Le Sommeil D'Endymion (blond)

Medium: Oil On Canvas
Size: 55 x 73.4 cm (21.7 x 28.9 in)

Provenance: Nicolas-Charles Count Oudinot, Duke of Reggio, Marshal of France (1767-1847), probably acquired circa 1810; Victor Count Oudinot, 2nd Duke of Reggio, General (1791-1863); Charles Count Oudinot, 3rd Duke of Reggio (1821-1889); Marie Oudinot de Reggio (1850-1932) and by descent.

The image of the sleeping shepherd boy, Endymion, his nude body illuminated by the Goddess Selene disguised as a moon beam, is one of the most renowned among French paintings of the turn of the 18th – 19th centuries. The recent discovery of this brilliant sketch, or variant, in the collection of the heirs of Marshal Oudinot, Duke of Reggio, adds a further dimension to the history of the composition. The extraordinary quality of the heads of the two protagonists, the treatment of the foreground and other particular details serve to indentify this painting, in the opinion of Sylvain Bellenger, as from Girodet’s hand and considerably superior to the only other version, sometimes attributed to the artist but more likely from his studio (Musée Girodet, Montargis). The numerous changes to the composition introduced here, while small, are clearly intended to alter the mood, the time of day and perhaps the season, and are fully consistent with the artist’s somewhat idiosyncratic approach to his art. The change to Zephyr-Cupid’s hair and the wreath are consistent with a date around 1808-1810, reflecting the move away from the more austere neo-classicism of the 1790s, with Zephyr’s tight curls in the Louvre painting more Roman in style.

Conceived in 1791 initially as a painted male academy, Girodet produced a work whose many layered meanings have been explored by art historians and wondered at by connoisseurs for two centuries. Like every student at the French Academy Girodet would have been familiar with the antique Endymion relief in the Albani collection but here the shepherd boy has stepped down from his seat while the elegant greyhound of the classical age has been substituted by a more robust, and perhaps more useful, canine guardian. Girodet also looked to another early source for this composition, a Roman bas-relief displayed in the Villa Borghese in which Silenus visits the reclining Endymion lying beneath a grotto. When exhibited at the Salon of 1793 (no. 296) it caused an immediate sensation and marked out the artist as one of the most brilliant of David’s young pupils. The painting was exhibited again, at the Salon de l’Élisée of 1797, the Salon of 1802 (hors catalogue), and at the Salons of 1814 and 1818. Despite requests from impassioned collectors to sell the work Girodet retained it in his studio hoping always to sell it to the state, finally succeeding when it was acquired by Louis XVIII in 1818, since which time the painting has been on public exhibition in the Louvre. It marks the beginnings of the transition from the austerity of neo-classicism to a more romantic sensibility of which Girodet’s later work, the Death of Atala, is a further example.

The Louvre Sleep of Endymion was engraved several times: by Normand, in a line engraving reproduced in Landon’s Annales (the closest in specific details to the Louvre painting), by Noel, and by Aubry-Lecomte, each of whom included the moonbeam, such a notable feature of the Louvre painting. A loose engraving of somewhat poorer quality that appeared in the Historique Girodet of 1832, plate 52, is also based on the Louvre version; this places Cupid’s bow across just one resting spear (pointing to the left, like no other version of the painting).

While Girodet, still at the French Academy in Rome, may be presumed to have produced drawings and painted sketches directly connected to its production, a fire in his studio has left us with few such images – a fine painted figure of the young man lying against a sketchily brushed in background that gives no hint of the setting in which he would ultimately be placed (Paris, Louvre, 48.6 x 56 cm), and a handful of drawings including one of the winged Zephyr-Cupid pulling back the branches, two drawings for the sleeping dog, and a drawing for his only garment, an ornate pair of sandals (Paris, Bibliothèque Nationale de France). Several small painted replicas have appeared on the market over time but these, despite the occasional competence in their execution, have been rightly dismissed as copies by other hands. Only one other version has hitherto been identified as having been produced under the artist’s direction, a smaller replica (measuring 89.5 x 117 cm) with a few variations that may be seen in the Musée Girodet at Montargis.

The Montargis painting, in the opinion of Sylvain Bellenger, author of the magisterial Girodet exhibition catalogue (2005), may have been executed by Henri Chatillon under the direction of the artist, possibly circa 1810, the date of an engraving by Chatillon dedicated to the artist’s adopted father. This is a finished replica and not a sketch, as sometimes suggested, and includes some small changes, but it lacks any provenance before the early twentieth century and its precise genesis is unknown. Unlike the Oudinot painting the changes do not reflect a significant rethinking of the composition, which aside from the small changes noted here, otherwise replicates the Louvre painting. Cupid’s bow, rather than lying resting across Endymion’s spear, has been moved to the left middle ground as if dropped from his grip while he holds back the branches to allow the moonlight, no longer marked by a beam, to shine through. Selene now admires the recumbent figure from the heavens rather than gently caressing his body, as in the Louvre painting – this represents a significant and obvious amendment which Chatillon, if he painted it, would not have done without Girodet’s direction. The black curls on Endymion’s right shoulder have also been omitted and the branches and leaves to the right of the figure of Cupid modified. Finally he has changed the colour and complexity of the robe on which Endymion lies; in the Louvre painting this has a dark yellow umber tone, with a Greek key pattern in dark brown along the edge. In the Montargis painting, and indeed in the Oudinot version, this has been replaced by a simpler grey-blue coloured robe omitting the decorative border, contrasting more effectively with the leopard skin beneath it.

The differences between the Louvre and the Oudinot paintings, however, are far more striking and rather than make Endymion the prime focus of attention of the viewer, both figures are now almost equally prominent. The artist has given the shepherd boy blond curls instead of the dark locks of the Louvre picture, which falling in a similar fashion are less abundant on each shoulder. Girodet played with changing hair colours in other works; in his portrait of Queen Hortense (Amsterdam, Rijksmuseum) he has painted her as a brunette although she was well-known as blonde haired. The two sketches for the head of Galatea for the Pygmalion and Galaeta of 1819 were given different hair colours, one dark the other blond; for the final version he chose blond hair for the statue brought to life by love. Zephyr-Cupid, in addition to brilliantly hued wings, also has a new hairstyle instead of the tight golden curls of both the Louvre and Montargis paintings (and the many copies of the former); his hair now flows more naturally, the curls looser, and the hint of a smile gives him a look of added mischief absent in the other two works. This impish boy, with muscled arms, strains to hold back the branches whose leaves in part have been only loosely sketched. Notably, as in the Montargis painting, the moonbeam has been removed although the young man’s body is still sharply illuminated. Only one spear lies beside Endymion’s body in the foreground but its sharpened point is emphasized by the moon light, while Cupid’s bow has been removed altogether. The configuration of branches and leaves to the right of Cupid’s body has been simplified, perhaps because of the relatively scale of the work, and are different to the Louvre and Montargis compositions and the engravings connected to both of them. The artist has painted the soles of Endymion’s sandals, which can be seen between his finely painted toes while the side of the sole beneath his left foot (the leg is curled beneath his right) is clearly visible, although he has not bothered to paint the elaborate straps and buckles.

Uniquely in the Oudinot version, the wreath held by Cupid-Zephyr has also been changed – it is no longer made of laurel but of tightly gathered flowers, becoming a more prominent accessory in the composition. As if reflecting in this a change of season, Endymion lies not on the rough earth but on soft, green grass, through which small pale violet flowers are scattered. Endymion rests on a similar robe to the Montargis painting but a small stream flows from the lower right side across one third of the bottom of the canvas, which has been extended on the right side so as to leave more space for the shepherd’s left hand (although this part remains rather sketchy), that now hangs at a more natural angle. In the Louvre painting the presence of water is only hinted at by a barely visible fleck of white towards the bottom right corner although the 1810 engraving by Chatillon includes some wavelets suggesting a stream. These changes together suggest a change of season, or mood of which there is no evidence in any of the engravings.

The Chatillon engraving of 1810 may be connected to the Montargis painting, but nonetheless includes the moonbeam although otherwise following the Montargis version closely. If indeed the Montargis painting is the work of Chatillon under the direction of Girodet, one must assume that Girodet decided that the moon-beam should be reinstated in the engraving. Chatillon’s own connection with the painting did not end there. In 1831, freed from Girodet’s guidance, he produced a large lithograph which made yet further changes – this, like the Oudinot painting has only one spear and Zephyr-Cupid’s face bears a rather crude leer, without the subtlety of expression found in the Oudinot painting. Another, later engraving, by Joubert also follows the Montargis painting but omits the moonbeam, placing the bow beneath Cupid and eliminating the Greek key pattern from the edge of the robe. The engraving by Duchesne follows Joubert with no moonbeam, no Greek key pattern, and the bow on the left but has a more complex tangle of foliage, closer to the Louvre version. These later engravings were produced a decade after Girodet’s death, a testament to the continued popularity of the painting.

Barthélémy Jobert in the Girodet exhibition catalogue (Girodet et la lithographie, Le problem de l’estampe de reproduction) has pointed out that the small differences between the engravings and the paintings they reproduced were far from casual inaccuracies. Aubry-Lecomte, one of the most accomplished engravers after Girodet and a pupil working in his studio, in his engraved version of the Danae of 1798, altered the image in several respects but under Girodet’s specific direction. Girodet was not only a brilliant draftsman but talented engraver himself and, aware of how important accurate representations of his images were to his reputation as well as their broad dissemination, carefully controlled their production. It was Girodet himself who would have determined that the bow was misplaced in his original and should be seen as having fallen from Cupid’s grasp, landing on the ground in front of the resting hound, and insisted on this amendment to the Endymion engravings made in his lifetime.

The Oudinot painting remained unknown, although in the mid-19th century when it was lined the prepared but unpainted edges were extended to enlarge the composition and then painted over (these painted additions have now been removed and the painting framed to its original size). These later interventions sought to disguise the sketchy and sometimes thinly painted areas which, done originally with oil glazes, have in part become transparent. Nonetheless, the remarkable quality and the many changes suggest that Girodet later reconsidered the composition and perhaps planned to make another large scale replica for which this work may have been the first idea.

There is no actual record of Girodet ever having been commissioned to paint a second version of the Endymion, however, which therefore leaves no clear explanation as to the precise genesis of either the Montargis or Oudinot compositions. A clue, however, may lie in a letter from Girodet to M. Mirbel, the agent for Louis Bonaparte, written in 1808 when this younger brother of Napoleon, married unhappily to Empress Josephine’s daughter Hortense, was King of Holland. Evidently Louis Bonaparte had asked his agent to try and purchase the Endymion for his own collection, but Girodet refused Bonaparte saying he was “not free to dispose of it” but that he would instead make for the King “not a copy, but a repetition of this painting with a few improvements (améliorations) in the accessories or again in the principal figure, and which will emphasize the originality of this repetition. If His Majesty deigns to agree to this project, I shall exercise it with the zeal inspired by the kindness with which he has always honoured me and because I have always been jealous of my reputation, and I will go immediately to work after this year’s Salon.” Mirbel passed this letter on to King Louis in a detailed report dated 19 April 1808, in which he also discussed the King’s proposed acquisition of several other works. In his report Mirbel noted that the repetition “would superior to the first” and that the character of Girodet was such that he would not do this simply for commercial reasons.

Girodet’s work load was considerable, resulting in a failure to complete several promised compositions, so this project may have got no further than a preparatory sketch to be shown to the King. There was ample precedent for such a submission, both in the oeuvre of Girodet and other 18th and 19th century artists offered state or royal commissions. Girodet had produced four high quality sketches for the series of the Four Seasons done in 1800 for the Casita del Labrador at the Palace of Aranjuez, for King Charles IV of Spain, which differ in many small details from the works finally installed in the king’s cabinet. These sketches, now in a private collection, belonged to Charles Percier, who designed the room, but were probably intended for the approval of either architect or king before the project commenced. In 1810, when given the commission to paint the Revolt of Cairo, he made a full scale composition sketch on paper (30.8 x 45.1 cm, Chicago, Art Institute) for the approval of Vivant-Denon who was in control of the overall project. This sketch also differs in numerous details from the final work. When commissioned by the crown in 1816 to produce posthumous portraits of Général Cathelineau and the Marquis de Bonchamps he again produced finished sketches, probably submitted to Count de Forbin for approval so they would conform to the proportions of the other four portraits ordered for the king’s gallery of royalist heroes, and once again there are small but obvious differences between the sketches and the final works.

We do not know whether Girodet was indeed given the go ahead to make this “repetition” but Louis Bonaparte is known to have been desirous of acquiring works by his contemporaries – in the same year, 1808, he had tried but failed to purchase the large scale Belisarius (exhibited at the 1795 Salon) by François Gérard, which the latter instead sold for a greater price to Louis’ brother-in-law and adopted nephew, Prince Eugène Napoléon (Eugène de Beauharnais). Girodet had also refused to sell the Endymion to the wealthy Italian collector of contemporary art Giovanni Battista Sommariva, who had acquired the small version of the Belisarius to which Gérard had made several small, but significant changes. The Belisarius was directly connected to the Endymion; the head of the latter, painted in 1791, is precisely identical in features and position, to the head of the great Roman General’s boy guide, who lies in his master’s arms, his head back. Yet, while Endymion is merely sleeping, the boy in the Belisarius has been bitten by a snake, his life quietly slipping away. Perhaps Gerard, a friend but also a rival of Girodet, intended this deliberately, killing of the boy who had inspired in Girodet such an extraordinary masterpiece.

There was an indirect relationship between Louis Bonaparte and Marshal Oudinot; in 1810 the former decided to give up his Dutch throne and the governorship of the Kingdom was instead consigned to Oudinot, who in 1810 had been the beneficiary of a large government endowment. Could the Oudinot sketch have been submitted to Louis Bonaparte for his approval and then perhaps given to the Marshal?

Knowing how jealously Girodet guarded both his reputation and the sanctity of the engravings after his work, this painting must be given to the artist himself. Although in some respects more complete than other preparatory sketches by the artist, unlike the Montargis painting it was clearly not intended to be presented as a finished work, suggests that it was intended as a model for the production of yet another version, perhaps the work promised to Louis Bonaparte. Sylvain Bellenger has proposed that it may alternatively have been produced by the artist for his own pleasure, placing the sleeping shepherd boy in late spring, which explaining the grassy bank on which he lies, the flowers on the ground and the wreath. Were these changes the very “améliorations” that Girodet offered to make in the repetition for Louis Bonaparte? This would explain why they would be sufficient to identify the work as an “original”, but not dramatic enough to change the sensibility and character of the subject itself.

original size: 52 x 70 cm

Painted circa 1808-10