Comtesse d’Haussonville

Louise d’Haussonville rests a finger against her chin and looks out from a room crowded with flowers, porcelain and the traces of a social evening. Her silvery dress spreads towards the lower edge of the canvas; behind her, a mirror brings the back of her head almost alongside her face. Everything appears settled. Yet follow the arms, or try to reconcile the figure with her reflection, and the apparent ease becomes much harder to explain.

Jean-Auguste-Dominique Ingres completed this portrait in 1845, after a project extending over three years. ^[1][2] The Frick’s full title is Louise, Princesse de Broglie, Later the Comtesse d’Haussonville. ^[3] The painter’s letters and the first reviews tell a less settled story than its composed surface suggests.

Looking at Louise

The face has a quiet clarity that makes the surrounding abundance seem momentarily incidental. Light passes across the forehead, the bridge of the nose and the chin. The centre parting of the hair gives a firm axis against which the tilt of the head registers. Her eyes meet the viewer without the body turning squarely towards us. The lips are closed, their edges delicately shaded. This is enough to establish an encounter without allowing us to name the sitter’s thoughts.

The hand at her jaw makes that encounter unusually intimate. A finger follows the underside of the face, while the other fingers curl back beneath it. Instead of acting as a decorative addition to an already complete likeness, the hand helps to shape the head’s silhouette. Remove it in imagination and the head seems much less securely held within the composition. The pale forearm leads upwards from a rounded elbow, carrying the eye from the dress towards the face.

Across the waist, the opposite arm follows a different rhythm. A bracelet marks the wrist, and a ring catches the light on the hand resting against the skirt. The fingers descend over the fabric instead of gripping it. The two arms form a broad enclosure in front of the body: one lifts the head, the other closes the figure at the waist. Their smooth surfaces stand out against the short, gathered sleeves, whose little creases and bows make the cloth seem much more agitated than the wearer.

It is worth following each contour before deciding that something has gone wrong. A sleeve partly conceals the connection between shoulder and upper arm; the forearms overlap; the waist draws inward behind them. In these crowded passages, visible pieces of the body do not supply an easy map of the whole. What we can see is how effectively the arrangement closes the figure into a continuous shape. An explanation of its pictorial effect need not settle every question about the pose’s physical reconstruction.

Dress, colour and the surrounding room

The dress occupies more of the canvas than the face and arms together. Ingres gives its upper part sharp, compressed folds, then lets long channels of light run down the skirt. Some are narrow and almost white; others spread into broad, cool planes. Darker creases interrupt their descent. The gathered band at the hem draws these vertical movements together near the bottom edge.

Tinterow identifies the dress fabric as silk taffeta. ^[4] In the painting, its colour shifts towards blue and lilac according to the neighbouring shadows and highlights. The difference between that restless cloth and the even modelling of the skin is especially clear at the elbow. Here a rounded area of flesh meets a tangle of angular folds. Both seem tangible, through quite different kinds of description.

Small touches keep the large cool areas from becoming monotonous. A pale trim edges the neckline. Gold jewellery gives warmer points of light near the wrists and fingers. The red ribbon above and behind the head is the strongest isolated colour accent. Its folds echo the more expansive folding of the dress, but its warmth pulls the eye upwards. Down at the right, a yellow shawl lies over a chair, with a white handkerchief nearby. The shawl occupies relatively little space; it nevertheless changes the balance of a picture otherwise dominated by grey, blue and white.

The room establishes its own pattern of solids and reflections. A gold mirror frame rises beside Louise. Below it, a mantel covered with blue velvet creates a dark horizontal ledge. Porcelain vessels and flowers break its surface into small intervals. The white wall panelling and pale upholstery at the sides contrast with the fringed velvet. One can move from the cool gleam of the dress to the softer blue furnishing and then to the painted brightness of a vase. The variety rewards slow looking without requiring every object to carry a symbolic message.

Ingres’s signature and the date are placed low on the left, inside the chair. ^[3] They are easy to miss because the dress commands so much of that part of the canvas. The location belongs to the painting’s network of carefully observed things: even its declaration of authorship sits within the room.

The woman behind the title

Louise was Madame de Staël’s granddaughter and the daughter of the duc de Broglie, a minister under Louis-Philippe. Her husband, Othenin d’Haussonville, was a diplomat. The Louvre’s educational account connects this family background to the painting’s representation of a powerful modern aristocracy. ^[5] The couple met Ingres in Rome in 1840. ^[4]

Such connections matter because the painting was made for particular people. The face before us belonged to a young married woman within a powerful family, and the elaborate room was part of a recognisable social world. Yet neither her ancestry nor the furnishings exhausts her identity. Louise later wrote about Robert Emmet and Lord Byron. ^[4] The Frick notes her literary achievement and explains that her husband required anonymous publication. ^[3]

That later activity makes it unhelpful to treat the painted woman as merely a collection of beautiful possessions. At the same time, the contemplative gesture does not prove that Ingres intended to portray a future author. Nothing in the visible room identifies one particular book or literary undertaking. We can recognise Louise’s life beyond the sitting without making her finger at her chin an illustrated biography.

Her family name can cause another confusion. Ingres’s Princesse de Broglie, painted in 1851–1853 and now at the Metropolitan Museum of Art, represents Pauline. ^[6] She was Louise’s sister-in-law. ^[4] These are two different sitters and paintings, despite their shared family connection and the prominence of luxurious blue or silvery fabric.

Ingres, The Princesse de Broglie, 1851–1853. ^[7]

How the portrait changed

The finished painting conceals the interrupted character of its making. In a letter to his friend Jean-François Gilibert, dated 26 June 1842 in Boyer d’Agen’s edition of the correspondence, Ingres reported that he had sketched Louise’s portrait. It was one undertaking among several: he had just finished a copy of an odalisque, was preparing to visit Dampierre and had another sitting arranged with Madame de Rothschild. The letter places the beginnings of this apparently self-contained painting within a crowded working life. ^[1]

By February 1843, Ingres wanted to begin Louise’s portrait again. He listed its new sketch alongside a copy of the duc d’Orléans, a cartoon of the Archangel Raphael, Rothschild’s head and the composition of The Golden Age. The list does more than explain a delay. It shows the portrait sharing his attention with religious work, royal patronage and a large decorative scheme. These projects coexisted on his schedule even when his letters expressed impatience with the demands of portraiture. ^[1]

Louise’s pregnancy also interrupted progress, as the Frick explains. ^[3] In June 1844, Ingres told Gilibert that he could not finish the portrait that year. He was frustrated by the gap between public recognition and work still waiting to be completed. Being honoured by having a street named after him in Montauban had not enabled him to bring this particular commission to an end. ^[1] The family’s travels in 1844 prolonged the interruptions. ^[4] Three years on the calendar did not mean three years of uninterrupted painting before the same model. An early oval oil sketch was overpainted and later recovered by Raymond Balze, according to Tinterow’s account. ^[4] Abandonment and completion belong to the same history.

The catalogue follows changes to pose, costume and accessories; studies carry annotations on light and hair. ^[4] Those particulars help explain why a well-developed drawing need not record the final decision. Pose, costume and illumination could remain open questions at different stages of the same project.

The Frick owns a graphite and black chalk study dated 1843–1844, with a shawl around the shoulders and an additional small skirt study. It estimates around eighty preparations originally, of which fifteen survive. ^[8] That is an estimate of the possible scale of preparation, not a contemporary census. What survives already gives us grounds to imagine repeated decisions; a total for what has vanished would add little certainty about the order in which those decisions were made.

A later witness supplies another piece of the process. In his 1856 account, Théophile Silvestre described Ingres adding a model’s arms to Louise’s unfinished portrait within a few hours. He presented this episode as an exception to the painter’s habit of prolonged revision. ^[9] It supports the use of a model to supplement the sitter, while leaving much unspecified: which studies survive from that session, exactly when it happened, and how the swiftly painted passage was subsequently finished. It offers no basis for assuming that Louise herself posed undressed.

Who drew the room?

Georges Vigne proposed assistant work in a Montauban room study because its furnishings and figure are handled differently. ^[4] The proposal concerns the evidence of a drawing, with consequences for how we imagine the preparation of the painting. A change in handling can invite questions about shared work without allowing us to name the maker of every object in the final room.

In his subsequent assessment of the exhibition, Tinterow emphasised Ingres’s finishing of the figures and costumes in these late female portraits, while accepting assistance with architectural elements and furnishings. ^[10] The distinction between preparing an interior and executing a whole painting matters. An assistant’s possible work on a room study does not displace Ingres as the painter of Louise. Equally, the portrait’s consistency does not require us to imagine that every preparatory task was performed alone. A workshop could help organise what the canvas would contain while the principal figure remained the focus of the artist’s own labour.

Ingres’s relief, and his remaining doubts

A letter to Charles Marcotte gives an unusually direct glimpse of the painter’s response to the finished work. Henri Delaborde printed an extract in his posthumous book on Ingres in 1870, dating the letter 28 June 1845. Ingres described the portrait as having tormented him, then reported a complete success during four days of showing it in his studio. Family and friends approved, with Louise’s father foremost among them. Nevertheless, he still felt that he had failed to capture all the sitter’s graces. ^[2]

This combination of public relief and private dissatisfaction is more revealing than a simple story of triumphant completion. The people whose approval mattered most were initially the sitter’s family and their circle. Their response could reassure him without bringing his own judgement to rest. The smoothly continuous surface gives little hint of that difficulty, but his account makes its ease seem an achievement reached through uncertainty.

Delaborde’s text transmits a letter rather than reproducing the original manuscript in full. It can therefore support what Ingres is reported to have said at that moment, without turning the passage into an exhaustive explanation of his feelings or his relationship with Louise. In particular, admiration for the sitter is insufficient evidence for a romance. Its subject is the difficulty and reception of a portrait. ^[2]

The letter to Gilibert of 27 July 1845 returns to the same combination of success and pressure. Ingres reported approval from the duc de Broglie, his family and their friends, and said that the portrait would remain available to see until his return, when it would be varnished. In the surrounding passage he also described the continuing demands of work at Dampierre and further portraits awaiting completion. ^[1] The repeated complaints were part of correspondence with a friend; they should not be read as a simple measure of how much care he gave this canvas. The painted hands and dress make that care visible in another form.

What are the objects doing?

Opera glasses sit on the mantel beside calling cards. A small blue evening bag hangs over a porcelain vessel; flowers stand among the harder surfaces. The Frick interprets the accessories as suggesting a return from the opera and the folded visiting cards as signs of missed calls. This reading connects the objects without documenting a particular evening. ^[3]

The objects also work without that narrative. The scattered cards introduce pale, irregular shapes against the blue ledge. Black opera glasses form a compact, dark mass beside porcelain painted with flowers. Actual flowers rise into the reflected figure, bringing fragile, uneven stems against the smooth surface of the mirror. These changes of scale and texture prevent the mantel from becoming a simple horizontal barrier behind Louise.

There is a suggestive contrast between the apparent casualness of these things and the discipline of their placement. Cards have been left out, a shawl dropped over a chair, flowers gathered in different vessels. Yet they spread the colour and light across the canvas with great precision. The portrait accommodates the impression of a used room within a composition that leaves very little to chance.

The different objects also allow two kinds of looking to coexist. From a distance, a vase may function as a dark upright or a glint of gold beside the sleeve. Close to the image, it becomes an individual object with a handle, a painted surface and flowers above it. Louise’s body holds the composition together, but the room repeatedly encourages the eye to leave her and come back.

A portrait that divided its first viewers

The studio audience was followed by the Bonne-Nouvelle exhibition in Paris in 1846. Baudelaire described eleven Ingres paintings gathered in a separate room, contrasting their display with the commotion of the annual exhibitions. He named the portraits of Bertin, Molé and Haussonville alongside Stratonice and the odalisques. This was an opportunity to consider works from different moments together, beyond the circumstances of their original commissions. ^[11]

Baudelaire valued the portraits as imaginative reconstructions of particular people. He also rejected the commonplace that Ingres’s painting was merely grey, pointing to the force and variety of its colour, and praised his attention to the surfaces and curves of the female body. ^[11] His review helps us understand why the dress and skin could matter as much as a measurable likeness. He saw the painter’s shaping intelligence in the finished image.

Théophile Thoré offered a different response. In the portion of his 1846 review devoted to the Bonne-Nouvelle display, he contrasted Louise’s abundance of flowers and mirrors with the austerity of the male portraits. He admired the bare arm supporting her head, recognised its similarity to the pose of Stratonice, but objected to the clash between the lilac bodice and blue velvet. ^[12] Details that make the room absorbing to one viewer could seem excessive or disruptive to another.

Social judgements were involved too. The Louvre’s teaching essay cites Charles Lenormant’s reading of Louise as an aristocrat formed in a world that valued ideas alongside inherited rank. ^[5] Such a judgement tells us how a critic related the image to contemporary society. It cannot establish that a particular class produces a particular character, or that we can read Louise’s beliefs directly from her furnishings.

At the 1855 Exposition Universelle it hung beside The Virgin with the Host, beneath The Apotheosis of Homer, with Molé as its pendant; the family lent it to Ingres’s memorial exhibition in 1867. ^[4] These public appearances gave a private portrait a place within the broader account of Ingres’s achievement.

Charles Blanc’s 1870 book admired Louise’s easy grace but described the grey and lilac atmosphere as dreamlike, weakening the impression of life. He contrasted it with the firmer, more palpable presence of earlier portraits, including Madame de Senonnes. ^[13] His judgement remains a judgement, not a diagnosis of Ingres’s eyesight. Read beside Thoré and Baudelaire, it shows how the same passages of colour and modelling could sustain incompatible evaluations.

The mirror: another view, another puzzle

Look behind Louise’s head and the mirror offers the back of her coiffure, the red ribbon and a sweep of neck and shoulder. The reflection lets us see what the principal figure withholds. Yet its position is disconcertingly close to the face. The two heads almost form a pair, each turned away from the other, while the reflected neckline opens a second pale shape against the wall.

The gold upright at the left helps declare the surface a mirror, and reflected furnishings make the room seem to continue beyond the canvas. But identifying the surface does not make every spatial relation self-explanatory. The front figure, her reflected back and the mantel objects compete for the same narrow area. Our eye readily accepts the reflected hairstyle as descriptive information, then hesitates over the body and the distance between the two views.

This tension is part of the experience of looking. A painted mirror is itself an arrangement of forms; it need not be treated as a diagram from which we can calculate the exact placement of a real sitter and an actual viewer. Nor does uncertainty about the reflection demonstrate that the artist lacked observational skill. The many adjustments of the drawings already show how carefully he considered what the painting would make visible.

Compare the later seated Madame Moitessier. There, the reflected face supplies a profile against the main view, and the hand at the head opens into a much broader gesture. The reflection seems to give the sitter a second direction of attention. Here, the back of Louise’s head remains anonymous beside the face that addresses us: ribbon, hair and neckline take the place of a second expression. The Louvre’s account also connects the mirror motif with Madame de Senonnes. ^[5] Tinterow identifies later responses in Flandrin’s portrait of his wife and Degas’s Thérèse de Gas Morbilli of 1869. ^[4] The portrait’s afterlife includes such transformations of its arrangement, as well as reproductions that sought to preserve it.

Was photography involved?

The Rhode Island School of Design’s drawing, accession 32.227, is understood by the museum as a model for an outline engraving made after the painting’s completion. Its even description across the image belongs to the problem of reproduction. ^[14] Tinterow describes it as traced from a photograph to assist Achille Réveil; he finds no proof of a photographic source for the original portrait. ^[4]

The difference is substantial. A photograph could record a finished painting and help translate it into lines for an engraver. That does not establish that a photograph of Louise supplied the likeness in the first place. Some drawings helped Ingres make the painting; this one helped circulate its image afterwards. The stages of making and reproduction have their own evidence, even when they concern an apparently identical composition.

Ingres, Madame Moitessier Seated, 1856. ^[15]

Ingres, Portrait of Madame de Senonnes, 1814. ^[16]

From family portrait to the Frick

The portrait passed from Louise’s family through Wildenstein to the Frick Collection in 1927. ^[3] The purchaser was the Collection’s trustees. ^[4] The image had outlived its sitter and moved beyond the family setting for which it had been made. Its detailed account of a private room now belonged to a public collection, where viewers with no knowledge of the Haussonvilles could encounter it through the face, gesture and objects.

The frame has a separate history. Nicholas Penny identified an early eighteenth-century style frame in illustrations of the 1846 and 1855 displays. It suited the objects within the painted room and was later transferred to the copy retained by the family. A different carved frame was adapted for the original, giving it a more classical surround. Penny also cautioned that survival of an early frame would not, by itself, prove Ingres designed it. ^[17]

This is a useful reminder that the painting’s presentation has changed even where its composition appears fixed. A frame can echo the room inside a portrait or make the canvas seem more emphatically separate from it. Louise’s image passed from a family setting into a museum, while the boundary through which viewers encountered her changed along the way.