Antoine Watteau 1684–1721
NG 2897 
La Gamme d’Amour (The Scale of Love)

Humphrey Wine, 2018

Extracted from:

Humphrey Wine, The Eighteenth Century French Paintings (London: National Gallery Company and Yale University Press, 2018).
Humphrey Wine and Virginia Napoleone, Former Owners of the Eighteenth‐Century French Paintings in the National Gallery: Appendix to ‘The National Gallery Catalogues: The Eighteenth‐Century French Paintings’ (London: National Gallery Company, 2018).

Probably 1717–18

Oil on canvas, 51 × 60.1 cm

Provenance¹

In the collection of Denis Mariette (1666–1741) by 1729;² in his posthumous inventory of 16 October 1741;³ by inheritance to his daughter and sole legatee, Jeanne‐Justine Chomel (died 13 December 1782?);⁴ sale of Sir John Pringle (1784–1869) (see the Appendix to this volume on the NG website), London, Christie & Manson, 13 May 1837, lot 53, £90 6s. to Pennell;⁵ Philips collection according to Goncourt;⁶ in the collection of Yolande Lyne Stephens (1813–1894) (see the Appendix to this volume on the NG website) when exhibited in 1874 (see under Exhibitions); her posthumous sale, Christie, Manson & Woods, London, 9 May 1895, lot 366, £3,507 to Agnew’s; sold by that firm to Julius (later Sir Julius) Wernher (1850–1912), mining magnate and philanthropist, on 13 May 1895 for £3,869 5s.;⁷ recorded in the Pink Drawing Room of Wernher’s London home at Bath House, Piccadilly, London;⁸ bequeathed by Sir Julius Wernher, Bt, 1912.⁹

Exhibitions

Paris 1874 (529); London 1898 (59); Bristol, Newcastle and London 2003.

Related Works

Paintings

A. Compositions by Watteau with one or more similar figures

In cases in which the composition was the subject of a print in the catalogue of prints by Dacier and Vuaflart published in 1922 (DV) (as part of a four volume work published 1921–9), the relevant DV number is given.

• (1) Berlin, Staatliche Museen zu Berlin, Gemäldegalerie, oil on canvas, 115 × 167 cm, inv. 474 B (fig. 1). An autograph work by Watteau, known as La Récréation galante, in which the principal group of guitarist and female companion in NG 2897 appears slightly left of centre.¹⁰ The scale of this group is virtually identical in both paintings, although there are small adjustments to these figures as between one painting and the other.¹¹ A painting attributed to Lancret in Wildenstein 1924, no. 377 (Munich, Bayerische Staatsgemäldesammlungen, oil on canvas, 127.5 × 192 cm, inv. 1161) has a similar figural composition to the Berlin painting, but a different landscape background.
• (2) Original lost. In the posthumous sale of the comtesse de Verrue, Paris, 27 March 1737, known from the engraving [page 548]by Jacques‐Philippe Lebas called Assemblée galante, and presumably autograph, by Watteau: see DV 139,¹² according to which the original painting was the same size as the print, namely 37.1 × 51.6 cm. It was most likely painted before the Berlin picture.¹³ The composition is close to, but not identical with, the Berlin picture. The guitarist in NG 2897 appears in reverse in the engraving, but his female companion is differently posed.
• (3) Lost. First recorded in the Jullienne collection, a composition known through a print by C.N. Cochin called Le bosquet de Bacchus (DV 265) on the right of which appears a child with its head on its hand, close to that in NG 2897. According to the print, the painting measures 1 pied, 7 pouces × 1 pied, 11 pouces, that is, 51.3 × 62.3 cm. The copy (on copper) of the painting recorded as DV 265 (in the same direction as the print) in Lady Wantage’s collection, and once regarded as the original, was sold at Drouot, 13 March 1989 (lot 53).¹⁴
• (4) Lost. A child with its head on its hand close to that in NG 2897 also appears in L’Occupation selon l’age, a composition known from a print by C. Dupuis (DV 208). According to Dacier and Vuaflart the original painting is the same size as the print, namely 32.7 × 41.6 cm. A version of the painting catalogued as a copy is in the National Gallery of Ireland, Dublin, inv. 1321, oil on canvas, 34 × 42 cm.¹⁵
• (5) A figure substantially like that of the guitarist in NG 2897, but in reverse, appears right of centre in Fête champêtre by Watteau with possible assistance of Jean‐Baptiste Pater (Art Institute of Chicago), but his female companion is not looking at him and her body is hidden by other figures in the painting. In addition an X‐radiograph shows that the woman with a fan who is seated left of centre was initially positioned rather like the woman holding the music in NG 2897.¹⁶
• (6) Los Angeles, Los Angeles County Museum of Art, L’Accord parfait (p. 568, fig. 2), oil on panel, 35.5 × 28 cm, inv. AC 1999.18.1 (DV 23). The couple walking away from the viewer in the Los Angeles painting is an inverted and differently costumed version of that in NG 2897.

B. Copies of NG 2897

• (1) A presumed copy on wood, 33 × 45 cm, Charles Iweins sale, de Brauwere, Brussels, 9–10 April 1892, lot 101,¹⁷ and in the Charles Iweins deceased sale, Van Halteren and de Brauwere, Brussels, 16 February 1898, lot 129.¹⁸
• (2) A painting on canvas, 27.9 × 35.5 cm, signed ‘Dietricy Pinx 1753’, (that is, Christian Wilhelm Ernst Dietrich, 1712–1774), was recorded by Davies 1957, p. 222, at Bagnères‐de‐Bigorre. In his manuscript note in the National Gallery dossier Davies called it: ‘A queer copy inverted & with the colours different … obviously done after Le Bas’ engraving’.
• (3) An inverted copy, called ‘The Love Song’, and attributed to Lancret, 78.7 × 62.2 cm, was sold at Christie’s, London, 6 April 1955, lot 175, £25 4s. to Waters. [page 549]

  

• (4) A copy on copper, 22.2 × 27.9 cm, Christie’s, London, 11 November 1955, lot 53.
• (5) A copy, called ‘Continental School, c. 1900’, oil on canvas, 49.5 × 60 cm, signed and inscribed ‘Dupres [D’aprés?] [?] Watteau, M Behar?’, Bonhams, Chelsea, 11 January 2000.
• (6) A variant copy in reverse attributed to the ‘Circle of Jean Antoine Watteau’, oil on canvas, 63.5 × 75.6 cm, Sotheby’s, New York, 5 June 2002, lot 79.
• (7) A variant copy in reverse attributed to the ‘École française fin XIXe – début XXe’, oil on canvas, 16.5 × 23 cm, Eurl SVV Patrick Declerck, Douai, 29 November 2010, lot 79.
• (8) Not a copy, but one may note that in Jean‐Baptiste Pater’s Le concert amoureux (London, Wallace Collection, oil on canvas, 57.1 × 46.2 cm, inv. 383) the guitarist has a broadly similar pose to the guitarist in NG 2897. An autograph version of Pater’s composition with slight modifications is at Charlottenburg Palace.¹⁹

Drawings

A. By Watteau, here proposed as connected with NG 2897,

and numerically ordered as in Rosenberg and Prat’s 1996 catalogue raisonné of Watteau’s drawings (RP).²⁰ Drawings rejected by Rosenberg and Prat as autograph works by Watteau are here listed under C, Drawings by others, and are prefixed by R rather than RP. Where the drawing in question also appears in the 1957–8 Parker and Mathey catalogue raisonné²¹ the relevant PM number is also given.

• (1) RP 399 (PM 636) (fig. 2). Of the two figures in the drawing, that at the left appears as the woman with the child just behind the principal figures in NG 2897. The drawing is dated to about 1715–16 in RP.²² The same figure is there connected also to a figure in the background of Dulwich Picture Gallery’s Les Plaisirs du Bal, but the figure in question in that painting is seated in a more upright position.
• (2) RP 409 (PM 690) (fig. 3). Dated in RP to about 1715–16. The child at the right appears without its hat at the right of NG 2897, as well as in L’occupation selon l’age (see under Paintings A (4)) and Le bosquet de Bacchus (see under Paintings A (3)).²³
• (3) RP 470 (fig. 4). According to Grasselli it is ‘clearly related to the couple [at the right] in the London painting but later than it’, and she suggested that the pose of the woman in the drawing was derived from the drawing now only known through a counterproof (RP 500), for which see below.²⁴ She dates NG 2897 to 1716–17 and RP 470 to about 1717–18, and sees a closer connection between RP 470 and RP 599, from there leading towards the Dresden Gemäldegalerie Plaisirs d’amour. The position of the legs and visible forearm of the man, the turn of the woman’s head, and the position of her left arm in that painting, are rather different to the positions of the corresponding figures in the Dresden picture. Instead, and notwithstanding slight differences, RP 470 is not preparatory to the Dresden picture, but is a study used for the couple at the extreme right of NG 2897, a possibility not mentioned by RP, who tentatively suggest its date as about 1716. The couple at the right of the rapidly drawn sketch RP 599 in Chicago, which is certainly connected to the Dresden painting, may be derived from RP 470 or from NG 2897.
• (4) RP 479 (PM 830) (fig. 5). A part of this drawing at the left measuring 12.8 × 4.7 cm is missing and has been restored, and the drawing may have been cut along the bottom. The figure at the left was used for the guitarist in NG 2897, as it was for Watteau’s Assemblée galante (lost, see DV 139 and Paintings A (2)) and his La Récreation galante in Berlin (see Paintings A (1)). The guitarist in the Berlin painting wears a ruff, which he does not in the drawing or in NG 2897. On the other hand the construction of the beret in NG 2897 – radiating folds attached to a central, circular retaining button – is unlike the loose, floppy construction of the beret in the drawing and the Berlin picture. The body of the guitarist at the left of RP 479 is more upright than in NG 2897, and the guitar is held at a less acute angle to the ground. RP 479 is dated by Rosenberg and Prat to about 1716.²⁵ The guitarist at the right of the sheet and the outstretched arm at the top were both used for Watteau’s La Perspective (Boston, Museum of Fine Arts, inv. 23.573), which may be dated 1715–16. The guitarist at the right of the drawing may have provided Watteau with his first thought for the position of the legs of the guitarist in NG 2897 (see Technical Notes, below).
• (5) RP 498 (fig. 6).²⁶ Once in the Goncourt collection, present whereabouts unknown, a counterproof. Rosenberg and Prat were unable to connect it to any known painting. However, when the standing figure at the right of this counterproof is reversed, as it would have been in the drawing from which RP 498 was taken, it corresponds closely to the woman in the background of NG 2897 seen leaving with her companion, especially in the emphasis on the vertical folds of her dress, the large gather of material in the small of her back, and the horizontal fold near the elbow of her left arm (which, as one would expect, appears as the right arm in NG 2897). The drawing is dated about 1716–17 in RP.
• (6) RP 500 (PM 621) (fig. 7). A counterproof in red and black chalks of a lost drawing, which must have served for the seated woman at the extreme right of NG 2897, is in a New York private collection, dated about 1716–17 in RP.
• (7) RP 503 (PM 641) (fig. 8). Damaged lower left, possibly cut down the right‐hand margin, although, as pointed out in RP, the forearm seems to stop short of the edge. The figure of the seated man has been recognised as preparatory for that of the seated man at the right of NG 2897.²⁷ Grasselli connected the arm at the right of the sheet to a figure of a man turning towards a woman holding a fan in Les plaisirs du bal (Dulwich Picture Gallery, inv. 156), but this was rightly doubted by RP. Although the arm has been said not to have been used for NG 2897,²⁸ in fact it closely corresponds, both in its position and in the dark accents in the drapery, to the proper right arm of the woman holding the book of music.²⁹ There is one smudged line of the man [page 551][page 552] that appears to be under the disconnected arm, but it is not possible to say with certainty from the strokes of chalk alone that the arm was drawn later.³⁰ However, the position of the disconnected arm at the edge of the sheet strongly suggests that Watteau added it after having drawn the figure of the seated man. Grasselli first dated the drawing ‘no earlier than 1717’³¹ but later revised this to 1715–16.³² According to RP, it is as likely on stylistic grounds to be of 1716 as 1717.

  

  

  

  

  

  

  

• (8) RP 522 (PM 786) (fig. 9). The half‐figure at top left was used for the woman holding the book of music in NG 2897 (and for the similar figure in the Berlin Récreation galante noted in Paintings A (1) above). There are a number of differences between this woman and the woman in the Berlin picture. The poor condition of that painting precludes drawing any firm conclusions from this,³³ but there is also a small difference between the woman as drawn and the one in NG 2897, in that the hair of the latter has been pulled back more tightly to make it straighter. RP 522 is dated by RP 1716–17.³⁴ RP note two recorded counterproofs after the drawing, one in the British Museum, inv. 1860,0616.137, the other whose whereabouts are unknown.³⁵
• (9) RP 591 (PM 815) (fig. 10). Used for the guitarist in NG 2897, as well as for the Berlin picture (fig. 1 and Related Works, Paintings A (1)) and the lost L’Assemblée galante (Paintings A (2)). However, the position of the fingers is different, as is, more obviously, the position of the legs. The costume is drawn with more precision than in RP 479 (fig. 5) and, as in the Berlin picture, but not NG 2897, the figure wears a ruff. In spite of the absence of a beret RP 591 would appear to be essentially a costume study and to have been developed from RP 479, which was more likely a study for the pose of the upper part of the body and the hands. It is dated 1717–18 by RP, but the style of the drawing seems close to that of RP 501, which is dated by RP to 1716–17.

  

B. Drawings by Watteau, but less certainly, or less directly, connected with NG 2897

• (1) RP 328 (PM 595). Seen by Roland Michel 1984b as connected to the main female figure in NG 2897 (and in other paintings), but not so regarded by RP. Certainly the turn of the head and the position of the hands are different. The drawing may have been made with no specific composition in mind, as suggested by Margaret Morgan Grasselli, or it may have been used for the lower half of that figure’s body, as Vogtherr has suggested,³⁶ but both the fall of drapery and the angle of the sole of the foot in relation to the ground differ.
• (2) RP 384 (PM 569). Not previously connected with NG 2897, but possibly an earlier study for the woman seated at the extreme right. Compared to NG 2897, however, the drapery has differences, the right hand is absent, and the [page 553] tilt of the head is slightly more towards the right. Possibly, the original of the drawing of which RP 500 (fig. 7) is the counterproof was developed from RP 384. Dated 1715–16 by RP.

  

• (3) RP 421. Not previously connected with NG 2897, nor probably directly used for it, but the head of the Virgin is sufficiently close to that of the woman seated at the right of the painting to suggest that it provided a type which Watteau would use in a number of his paintings. Of these the head in NG 2897 is the closest to the head in the drawing.³⁷ An engraving after this drawing was made by Jean de Jullienne. An inscription on a proof of the engraving in the Bibliothèque nationale, Paris reads: ‘Le dessein original est du Schidone il est des deux tiers plus petit et on le voit dans la Collection de Mr Mariette. Watteau qui le trouvait Charmant la dessiné tel que le voici et c’est Mr de Julienne qui l’a gravé’. It has been suggested that Watteau copied Schedoni’s drawing while it was in the Crozat collection and that Pierre‐Jean Mariette acquired it at the Crozat sale in 1741, the inscription therefore being later.³⁸ The inscription most likely refers to Pierre‐Jean Mariette, but could conceivably be a reference to his father, Jean, or even to Pierre‐Jean’s uncle and first recorded owner of NG 2897, Denis Mariette. No drawings, however, were recorded in Denis Mariette’s posthumous inventory.
• (4) RP 431 (PM 436) (fig. 11). Discussed under Subject, below. Date 1715–16 in RP.
• (5) RP 490 (PM 579). Red, black and white chalks, stomping and pencil on light brown paper, Paris, Louvre, inv. 3336. The figure at the left of this sheet bears some resemblance in reverse to the principal female figure in NG 2897, but, [page 554] contrary to the views of Goncourt and of Parker and Mathey,³⁹ insufficient to be convincing as connected to the painting, and has been rejected as such by RP.

  

• (6) RP 599 (PM 858). The date suggested in RP is 1717–18. See commentary under Drawings (A 3).

C. Drawings by others

• (1) R 35. Basel, Öffentliche Kunstsammlung, Kupferstichkabinett, inv. 1948.182.11. A drawing by an eighteenth‐century(?) French hand of the principal figures has rightly been rejected by RP as by Watteau.
• (2) R 124. Dijon, Musée Magnin, inv. MMG 1938‐971. Possibly a copy by a later eighteenth‐century French hand of the head of the woman at the extreme right of NG 2897, although the expression is less tranquil.
• (3) R 519. Paris, Musée Cognacq‐Jay, inv. J.188. As noted by RP, a partial copy (in reverse) of RP 409 (see Drawings (A 2) above), which has been attributed to Boucher, an attribution now doubted on stylistic grounds.⁴⁰
• (4) R 625. Whereabouts unknown. As stated by RP, a counterproof of a copy after RP 384 (for which see Drawings (B 2) above).

Engravings

• (1) DV 199 (fig. 12). Identified in DV as an etching, it is in reverse by Jacques‐Phillipe Le Bas made in or before 1729, published by François Chereau and titled La Game [sic] d’Amour at left;⁴¹ Amoris Diagramma Musicum at right. No verse is inscribed on it.⁴² A counterproof of this etching, hence in the same direction as NG 2897, and not noted by Dacier and Vuaflart, exists in the Collection Edmond de Rothschild, Paris, Louvre, département des Arts graphiques, inv. 21747 LR.⁴³

  

• (2) Etching in reverse by François Boucher after the bust‐length study of a child on the right‐hand side of RP 409.⁴⁴
• (3) DV 199A. Identified in DV as an etching, it measures 26.6 × 34.5 cm and is inscribed in the plate at left: ‘C.P. Caes Maj.’; in the centre: ‘No. 20’; and at right: ‘Mart. Engelbrecht ex. A.V.’.
• (4) Rosenberg and Prat 1996, engraving no. 67. An etching by Pierre‐Charles Trémolières of a standing female figure seen from behind. Although connected by RP to NG 2897, there are too many differences between it and the figure in the background of the painting for this to be said to be after it. Jacques Vilain identifies the etching as after a study of a figure in Watteau’s La Perspective now in the Museum of Fine Arts, Boston.⁴⁵

Technical Notes

The impasto has been somewhat flattened during a lining treatment, as was noted in 1946 in the Conservation dossier, and there is wear and retouching at the edges. There is a vertical tear some 4 cm long above the woman’s head, a loss behind her head, and another to the left of the tree. There is also a very small loss to the exposed part of the singer’s proper left breast and a few others at the very bottom of the canvas. NG 2897 has been retouched in front of her head, between her bodice and the music book, in her companion’s right hand and in the background figures. The music book itself is mainly intact, but the branch to the right of the herm has been reinforced. There is a splash of red paint, which is most likely original, in the singer’s skirt. There is a heavy craquelure in parts of the picture. The pattern of cracks in the singer’s skirt suggests that, as was typical of Watteau, a layer of paint was applied over an earlier layer that was not yet completely dry. The painting was last cleaned and retouched in 1958.

The support is a rather coarse plain‐weave canvas, strongly cusped all round, which was lined prior to NG 2897’s acquisition in 1912. Its reverse bears a label (possibly dating from the nineteenth century) on which has been written in ink: ‘74’. From visual observation the painting has a ground consisting of a layer containing red‐brown ochre, and a warm mid‐brown layer of mixed pigment on top.⁴⁶ Prussian blue has been detected microchemically in the greenish‐blue paint of the sky, where it is mixed with white and small amounts of green earth and Naples yellow (lead antimonite yellow). Naples yellow has been confirmed by analysis in the skirt of the woman holding the music book. Green earth occurs widely in the foliage and background greens. The red lake used in the guitarist’s costume has faded.⁴⁷

The X‐radiograph suggests some changes in the area of the guitarist’s proper right hand, and that the background was painted leaving spaces in reserve for the figures in the background. NG 2897 has been examined by Rachel Billinge using infrared reflectography (fig. 13). Her report notes a simple, linear underdrawing for all the figures in the painting, made with paint using a brush.⁴⁸ There is visible to the naked eye a line near the guitarist’s proper right calf (fig. 14). This is in fact underdrawing rendered visible by [page 555] the thin final layers of paint becoming more transparent with age. The underdrawn outline of the calf began to the left of the painted position at knee level, crossing under the painted calf to form the ankle, itself level with the toes of the painted left foot. Thus the underdrawn right foot was closer to the painted left foot, and its toe ended where the rose is now painted on the shoe. It may be noted that the angle of the underdrawn outline is close to that of the proper right calf of the right‐hand guitarist in RP 479. There is also what might be a bow underdrawn, but not painted, at the right knee. The left leg seems also to have been moved slightly to the right. The guitarist’s proper right shoulder was underdrawn slightly higher, and the outline of his jacket to the right of his right hand is clearly visible in the reflectogram. A series of lines projecting out from the guitar a short way along his leg are also underdrawing, possibly for a slashed or fringed bottom edge to the jacket. Also visible in the reflectogram is underdrawing for what appears to be a smaller guitar. The underdrawing for the [page 556] main female figure is more closely followed in the finished painting, albeit with slight changes to the neckline of her dress, and to her left hand, which seems to have been drawn slightly lower. Her ear was first painted flatter against her head, and so to the left of its final position, before being repainted as it is now seen. The smaller background figures also show some underdrawing, with only slight changes to the group with the child on the right. The reflectogram is difficult to interpret in the area of the standing couple to the right of the guitar (fig. 15) since it shows the paint used for the figures’ draperies as well as underdrawing for them roughly in their painted positions. Also visible in this area of the reflectogram are more lines and dark areas, which may be an initial drawing and blocking‐in for two standing figures slightly smaller in scale and lower down.

Discussion

By the eighteenth century the music‐making couple had a long tradition in Western art, especially in Northern and Venetian painting,⁴⁹ and the theme was also long established in pastoral poetry. The links between poetry and Watteau’s paintings were recognised by his contemporary, the comte de Caylus.⁵⁰

The compositional arrangement of the principal couple in NG 2897, and in the other paintings by Watteau in which they appear, may have had their origin in a copy Watteau made of a lost drawing after a Venetian draughtsman, possibly Domenico Campagnola (1500–1564) (fig. 11).⁵¹ The pose of the violinist is close to that of the guitarist in NG 2897, and his position in relation to the seated figure at whom he is looking is close to the relative positions of the main protagonists in the painting.

The popularity of the guitar, which on account of its soft sound was perceived as an instrument to be played among intimates,⁵² was in temporary decline in court circles when NG 2897 was painted.⁵³ Within Watteau’s oeuvre, however, guitarists appear frequently, and the guitarist as posed in NG 2897, or a figure very like him, and a female companion, appear in other known compositions (see Related Works, Paintings). Although Watteau counted amateur guitarists among his personal friends,⁵⁴ there is no reason to suppose that any of the figures in NG 2897 are portraits. The painting’s title, as with the titles of most of Watteau’s works, derives from the title of the print made some years after it, in this case by Le Bas, published by François Chereau (see Engravings (1)), although that does not exclude the possibility that the print adopted the title by which NG 2897 was already known. ‘La Gamme d’Amour’ (‘The Scale of Love’) may be taken as a reference to the musical scale, to the various stages of flirtation and seduction, and to the music which here provides their catalyst. The word gamme also had various secondary meanings (knowledge, ability and custom) which might have enriched the print for contemporaries, as might the figure of speech describing a person as ‘hors de gamme’, meaning that he had lost his way ‘comme un Musicien qui a perdu son ton’.⁵⁵

The guitarist is dressed in theatrical costume of a kind then available for sale or for hire in Paris,⁵⁶ and whose archaism, it has been suggested, was part of Watteau’s poetic distancing of his subject.⁵⁷ His companion wears a costume consisting of a separate bodice and skirt which was strictly informal attire for the fashionable woman and worn in the privacy of home or for country pursuits.⁵⁸ It was not specific to any particular period such as might assist with the dating of NG 2897. She appears to look at the guitarist’s proper left hand to gauge the note he is about to play, and together they might be reaching an understanding on the embellishments of the musical score with the implied meaning of mutual flirtation.⁵⁹ To the right, a seated couple make moves towards each other, and in the background another couple leave the scene to do whatever the spectator imagines. Between the guitarist and the seated couple is another seated woman looking pensively at a child leaning on her lap. The child looks out at the viewer. Possibly, one is meant to read the mother as being the same woman as those in the background at a subsequent stage of her life, and the child as the fruit of the offstage liaison possibly intended to be implied by the departing couple.⁶⁰

Although the guitarist is in fantasy costume, the guitar is an accurate rendition of a type made by the Voboam dynasty of instrument makers in Paris in the years 1650 to 1730.⁶¹[page 557] The prototype, of which there were numerous varieties, and of which examples still exist, consisted of a body of ebony and was embellished with marquetry decoration in alternate coloured woods around the soundboard and sound‐hole,⁶² and ebony ‘moustaches’ on either side of the bridge. In NG 2897 there are two lines of inlay around the sides of the guitar, marquetry decoration around the edge of the soundboard but not clearly around the sound‐hole, which has a single thin pale inlay and a wider dark (ebony?) inlay outside it, and an ebony ‘moustache’ on either side of the bridge. The back of the instrument cannot be seen, but neither the head nor the finger‐board appears to be decorated, so that overall the guitar in NG 2897 appears relatively simple in its decoration. Florence Gétreau sees similarities between the guitar in NG 2897 and one (more elaborately decorated) made by Jean‐Baptiste Voboam in 1708 (fig. 16).⁶³

The musical score is less convincing than the instrument and probably does not represent a real piece. There may be indications of a title at the top of the left‐hand page, but it is impossible to confirm this. The music shows a single line melody, rather than a score for singer and guitar accompaniment, and there are no numbers denoting the chords that one might expect for guitar music.⁶⁴ The guitarist, who has a suitably prominent thumbnail, appears to be in the course of striking, rather than plucking, the guitar strings ready to sound a chord, but in what note is unclear.⁶⁵ Although it has been said that the position of his proper right hand is such as to produce a softer, more persuasive sound appropriate for playing con amore,⁶⁶ the music may represent only the singing part, with the guitarist studying it to decide what chord to play. On the other hand, if the music is supposed to represent a singing score, it is odd that it includes no words. Consequently, it is possible that the woman is holding a guitar score which the guitarist is leaning over to decipher, and that he is not about to strum a chord but merely resting his hand on the instrument. This would explain its distance from the sound‐hole.⁶⁷

At the top centre of the painting is a herm. In that it appears to be surmounted by a bust with portrait‐like qualities, it is unusual in known compositions by Watteau. Its full beard and hairstyle make it unlikely that the portrait, if that is what it is, is of a contemporary of Watteau, and in any event, when he did incorporate contemporaries into his compositions, it was in the form of human beings, not statues. If the herm is meant to represent anyone, it is most likely to be of an ancient, or, less plausibly, a sixteenth‐century, personality, perhaps a philosopher or poet. Pythagoras has been suggested on the basis that he was said to have discovered a musical scale based on a mathematical ratio, but no supposed representation of him is sufficiently like. No other ancient can convincingly be said to have been represented by Watteau in NG 2897, although the Greek poet Anacreon might have the best claim.⁶⁸ The term is of a type of which examples by Antoine Girardin (born 1647?) and by Jean Raon (1630–1707) still exist.⁶⁹

Date

Watteau’s paintings are generally difficult to date accurately because there are so few of certain date against which they can be compared. The date of 1717–18 has been proposed for NG 2897 on stylistic grounds.⁷⁰ A consideration of its date needs to take account of the drawings related to it. Admittedly, their use in assessing the date of the painting is limited for two reasons: the dating of Watteau’s drawings is uncertain, and most drawings by him were not made for a specific painting, so that, even if a drawing could be securely dated, it could not be stated with confidence that a related painting closely followed. Subject to those important qualifications, some suggestions can be made. Watteau’s drawing after Campagnola(?) (fig. 11) has been dated to 1715–16,⁷¹ and if it is right that the compositional concept of the guitarist and his companion was influenced by it, then this drawing provides a terminus post quem. All the drawings by Watteau directly connected to NG 2897 (see Drawings A (1–9), above) have been dated by Rosenberg and Prat to within the years 1715–17 except for RP 591, dated by them 1717–18, but here about 1717. Although as a general proposition it cannot be confidently asserted that a painting by Watteau was close in date to a related drawing, it is here proposed that if it can be shown that a drawing was used for only a single known composition, it is more likely that the painting is closely contemporary with the drawing.⁷² The more drawings that [page 558] fulfil the condition of single use the greater the likelihood that the connected painting is contemporary with them. There are four sheets connected with NG 2897 which, based on Watteau’s known compositions, contain figures that for the reasons given above in relation to drawings here accepted as autograph, are connected only with the National Gallery painting. The first is RP 470 (fig. 4, see Drawings A (3)), there dated about 1717–18. The second is RP 498 (fig. 6, Drawings A (5)), previously unconnected to any known composition and there dated 1716–17. The third is RP 500 (fig. 7, Drawings A (6)) also there dated about 1716–17. The fourth sheet is RP 503 (fig. 8, see Drawings A (7)) there dated 1716 or 1717. Based on the dating of these four drawings in RP, itself admittedly a matter of conjecture based on stylistic comparisons, a date of 1717 or the following year for NG 2897 seems likely.

The question then arises whether such a date for NG 2897 is consistent with the likely date for the (located) painting with which it shares the figures of the guitarist and female companion, namely La Récréation galante in Berlin (fig. 1, Paintings A (1)). That painting had generally been dated 1717–18. Rosenberg has opined (without reasons) that it is later than La Gamme d’Amour, and such an opinion would sit well enough with a hypothetical date of 1717 for the London painting.⁷³ However, Christoph Vogtherr has recently proposed dating the Berlin picture to between 1717 and 1719, and has convincingly argued that it is likely to be later than the lost Verrue composition.⁷⁴ He further considers, principally on the basis of the large number of pentimenti in the Berlin picture, that it preceded the London painting.⁷⁵ Vogtherr has noted that the left half of the Berlin picture, that is, the half which includes the guitarist and his female companion, was according to X‐radiographs originally a different composition with smaller figures, and that at a late stage in working out the composition of the picture as a whole, he overpainted the left half to include a guitarist like that in the Verrue composition and his female companion (as well as other alterations in this area).⁷⁶ He has also noted that the pose of the guitarist’s companion (hereafter called ‘the singer’ although she is not necessarily holding a songbook) makes more sense in the Berlin picture on account of the step, which is part of a wide platform.⁷⁷ This gives the singer’s proper left knee a place to rest, whereas in NG 2897 the lack of a step renders the singer’s balance awkward, if not precarious. All of this, it is argued, points to the chronological precedence of the Berlin painting. In addition, the scale of the group of the guitarist and singer in NG 2897 is virtually identical to that of the same group in the Berlin picture, suggesting that some mechanical means, possibly an oil counterproof, was used by Watteau to transfer the design from one canvas to the other.⁷⁸ A visible pentimento in the Berlin picture (fig. 17) appears to confirm that the transfer was from the Berlin picture to NG 2897. This shows that the guitarist’s jacket cuff originally extended the length of his forearm (as in RP 479, fig. 5), but that this was changed so that it extended only halfway along his forearm (as in RP 591, fig. 10). There is no such pentimento in NG 2897 in which the cuff always extended no more than halfway along the forearm. This suggests that if the design was transferred by way of counterproof, the counterproof was taken after the change to the cuff in the Berlin picture.⁷⁹

Is the supposition that the Berlin painting was the earlier of the two consistent with those paintings’ respective relationships with the related drawings that they have in common? Certain details in some drawings are closer to the London painting than to that in Berlin, suggesting that the London painting was the earlier. For example, in RP 522 (fig. 9) the singer’s headdress has no gap between its horizontal and vertical components, and in this respect the London picture is the closer; likewise, her earring in the drawing appears in the London picture but not in the Berlin picture; and her cleavage which is apparent in the London picture and in the drawing is not visible in the Berlin picture.⁸⁰ However, if these points suggest that the London painting came first, other drawings and details of the paintings suggest that the contrary is true. Thus the guitarist’s costume has no ruff either in RP 479 (fig. 5) or in NG 2897, whereas one exists in RP 591 (fig. 10) and in the Berlin picture. In addition, the waviness of the singer’s hair in RP 522 is close to the Berlin picture, whereas in NG 2897, it is more tightly pulled back making it straighter; the highlights on the guitarist’s beret in the Berlin picture are close to the white chalk highlights in RP 479, and the construction of the beret itself in NG 2897, with its radiating folds attached to a central, circular retaining button, is unlike that in the drawing, in which the loose, rather floppy construction of the beret is repeated in [page 559] the Berlin picture; finally in both RP 479 and in the Berlin picture the guitarist’s right thumb is playing, or on the point of playing, the instrument. In NG 2897 its position, which is slightly higher in relation to the guitar, is ambiguous – it could be merely resting on the instrument rather than about to play.

Thus, neither the paintings themselves, nor their relationship to the related drawings, tells us conclusively which of the London and Berlin pictures was first, and if one might be tempted, based on the evidence above, to lean towards the chronological precedence of the Berlin picture, the infrared reflectogram of NG 2897 (see Technical Notes, above) suggests the opposite. Since the figures of guitarist and singer in both paintings are in the same direction, if an underdrawing was indeed used to transfer the design of one painting to another, a support would first have had to be applied to the wet paint of the first painting, so producing a reverse image. Then the still‐wet counterproof would itself have had to be applied to the canvas of the second painting in order to right the reversed image. The infrared reflectogram of NG 2897 shows that the underdrawing for all the figures was made with paint and brush, a procedure which would have been unnecessary had the main lines of the figures already been applied by way of a counterproof, and indeed nothing in the reflectogram suggests that a counterproof, or any other form of mechanical transfer, was used to make the underdrawing.⁸¹ On the contrary, the underdrawing of the figures in NG 2897 is consistent with its having been used to create a counterproof. Furthermore, it is clear from the reflectogram that the position of the guitarist’s legs, principally his proper right one, was worked out on the canvas (the right leg of the guitarist at the right of RP 479 (fig. 5) having been conceivably the starting point, given the similarity of the angle of the underdrawn line with that leg). The reflectogram also shows that a virtually identical change was made to the angle of the singer’s ear in NG 2897 (fig. 18) as was made in the Berlin painting (fig. 19), bringing both into closer correspondence with the ear of the woman at the top left of RP 522 (fig. 9). Hence, even if the Berlin picture as a whole was started before that in London, it cannot with any confidence be asserted that the figures of guitarist and singer in the Berlin picture were completed before their London counterparts. If anything the technical evidence of NG 2897 suggests the reverse, and it is most likely that Watteau was working on both paintings at more or less the same time, at least so far as the figures of the guitarist and singer are concerned, turning from one canvas to the other and back again and making small adjustments as he did so.⁸² If, therefore, a date of 1717–19 is right for the Berlin picture, so it is for NG 2897. That date bracket can, however, be narrowed to 1717–18. That is because 1718 is the date of Nicolas Vleughels’s Apollo, Melpomene, Calliope and Clio, a composition now known through Edmé Jeaurat’s print (possibly in reverse) made a year later (fig. 20).⁸³ Watteau and Vleughels shared lodgings in 1718–19, and it is surely no coincidence that the seated figure of Calliope holding a book in the latter’s painting is, save for the direction of her gaze, so close to the figure of the singer [page 560] in the paintings in London and Berlin. The date of 1718 is therefore proposed as the terminus ante quem for the Berlin and London pictures.

Critical Appreciation

The Goncourts alluded to NG 2897, or at least to Le Bas’s print after it, in an evocatively erotic passage in ‘La philosophie de Watteau’, published in the 7 September 1856 issue of L’Artiste (and reprinted in 1860 with a few adjustments and with Caylus’s biography of Watteau added in the second instalment of L’Art du XVIIIe siècle): ‘All the fascination of women in repose: the languor, the idleness, the abandonment, the mutual leanings one on the other, the outstretched limbs, the indolence, the harmony of attitudes, the delightful air of a profile bowed over a lute, studying the notes of some gamme d’amour, the breasts’ receding, elusive contours, the meanderings, the undulations, the pliances of a woman’s body: the play of slender fingers upon the handle of a fan …’⁸⁴ Paul Mantz called NG 2897 a ‘beau tableau’ when he saw it exhibited in 1874,⁸⁵ and the title of the painting was among the chefs‐d’oeuvre of Watteau inscribed on the monument begun in the 1860s and inaugurated in 1884 in the artist’s native town of Valenciennes.⁸⁶ Lady Dilke was enthusiastic for the painting itself: Although it has suffered somewhat in parts, this beautiful work … is nevertheless a rare example of the master at his surest and strongest. The wonderful red which flashes like the gleaming of jewels from the hair and breast knots worn by the lady, is supported by the duller crimson of her lover’s cap and cloak and the rich purples and golden browns of her own dress, whilst pale pinks and blues and yellows play fitfully over the broader masses of colour, the hues worn by the secondary figures serving merely to break the uniformity of tint in the background. The painting of the flesh and the drawing of the hands show each touch running freely from the brush, yet defining every form with magical certainty of accent, and present the supreme attraction of Watteau’s most spontaneous art.⁸⁷ Lady Dilke had possibly first seen NG 2897 in 1895 when it was auctioned at the Lyne Stephens sale. There, according to one report, ‘the auctioneer expressed a hope that the picture would find a place in our National Collection. The bidding started at 1,000 guineas, and the picture was bought for 3,350 guineas by Messrs Agnew’.⁸⁸ There would have been another opportunity to see it when it was exhibited at the 1898 exhibition of French pictures at the Guildhall, London. In its review of the exhibition The Times described NG 2897 as ‘Marvellously rich in colour, splendidly drawn and painted, this brilliant little picture must rank among the best of the smaller works of this great master, whose pictures are now almost unattainable’,⁸⁹ and the French art critic and writer Louis de Fourcaud commented: ‘This masterpiece has suffered somewhat over the years; it has lost nothing of the spiritedness of the brushstrokes, of its expressive design, of the warm harmony of its wondrously crystallised colour’.⁹⁰

When NG 2897 was bequeathed to the National Gallery in 1912, the Burlington Magazine wrote that it desired to draw special attention to the circumstances of this bequest, seeing that the absence of any painting by Watteau had long been felt to be a reproach upon the gallery. For munificence, so carefully considered, and appropriate, the thanks of the nation should be increased proportionately, and we venture to express our hope that the excellent example of Sir Julius Wernher may be followed by other wealthy owners of works of art, who may desire to become benefactors to the National Collections.⁹¹ Early in 1914 the National Gallery’s French room, show‐ing French art of all centuries, re‐opened following a rearrangement. One wall was devoted to seventeenth‐ and [page 561]eighteenth‐century pictures. La Gamme d’Amour, to this day the only painting by Watteau in the Gallery, was hung in the middle in the ‘place of honour’.⁹² From time to time since then, doubts about the autograph status of NG 2897 have been expressed (see under General References), but it is now universally accepted as autograph.⁹³

General References

Goncourt 1875, no. 136 (‘de la plus belle qualité’); Phillips 1895 (as 1718–19);⁹⁴ Temple 1898, p. 4 (reproduced); Fourcaud 1905, pp. 113–14; Foster 1905, vol. 1, p. 103 (as by Watteau, and in the Wernher collection, and said to be illustrated in the Edition Royale of Foster’s book); Zimmermann 1912, no. 124 (among the works of doubtful attribution, although the author was apparently judging from a poor‐quality reproduction); National Gallery: Descriptive and Historical Catalogue 1913 (as about 1717);⁹⁵ Dacier and Vuaflart 1921–9, DV 199; Réau 1928, pp. 1–59, no. 127, p. 41; Adhémar 1950, no. 169 (as the end of 1716); Huyghe in Adhémar 1950, p. 46 (as 1718–20); Davies 1957, pp. 222–3 (‘Probably a rather late work’);⁹⁶ Mathey 1959, p. 69 (as about 1718–19?); Laclotte 1965, p. 84; Nemilova 1964, p. 107 (as 1716–17); introduction by G. Macchia in Montagni 1968, no. 161; Camesasca 1971, no. 161 (where, following the opinion of Montagni 1968, the attribution was doubted);⁹⁷ Ferré 1972, vol. 3, p. 82, no. A15 (as autograph); Roland Michel 1981, no. 210 (as about 1717?); Ferré 1984, pl. 15; Wilson 1985, p. 78; Grasselli in Moureau and Grasselli 1987, p. 98 (as 1716–17); Rosenberg in Moureau and Grasselli 1987, p. 107 (as before or after La Recréation galante in Berlin?); Temperini 2002, p. 76 and no. 50 (as about 1716–17); Glorieux 2011, p. 195 (as about 1716–17) and fig. 143 on pp. 200–1; Rosenberg and Prat 2011, pp. 99, 142 and 156; Raymond 2013, pp. 16, 115, 121–3 and 134; Eidelberg online, p. 3 (as autograph).

Notes

78 See note 11. While noting an extremely close correlation in the main group between Watteau’s Fête galante with musicians and a bust of Dionysos (Potsdam) and Les Charmes de la Vie (London, Wallace Collection), such that ‘mechanical means were applied for the transfer’, Christoph Vogtherr has proposed that roughly two years separate the Potsdam picture from that in the Wallace Collection: Vogtherr [page [564]][page 565] 2009–10, pp. 181–3. On Watteau’s use of oil counterproofs see Eidelberg 1977, pp. 173–204. (Back to text.)