The Gallery holds the most extensive collection of portraits in the world. Search over 215,000 works, 150,000 of which are illustrated from the 16th Century to the present day.

Notes on frames in the exhibition, Portraits by Ingres

by Nicholas Penny

February 1999

That Ingres would have tried to control the presentation of his paintings and drawings is to be expected from his character, and his interest in picture frames is documented in his letters. However, not many of his portraits remain in their first frames - and we cannot be sure that these few were designed or selected by him or merely approved by him or, indeed, chosen by patron or client without consulting him.

The oval portrait of Madame Rivière (no. 9, Musée du Louvre) which was exhibited at the Paris Salon of 1806 has a neo-classical frame with olive branches in the spandrels and a frieze of anthemion and lotus. It is not an unusual pattern for Paris at that date - Jacob Simon notes that an "exactly contemporary example may be seen in the Wellington Museum at Apsley House on Robert Lefèvres' The Empress Josephine with detailing scaled down to suit the picture but otherwise not dissimilar". Had Ingres actually designed Madame Rivière's frame we would expect it to match those around the portraits of her husband and daughter.

The portrait of Ingres's close friend Gilibert of 1804-5 (no. 5, Musée Ingres, Montauban) is unusually framed in what appears to be a heavy slip from a larger frame, part-regilded. It has an elegant frieze of miniature foliate design in composition. This is reputed to be an original frame and it may be a fragment of one since the unusual crescent motif at the corners is also found in the portrait of Mademoiselle Rivière of 1806 which suggests the involvement by Ingres or at least his framemaker in the framing of both portraits.

The frame (obviously re-finished) around the portrait of Bonaparte as First Consul (no. 2, on deposit Musée d'Armes, Liège) is a standard type of French "Empire" frame generally associated with the second decade of the century, and so probably later than the picture, although the densely designed acanthus, oak leaf and scrolls applied at the centre and corners in the hollow do relate to the classical motifs woven into the carpet and printed in the fabric of the wall in the room in which Bonaparte stands.

Many of the early portrait drawings were made by Ingres in Rome. It is known that his sitters were invited to have their drawings (mounted on boards of standard size) framed and glazed by a neighbouring Roman craftsman. But none of the drawings in the exhibition retains its original frame although some are in neo-classical frames of the right period. Those kept in museums have broad, heavy mounts of a type unfamiliar to Ingres which have been devised for storage piled in solander boxes, for handling by visiting scholars and for framing in fairly large standard sized frames. One of the frames here, around Mademoiselle Jeanne Hayard (no. 51, private collection), is a collector's frame of c. 1900, reviving the neo-classical style. Unusually, it is made of dark, patinated bronze with a sight edge and miniature applied garlands which are gilded.

Several of the later drawings of the 1820s and 1830s do appear to retain their original frames. The drawing of Madame Godinot (no. 106, private collection), which is unusual in that it remains on its original mount, has a frame of ogee section with Greek palmette ornament which seems large in scale (the effect is very different to that of the painted Rivière portrait), but it may have matched the ornament elsewhere in the room in which it was originally hung, both on furniture as well as other frames. Many other frames of this type survive from the 1820s so if this frame were selected by Ingres in 1827 it would not have been an unusual style. To many of us used to seeing drawings in large mounts or with borders this close framing may seem tight but this was the normal practice and one of the drawings, The Kaunitz Sisters of 1818 (no. 77, Metropolitan Museum of Art, New York), includes a note by Ingres indicating a portion of the paper to be concealed by the frame.

The frames round the portraits of the architect Victor Baltard and his wife (nos 114 and 115, private collections) are also close but they are far simpler. The burnished pattern of naturalistic plant forms on the matt gilding of the convex frieze is typical of the 1840s and possible for 1837, the date of these portraits. The fact that these frames are identical in style, although the drawings have been separated for over half a century, suggests that they must have been both framed like that soon after they were made, presumably when the couple returned to Paris in 1838 or 1839. It is interesting that in 1840 Ingres recommended his friend Gatteaux to consult with Baltard ("excellent homme de talent et de goût") concerning the ornament for the frame of his Antiochus and Stratonice (Boyer d'Agen, Ingres d'après une correspondance inédite, Paris 1909, p. 294). Comparison with the Godinot portrait suggests that the style of frame for these portrait drawings changed much more radically than Ingres's own style as a draftsman. Here too, however, there are many frames which would have surprised the artist, including the very lean, silvered frame for the refined drawing of the slim Lizst (no. 116, Richard Wagner Stiftung, Bayreuth), a modern frame chosen without reference to historical precedent but with careful attention to the work of art.

Resuming our review of the paintings we meet a number of frames which appear at first sight to be plausible originals. The portrait of Madame de Senonnes (no. 35, Musée de Nantes) of 1814 is in a genuine French neo-classical frame not dissimilar in character to that on the earlier portrait of Madame Rivière but on close examination the composition ornament is assymetrically disposed and the corners look altered, so this frame must have been cut down for the painting. The plain, rather fat, ogee moulding around the expansive Joseph-Antoine Moltedo of c. 1810 (no. 27, Metropolitan Museum of Art, New York) is a characteristic Roman frame, plausible, although of a pattern established in the eighteenth century, but it has been found for the picture relatively recently. The plain hollow moulding around the portrait of the artist François-Marius Granet (no. 25, Musée Granet, Aix-en-Provence) seems to be the preferred frame of the sitter and is found round many of his paintings in the same collection. It may have been the portrait's first frame but was not made for it since its fixings show that it was originally hung from one of the long sides (around a landscape?)

Perhaps the most striking and interesting frames in the exhibition are those around the portraits of Monsieur and Madame Leblanc (nos 88 and 89, Metropolitan Museum of Art, New York). These are obviously old, they were not fitted by the Museum, or by their previous owner, Edgar Degas, and so it has reasonably been assumed that they are the original frames made in Florence in 1823 when the portraits were finished. However, although the lotus leaves and daisies on the broad, gentle double curve of the main section have a neo-classical look, they are not part of the ornamental vocabulary favoured in Florence in the 1820s, and are combined with a narrow, sanded flat and a band of strap-work near the sight edge which is typical of French frames of the late seventeenth century which were revived in France and England in the 1830s. This suggests that they were made later and in France (as does the use of pine, visible behind) and so the portraits were probably reframed in Paris where the artist and the sitters moved, perhaps when one of the portraits (Madame Leblanc) was shown at the Salon in 1834 or slightly later. Ingres may well have been consulted, indeed he may well have proposed the change, and similar ornament, as Paul Mitchell and Lynn Roberts have noted, was employed for at least one other frame around his paintings. The pale gilding seems perfect for the colours of these portraits.

The Leblanc portraits are surely designed to hang opposite each other. This is confirmed by the lighting which comes from different directions. (Ingres cared about the direction of the lighting even in his portrait drawings, as can be seen in the note in his hand pasted to the back of the portrait of Dr Thomas Church (no. 68, Los Angeles County Museum of Art) in which he reminded the owner to hang his drawing so that the light fell from the beholder's left-hand side.) The Leblancs would have gazed at each other and at their guests from across a room - a style of arrangement hard to replicate in a book or lecture theatre. This made it important that each portrait dominated the wall upon which it hung, and that would have determined the broad frames they have been given - an example of how picture frames belong also to the decor of a room, and relate closely to the way pictures are intended to be seen.

In 1833, the year before Madame Leblanc was shown at the Salon, Ingres exhibited his consummate portrait of Louis-François Bertin (no. 99, Musée du Louvre). Its highly distinctive frame, carved, with vine leaves, birds and lizards, has not been sent to the current exhibition but is reproduced in the catalogue (fig. 180) and appears to be that visible in photographs of the Ingres installation at the 1855 Paris Universal Exhibition, albeit then with a cresting or shield at top centre (fig. 301). The frame is compared in the catalogue to that on Raphael's Baldassare Castiglione (fig. 181, Musée du Louvre) with the suggestion that by designing such a frame Ingres may deliberately have been declaring his pride in having created an updated, modernised version of the Renaissance masterpiece. Several important considerations do not support this theory. Firstly, there are many more Raphaelesque paintings by Ingres. Secondly, the similarity could easily have been made closer and seems not to have been noted at the time. Thirdly, very similar ornament was used in the frame for Ingres's Joan of Arc which would have diminished the point. The naturalism of this ornament is part of the same taste that determined the burnished leaves on the Baltard frames later in the same decade.

It is clear from Ingres's portraits that he was keenly attentive to changes in fashion - the latest styles of furniture, wall covering, porcelains can be found in his portraits as well as novelties in jewellery, dresses and shawls. Nevertheless, for art historians he has tended to remain quintessentially neo-classical. The portrait of Cherubini (no. 119, Cincinnati Art Museum) painted at the end of the artist's stay in Rome of 1835-1841 can be seen in an early photograph reproduced in the catalogue (fig. 227) in an unmistakeably mid-nineteenth century frame, perhaps the original one. Today it has a neo-classical frame with a shallow fluted hollow probably dating from the 1780s and bearing the stamp of the important Parisian maker, Chérin. The same process may be observed in other cases. Thus the portrait of the Comte de Pastoret of 1826 (no. 98, Art Institute, Chicago) is in an especially fine (and unaltered) early neo-classical Parisian frame of the 1780s. With Ingres's portraits of the 1840s and 1850s the disparity is even more marked. The Vicomtesse d'Haussonville of 1845 (no. 125, The Frick Collection) had a frame of early eighteenth century style - a frame to match the objects in the picture - which may be glimpsed in two illustrations in the catalogue (one of 1846 and the other of 1855 - figs 299 and 302). It was switched to the copy of the portrait which was retained by the family. The fine carved frame adapted for the painting is not, however, historically anomalous, although distinctly more classical in style. The portrait of Baronne James de Rothschild of 1848 (no. 132, private collection) can be seen in a photograph of the hall at the Château de Ferrières taken in about 1880 (fig. 259) in a frame as deep and rich as its present one is narrow (Louis XIII style). In the exhibition it is possible to view the Rothschild portrait together with the National Gallery's own Madame Moitessier completed in 1856 (no. 134), which certainly does retain its original frame.

The Moitessier frame was, however, also removed from the painting at the instigation of Kenneth Clark who candidly admits in his autobiography that he feared that it would have made the portrait unpalatable to the Trustees who approved its purchase in 1936. The frame was then used for a Turner and only restored to the painting in 1970. The section is that of a mid-eighteenth century French frame just as the sitter's skirt is a revival of the broad skirts of that period. And the flowers and buds and leaves which smother the frame have evidently been suggested by the pattern of Madam Moitessier's dress. It is assumed that the frame was chosen by Ingres and indeed designed by him but since the ornament is cast in composition the boxwood moulds for each element were perhaps not specially cut for him. Flowers do in fact feature prominently on other frames of this date, including some around portraits by Winterhalter (who was in even greater demand than Ingres) but rarely in such profusion.

From his letters it is clear that at least by 1822 Ingres had highly original ideas about framing. In that year he wrote to his friend Gilibert recommending a plain black wood moulding of Flemish seventeenth-century character for his Vow of Louis XIII (Boyer d'Agen, op. cit. p. 90). Whatever his attachment to a classical tradition it did not preclude a promiscuous eclecticism in frame style. For the Antiochus and Stratonice he proposed "le cadre le plus large, le plus riche et le plus grec possible" (Ibid. p. 292) and for the Odalisque à l'Esclave not many months later "un assez beau cadre, bien large et aussi baroque que possible" with exotic ornaments, perhaps those "que l'on appelle, je ne sais pourquoi, 'gothiques modernes'" (Ibid. pp. 298-9). In the exhibition only the Moitessier frame survives as evidence of this originality.

These notes were compiled after the seminar organised by Kathy Adler, Head of the National Gallery Education Department, and Christopher Riopelle, Curator of Nineteenth Century Paintings and co-curator of the exhibition, shortly before the opening of the exhibition at the National Gallery. It incorporates much information supplied by Jacob Simon. The references to Boyer d'Agen are owed to Lynn Roberts. John England was able to examine the backs of many of the frames and supplied information on structure and materials. Corrections and additional information will, it is hoped, be supplied by those who attended the seminar as well as by others. The exhibition was shown in London, Washington and New York in 1999

Nicholas Penny, National Gallery, London