, Oil on canvas, 41 x 61 cm. Museum of Fine Arts, Boston.[pic]

6) Ormond and Kilmurray, op. cit., p. 76.

7) Ormond and Kilmurray, op. cit., p. 109.

8) Trevor C. Fairbrother, John Singer Sargent and America, New York and London, 1986 (PhD thesis prepared for Boston University Graduate School, 1981). 

9) Both Edward and his brother Alfred died before their portraits were completed. However Alfred’s image [pic] was taken much further than Edward’s.[pic]

10) The lines can be seen by looking closely at the painting, but were not picked up with the infrared techniques in this case, indicating that Sargent’s drawing material here (and perhaps in other cases too) is not infrared-sensitive.

11) Barbara Weinberg, ‘Sargent and Carolus-Duran’, in Simpson, et. al., cit., p.25

12) ibid., p.25

13) ibid., pp. 19-20

14) Ormond and Kilmurray, op. cit., pp.41-43.

15) William Rothenstein [pic], Men and Memories: Recollections of William Rothenstein 1872-1900, London, 1931-32, pp. 192-92

16) Vagts, op. cit,. unpaginated.

17) Ormond and Kilmurray, op. cit., p. xxiii

18) Stephen Hackney, 'Colour and tone in Whistler's "Nocturnes" and "Harmonies" 1871-2', The Burlington Magazine, vol. cxxxvi, no. 1099 (October 1994), pp.695-99.

19) Ormond and Kilmurray, op. cit., p. xxiii

20) William C. Loring, 'An American Art Student in London', Archives of American Art Journal, vol. xxiv (1984), p.18

21) The position of shoulder straps seem to have been a recurring problem for Sargent. See: Hilton Kramer, 'The Case of Madame's Missing Shoulder Strap', The New York Times, 1 February 1981. Also discussed in Ormond and Kilmurray, op. cit., pp. 113-16

22) Graham Robertson,[pic]1894, Oil on canvas, 230 x 119 cm. Tate Gallery, London.

23) Lord Ribbledale, 1902. Oil on canvas, 258 x 144 cm, National Gallery, London.

24) Ormond and Kilmurray, op. cit., p. 133, 144, 157, 171

25) Essie, Ruby and Ferdinand, children of Asher Wertheimer, 1902, Oil on canvas, 161 x194 cm, Tate Gallery, London.

26) E.  Kilmurray, personal communication.

27) Sally Woodcock, personal communication.

28) Vagts, op. cit., unpaginated.

29) Reproduced in Vagts, op. cit., unpaginated.

30) James Hamilton, The Misses Vickers: The Centenary of the Painting, by John Singer Sargent, Sheffield Arts Department, July 1984, p.61. The Misses Vickers, 1884, Oil on Canvas, 138 x 183 cm. Sheffield City Art Gallery

31) Joyce Townsend used optical microscopy to identify the pigments, and together with Marianne Odlvha of Birkbeck College interpreted the results of GC-MS carried out by R.E. Tye of Kings College , London, on the paint medium.

32) Marjorie B. Cohn, Wash and Gouache: A Study of the Development of Watercolour, Centre for Conservation and Technical Studies, Fogg Art Museum, and The Foundation of American Institute for Conservation, Cambridge, MA, 1997, p.66.

33) Judith C. Walsh, 'Observation on watercolour techniques of Homer and Sargent' in American Traditions of Watercolor: The Worcester Art Museum Collection, New York, 1987. p. 44.

34) Aneta Zebala, 'An investigation of John Singer Sargent's Murals "The Triumph of Religion" at the Boston Public Library', pre-print of the Art Conservation Training Programs Conference, 1988, and Vagts, op. cit.

35) Mary Crawford Volk, 'On the Cleaning and Technical Analysis of El Jaleo, 1990-91, in John Singer Sargent's El Jaleo, exh. Cat., Washington, DC, 1992, pp. 123-26.

36) Analysed and published by Katrinka Leschey in Fine Arts 202, Technical Examination of Works of Art, Harvard University, 1983, and quoted by Zebala, op. cit.

37) Permission to sample the paint was granted by The Royal Academy of Arts. Optical microscopy was used to identify the pigments.

38) Joyce H. Townsend, 'Whistler's oil painting materials', The Burlington Magazine, vol. cxxxvi, no. 1099 (October 1994), pp. 690-95.

39) R.E. Tye of Kings College used GC-MS to analyse the paint medium and the results were interpreted by Joyce Townsend and Marianne Odlyha of Birkbeck College.

40) The Sargent varnishes have not been analysed, but their appearance, brittleness, refractive index and response to solvent cleaning test are all consistent with mastic spirit of this age taken from Tate paintings and analysed.

41) Vagts, op. cit., unpaginated.

42) A consequence of his technique to be seen today is a tendency for dark paint to become milky or bloomed, i.e. to take on a whitened appearance. This is not peculiar to Sargent: it is seen sometimes in his contemporaries' work too, most often in unconserved paintings. In Sargent's work is is related to the glossy dark brown tube paints which he used and possibly to the presence of a thick layer of glue size on the canvas. We hope that further studies and more detail analysis  of materials used in the formation of tube paint in the later nineteenth century, such as waxes, stearates and substitutions for traditional pigment Vandyck brown, will lead to a better understanding of this subject.

43) Marc Simpson in Simpson, op. cit., p. 31.

44) Walsh, op. cit., p. 62.

45) Weinberg, op. cit., p. 19.