The Tree He Cut Down

Behind the head of the descending god, crossing the blue, there is a palm crown: long yellow-green and dark-green sweeps radiating from a single point, a dozen movements of a loaded brush. No trunk is given it — the crown simply emerges from the dark mass behind the god — and it occupies the upper right as a decorative spray, half foliage, half flourish. In the great canvas in the Wallace Collection for which this small one is the trial, it is gone. The Wallace’s own account of what changed between sketch and finished picture says that Boucher removed “a distracting palm tree near Mercury’s head to draw attention to the traditional pyramidal composition”.¹

That deletion is the most useful fact about this painting, and the reason to take the small canvas seriously rather than treat it as a rough draft of a better thing. A picture built on a hidden programme does not lose a palm to a pyramid. Boucher sacrificed it because it was in the way of a shape, and everything else in this dark, coral-flecked little canvas follows from that.

The Picture

The composition is a diagonal that resolves into a triangle. Mercury comes down from the upper centre-right on a grey-white cloud, his right arm flung up and out to the left — the highest point of the picture, an open hand, fingers spread against dark rock — his left reaching down to the right, where the caduceus lies across the paint as a dark brown rod with pale yellow-ochre serpent-loops drawn over it and two yellow wing-strokes at its head. He wears a coral-red winged cap over dark, barely indicated hair. His head is bowed and turned down to his own right, towards the viewer’s left: he is looking at the women below him, not at us. Two bare legs, the toes of both touched in coral, emerge from the cloud, and where they join the body nothing is resolved; the chest is soft and rounded, the anatomy generalised rather than emphatically male.

Below, five women lie along a riverbank in a frieze running from the left edge to the animals at the right. At the far left a nude seen from behind leans in toward a second woman whose head is tipped back so far that her face is upside-down to us, a yellow-gold drapery over her shoulder. Then the nurse: seen from behind, leaning into the group, her hair a heavy golden-blonde mass gathered into a thick plait, her chemise slipping from one shoulder. The infant is in her arms, chubby and pale, wrapped in white, turned into her body in the attitude of a child at the breast — though the breast itself, like his face, is turned away from us. He is not being displayed, and he is not being handed upward to the god.

The pivot is the fourth figure, reclining at the centre with a green-grey leafy wreath in her chestnut hair and a great sweep of strong blue across her lap. Her face is tilted up and back, the whites of her eyes showing, her mouth slightly open: she is the only one of the five who looks at Mercury. An arm reaches across at shoulder height between her and the fifth nymph, whose head is thrown far back in her turn, and where the hands meet sits a small knot of yellow-gold paint. It may be a vessel, it may be cloth; the paint does not decide, and it would be dishonest to pretend otherwise.

The foreground is where the landscape argument is made. A golden jar lies on its side at the bottom centre, its mouth turned toward us and tipping right, white water pouring from it into a shallow run of water along the lower edge, rushes beside it. Bare legs cross the foreground, the toes accented in hot coral — an accent Boucher repeats at every extremity in the picture, at fingertips, at toes, at a nostril and the corner of a mouth. At the right, behind the last nymph, a warm russet-brown mass the size of a figure fills the middle ground: a rounded, coral-edged contour that could be an animal’s flank, a drapery or another body, and the reproduction will not decide. Beyond it a goat’s head and shoulders emerge from the dark, turned away to the right — two pale horns swept back, a dark eye, a red-brown muzzle, a long white-grey beard, perhaps a dozen strokes. At the upper left a white cascade falls down the rock in three or four strokes, as bright a white as anything here.

There is more in the shadow than one sees at first: at the middle left, in unlit brown monochrome, a female torso with the breasts modelled, an arm extending right, and beside it what looks like a second head. Cincinnati’s record counts “five female nudes representing the Nymphs”,² which is right for the lit figures; whether the brown ghosts behind them were meant as further nymphs, or as the ground the group emerges from, a photograph cannot settle.

A fine, dense, net-like craquelure runs over the whole surface, most legible across the pale flesh and the grey cloud; the darks at left and lower right have sunk, so that foliage and rock read now as near-black masses; the canvas weave shows through the thin passages; the whites stand proud. Single loaded strokes do the work of a leaf, a horn, a fold.

Making

Boucher painted this on a canvas of 59.7 by 73.7 centimetres for a picture that would end up 230 by 273.³ Cincinnati calls it “a modello, or preparatory oil study”.⁴ The word covers two things: a sketch in which a painter finds out what he thinks, and a presentation piece by which a patron is shown what he will get. The paint favours the first, and so, in more technical language, does the Wallace: the sketch “appears to predate The Abduction of Europa due to its fragmented brushwork, more straightforward use of color, and some anatomical inaccuracies, such as Mercury’s legs”.⁵ The pendant tells the same story in reverse: in the surviving sketch for Europa, Boucher borrowed the angle of the bull’s head from Natoire and then changed it to a frontal pose in the final picture.⁶ Twice over, in one commission, the sketch preserves a first thought the big canvas throws away. That, and not the finish, is what these objects are for.

No published technical examination of either the Cincinnati sketch or the Wallace canvas could be traced: no X-radiograph, no infrared reflectography, no ground or pigment analysis, no lining history, no condition note from either museum. What can be said must therefore be said comparatively. On a large Boucher of 1740 in Kansas City, examination found a double ground — reddish-orange beneath off-white — no underdrawing detectable by infrared or magnification, extensive wet-over-wet application indicating completion over a short period, and thin fluid paint of low impasto.⁷ On a small Boucher of 1759 in London, a grey ground on fine-weave canvas, with X-radiography showing him adjusting a contour as he went.⁸ That is the technology one would expect here — probably no drawing underneath, everything decided in the paint, wet into wet, at speed — but nothing about this canvas has been examined, and that is an inference from other Bouchers, not a finding about this one. The same caution governs the blues. Prussian blue reached Paris around 1710 and was the only blue found in all twenty-five analysed pictures by Pater dated 1728 to 1736;⁹ the Wallace says it “forms the basis for Boucher’s foliage”.¹⁰ Nothing here has been sampled.

One drawing for this picture can be traced, and it has had an interesting career. In the British Museum there is a Head of Mercury in red and white chalk on buff paper, 150 by 137 millimetres, dated 1734, catalogued as “for a painting of the Mercury confiding the Infant Bacchus to the Nymphs”, the god “looking slightly left and downwards, wearing a winged hat”.¹¹ That is precisely the head in the Cincinnati canvas. For decades the sheet was filed under Boucher’s much later picture of the same subject; the British Museum has redated it on the grounds that the 1769 Mercury is in profile and bare-headed and that the sheet is stylistically earlier. Red and white chalk on buff paper is trois crayons minus the black — the Watteau method Boucher had spent his twenties reproducing, having etched over a hundred of the roughly 350 plates of Jullienne’s Figures de différents caractères.¹²

The Commission

The pictures were painted for François Derbais, an avocat au Parlement who died in 1743, and hung in the billiard room and on the staircase of his hôtel in the rue Poissonnière. The Louvre’s record of another canvas from the group states it flatly: painted in 1732 for the decoration of the salle de billard of Derbais’s Paris house.¹³ An inventory of his goods taken on 2 March 1743 lists the Aurora and Cephalus and the Venus and Vulcan in that room.¹⁴ By about 1735 he owned at least eight large Bouchers, and the painter took his own prospective clients there to look at the walls: the lawyer’s house was his showroom.¹⁵

He was not a connoisseur who had bought his way in: his father Jérôme was marbrier ordinaire des Bâtiments du Roi, the King’s marble contractor, and he grew up inside the royal works.¹⁶

The claim that Boucher did all this for nothing is repeated on every modern label, and it deserves looking at. The earliest source is Pierre-Jean Mariette: Derbais had filled his whole house with Bouchers, “which had been very easy for him”, Mariette relates, “for Boucher, seeking at that time only to make himself known, would, I believe, have done them for nothing”.¹⁷ That is the conditional, hedged with je crois. He did not say no money changed hands; he guessed at a young man’s motives. What has hardened into “painted without fee” is a mid-eighteenth-century supposition, and no contract, receipt or price for the Derbais pictures can be traced.

The guess is nonetheless plausible, because of where Boucher stood in 1734. The Salon did not become a regular public fixture until 1737, and he did not exhibit before that year; his tapestry work for Beauvais began in 1736.¹⁸ On 30 January 1734 he was received into the Académie with Renaud et Armide.¹⁹ In the year of this sketch, then, a history painter of thirty had two theatres available to him, the Académie’s chamber and a private wall, and he used both at once. The Derbais canvases are not decoration that happens to be large. They are an application, at royal scale, hung in a room where men played billiards.

Reading

Museum labels for this subject cite Ovid, and Cincinnati’s is no exception.²⁰ Ovid is not the picture’s source. In the Metamorphoses the child is torn from the burning Semele, sewn into his father’s thigh, reared in secret by his aunt Ino, and then “handed over, the Nyseian nymphs hid him in their caves and gave him the nourishment of milk”.²¹ There is no Mercury in it. The courier who fills the top third of Boucher’s canvas comes from Apollodorus, where Zeus, evading Hera’s anger, turns Dionysus into a kid, and then “Hermes took him and brought him to the nymphs who dwelt at Nysa in Asia, whom Zeus afterwards changed into stars and named them the Hyades”.²²

That one sentence supplies three things the picture needs, two of them found nowhere else. It supplies Mercury — as, in their different ways, do Diodorus and Nonnus, though Apollodorus is the tersest. It supplies the goat — the kid into which the child is changed, at exactly the moment Boucher paints, though Boucher has painted a bearded he-goat rather than a kid and may simply have been enjoying the animal. And it supplies the sequel the picture does not show: these women become the rain-stars in the head of Taurus. Boucher gives them a waterfall, a spilled water-urn and a run of water. That observation is mine and is unpublished; whether it is design or landscape habit cannot be demonstrated, and no eighteenth-century source connects the two. But the future rain-stars are here surrounded by water.

What is most striking is everything Boucher declines. Set the canvas against the tradition: the Roman sarcophagi of the infancy of Dionysus give the nursing nymph, Silenus, the first bath, a panther cub, satyrs and maenads;²³ both of Poussin’s treatments give the grapes squeezed into the bowl and the goat, and the National Gallery names the seated figures in its picture as the foster-parents Ino and Athamas.²⁴ Boucher’s own version of 1769 restores the putti, the grapes, the thyrsus and an eagle with a lightning shaft.²⁵ In everything the picture of 1734 lets us read there is no Silenus, no satyr, no maenad, no panther, no thyrsus, no grape, no vine, no putto, no bath. He keeps precisely what Apollodorus gives him, a courier and a goat, and throws out the entire apparatus of wine. The subject becomes the handing over and the nursing, and nothing else. The argument is mine and no published statement of it could be traced, and two things must be conceded with it. The 1769 picture restores everything the sketch leaves out, so the austerity was not a principle Boucher held for life. And the nearest precedent is one his own teacher set: Jean-François de Troy, with whom he briefly studied, signed and dated a Bacchus enfant confié par Mercure aux nymphes du mont Nysa in 1717, in the same landscape format.²⁶

Beneath the Surface

The picture offers a great deal to an esoteric reading, and it is worth setting out how much of the offer is real.

Mercury as psychopomp is ancient and uncontested: he is the god who conducts souls. A picture in which he descends from heaven to deliver a newborn into the world of bodies is, formally, a descent-of-soul image, and the Neoplatonic tradition made Dionysus the emblem of exactly that descent — Plotinus is the locus classicus, “the souls of men, seeing their images in the mirror of Dionysus as it were, have entered into that realm in a leap downward from the Supreme”.²⁷ Edgar Wind called Hermes “the mystagogue … who guides the souls from outward appearances back to the inward One”.²⁸ The observation about Boucher’s composition is this writer’s own and unpublished; the material behind it is genuine, and entirely about Florence and Rome around 1500.

The Orphic reading is weaker. The Zagreus story concerns a different Dionysus, born of Persephone, and its scene is a dismemberment rather than a delivery. And the composite Orphic myth has itself been argued to be a modern construction: Radcliffe Edmonds holds that no ancient source before the sixth century combines all its elements, and that the doctrine of Orphic original sin was built on a Christian template by nineteenth-century scholars.²⁹ An Orphic reading here would rest on something no one in the 1730s held.

The alchemical resemblance is the most seductive and the most instructive. Michael Maier’s Atalanta fugiens of 1617 has an emblem whose motto is “Nutrix eius terra est”, the earth is his nurse, and whose pictura shows the philosophical child nursing at the terrestrial globe, flanked by Amalthea, the goat who suckled Jupiter, and the she-wolf of Rome.³⁰ Boucher’s canvas has a divine infant in the attitude of nursing, a horned goat, and a caduceus-bearing Mercury on a cloud. The resemblance is looser than it looks — Maier has no nurses and no Mercurius in the plate, and his child feeds from the earth itself — and it is worth nothing at all: both images inherit the same classical nurse-and-goat topos, and the caduceus that is the alchemists’ sign of reconciled opposites is, in the ordinary emblem books, merely Mercury’s stick. Alciato reads it as eloquence and intelligence rewarded with plenty, and sets it “between the horns of Amalthea” — the horn of plenty, not a live goat.³¹ No alchemical reading of this Boucher, or, so far as can be traced, of any Boucher, has been published. The juxtaposition above is offered as what it is: a demonstration of how quickly the eye manufactures esoteric meaning where none is on offer.

Freemasonry deserves a paragraph because the claim circulates. Boucher appears on at least one published list of eighteenth-century masonic painters, with the lodge L’Amitié at Bordeaux and the date 1765 beside his name; that listing says something narrower, namely that he was the author of a famous masonic drawing for the lodge.³² The object is real and catalogued, a Diplôme des Francs Maçons pour la loge de Bordeaux signed “F. Boucher inv. 1765” and engraved by Choffard the following year.³³ It is a design commission, from an artist who designed for anyone who paid, thirty-one years after this picture. No lodge register and no initiation record could be found, and the Paris lodges’ great public moment, from Ramsay’s oration to the prohibition of 1738, postdates the Derbais commission entirely. One specialist study stands unread between this verdict and a definitive one.³⁴

The plainest statement of the case against all this was made by the greatest practitioner of the method. “The reading of a picture in the light of a text”, Wind wrote in his preface, “may easily destroy both text and picture.”³⁵ There is no evidence of any kind that Boucher or Derbais had an interest in hermetic, Orphic, Neoplatonic, alchemical or masonic matter; Derbais is documented as a lawyer with a billiard room and eight large Bouchers. And the palm is the best evidence there is: a painter working to a coded programme is unlikely to remove an element of his iconography to improve a triangle.

What Isn’t There

Several things commonly said about this picture and its maker do not survive checking. That Boucher painted the Derbais pictures without a fee: a conditional guess of Mariette’s, not a record of a transaction. That Mercury sewed the infant into Jupiter’s thigh — stated on the Wallace Collection’s own public page: no ancient source says so.³⁶ That Boucher said nature was trop verte et mal éclairée, too green and badly lit: a full-text search of HathiTrust for the exact phrase returns 109 full-view results, of which exactly one predates 1860, Charles Blanc’s Les peintres des fêtes galantes of 1854.³⁷ No eighteenth-century source for it has been found. It is a nineteenth-century mot, routinely used to explain pictures painted a hundred and twenty years before it was printed.

Diderot’s demolition of Boucher is quoted against everything he ever painted, and it is a late verdict. The famous sentence — “the degradation of taste, of colour, of composition, of character, of expression, of drawing, has followed step by step the depravation of morals” — belongs to the Salon of 1765 and describes a man of sixty-one who was by then premier peintre du roi.³⁸ Four years earlier the same critic had written of Boucher’s pastorals that “this man has everything, except truth”, and that “this painter is in painting roughly what Ariosto is in poetry”.³⁹ There is in any case no record that he ever saw the Derbais canvases: they hung in a private billiard room and were never at a Salon.

And a correction to the catalogues: Ananoff and Wildenstein’s catalogue raisonné of 1976 filed the British Museum’s chalk Head of Mercury under the 1769 picture, at number 676/7, and the British Museum has since redated it to 1734 and reassigned it here — one drawing and one small canvas, agreeing with each other and disagreeing with the standard catalogue.⁴⁰

Afterlife

The room Boucher filled for Derbais is now in six places: the two great canvases in the Wallace Collection, Venus Asking Vulcan for Arms for Aeneas in the Louvre, Aurora and Cephalus at Nancy, a Triumph of Venus at the Romanian Embassy in Paris, the oil sketch for Europa at Amiens, and the sketch for the Mercury in Cincinnati.⁴¹ Four putto-Seasons served as overdoors, one of them probably in the billiard room itself.

The dispersal can be dated at one point. Aurora and Cephalus and the Venus and Vulcan came back on the market together in the posthumous sale of Claude-Henri Watelet’s collection on 12 June 1786, bought by Paillet for the Crown at 3,121 livres; they went to the Louvre that year, to Versailles before October 1800, and the Aurora to Nancy in April 1801.⁴²

How the two big canvases reached the 4th Marquess of Hertford is not documented anywhere reachable. The endpoint is fixed at 1843, with Lady Wallace’s bequest to the nation in 1897.⁴³ The century between Derbais’s death and Hertford’s purchase is blank. What can be said is that 1843 is eleven years before Blanc’s book and nineteen before the Goncourts’ essay: the collector was ahead of the revival, not riding it.

One near-relation is gone. The Musée des Beaux-Arts de Caen owned a version of the composition, oil on canvas, 58 by 75 centimetres — within two centimetres of the Cincinnati canvas — catalogued as school of Boucher and bequeathed to the city in 1848. The record reads: présumé détruit lors des bombardements de juin et juillet 1944.⁴⁴ Whether it was a copy or a second sketch can never now be known.

Of the Cincinnati canvas’s own history before 1989 nothing is published: no owner, no sale, no price. The credit line records only that it was bought with two endowments and with funds released by exchanging an earlier gift — American museum usage for a deaccession.⁴⁵ The picture Mariette supposed had been painted for nothing was paid for by selling something else.

The Palm

Look again at the spray of green behind the god’s head. I take it for a sign of the East, roughly where the ancients put Nysa — Arabia, Phoenicia by the streams of Egypt, “Asia”, India, and in one account Libya. That reading is mine and published nowhere; no eighteenth-century source glosses the palm, and a palm in a Boucher landscape may be nothing but a palm. It is in any case one of the most enjoyable things here: a dozen sweeps of a brush, yellow-green over dark, thrown across the blue with the confidence of a man who has just found out he can do this.

Boucher painted it out. He chose the pyramid over the palm, the shape over the sign, and in doing so told us more plainly than any document what he thought he was making: not a text to be decoded but a wall to be filled, and filled better than anyone else of his generation in Paris could fill it. The small canvas keeps the tree because the small canvas is where he was still deciding. That is worth more than the finish it was made to serve.

Things You May Not Have Noticed

1. A tree that no longer exists. The palm crown behind Mercury’s head is gone from the finished canvas. The Wallace Collection’s own account of the changes says Boucher removed “a distracting palm tree near Mercury’s head to draw attention to the traditional pyramidal composition”. You are looking at a tree he decided to cut down.⁴⁶
2. The museum’s own cited source does not contain the main figure. Cincinnati cites Ovid. In Ovid the child passes from Jupiter’s thigh to Ino and then to the nymphs, with no divine courier at all. Mercury comes from Apollodorus, Diodorus and Nonnus; the goat and the nymphs’ future as the Hyades come from Apollodorus alone.⁴⁷
3. Mercury’s head has a drawing, and it spent decades filed under the wrong picture. British Museum 1933,1014.22, red and white chalk on buff paper, 150 × 137 mm, dated 1734. Ananoff and Wildenstein catalogued it under the 1769 version; the British Museum redated it on the grounds that the later Mercury is in profile and bare-headed, and that the sheet is stylistically earlier.⁴⁸
4. The whole legend of the unpaid commission rests on a conditional verb. Mariette wrote that Boucher, seeking only to make himself known, would, I believe, have done them for nothing.⁴⁹
5. The patron was a marble-man’s son. François Derbais, avocat au Parlement, died 1743; his father Jérôme was marbrier ordinaire des Bâtiments du Roi.⁵⁰
6. The lawyer’s house was the painter’s showroom. Boucher brought prospective clients to the rue Poissonnière to look at the walls. Derbais owned at least eight large Bouchers by about 1735.⁵¹
7. The subject was his own teacher’s. Jean-François de Troy, with whom Boucher briefly studied, had signed and dated a Bacchus enfant confié par Mercure aux nymphes du mont Nysa in 1717, now in Berlin. Louis de Boullogne showed one at the Salon of 1704; Collin de Vermont painted one in 1725.⁵²
8. A version of this picture was lost in the Battle of Normandy. Caen owned one at 58 × 75 cm, catalogued as school of Boucher, “presumed destroyed” in the bombardments of June and July 1944.⁵³
9. “Nature is too green and badly lit” is a nineteenth-century remark. Its earliest traceable printing is Charles Blanc, 1854. No eighteenth-century source for it has been found.⁵⁴
10. Diderot’s demolition is a late verdict. “The degradation of taste … has followed step by step the depravation of morals” belongs to the Salon of 1765, on a man of sixty-one who was by then premier peintre du roi. Four years earlier Diderot had allowed that Boucher “has everything, except truth”, and compared him to Ariosto.⁵⁵
11. Figures in the dark. Brightening the shadow at the middle left reveals a torso with the breasts modelled, an arm, and what looks like a second head, painted in brown monochrome and never lit. The museum’s record counts five nymphs.⁵⁶
12. It was bought by selling something else. “By exchange” in Cincinnati’s credit line is American museum usage for a purchase funded by a deaccession.⁵⁷

Select Bibliography

Alexandre Ananoff and Daniel Wildenstein, François Boucher, 2 vols, Lausanne and Paris, 1976.

Radcliffe G. Edmonds III, “Tearing Apart the Zagreus Myth: A Few Disparaging Remarks on Orphism and Original Sin”, Classical Antiquity 18, no. 1 (1999), pp. 35–73.

Jo Hedley, François Boucher: Seductive Visions, London (The Wallace Collection), 2004.

John Ingamells, The Wallace Collection Catalogue of Pictures III: French before 1815, London, 1989.

Pierrette Jean-Richard, L’œuvre gravé de François Boucher dans la collection Edmond de Rothschild, Paris, 1978.

Françoise Joulie, Esquisses, pastels et dessins de François Boucher dans les collections privées, Versailles (Musée Lambinet), 2004.

Alastair Laing and Pierre Rosenberg, François Boucher 1703–1770, exh. cat., New York (Metropolitan Museum of Art), Detroit and Paris (Grand Palais), 1986–87.

Ewa Lajer-Burcharth, The Painter’s Touch: Boucher, Chardin, Fragonard, Princeton, 2018.

Pierre de Nolhac, François Boucher, premier peintre du roi, 1703–1770, Paris, 1907.

Katie Scott, The Rococo Interior: Decoration and Social Spaces in Early Eighteenth-Century Paris, New Haven and London, 1995.

Edgar Wind, Pagan Mysteries in the Renaissance, revised and enlarged edn, Peregrine Books, Harmondsworth, 1967.

Humphrey Wine, The Eighteenth Century French Paintings, National Gallery Catalogues, London, 2018.