Pierre Mignard 1612–1695
NG 2967 
The Marquise de Seignelay and Two of her Sons

Humphrey Wine, 2001

Extracted from:
Humphrey Wine, The Seventeenth Century French Paintings (London: National Gallery Company and Yale University Press, 2001).

Oil on canvas, 194.5 × 154.4 cm

Signed and dated bottom right: P.MIGNard/PINXIT. 1691

Provenance

Collection of Quintin Craufurd (1743–1819), his posthumous sale, Paris, 20–21 November 1820, Delaroche & Paillet (lot 162, 299.95 francs(?) to Dalbatre);¹ collection of Sir Richard Wallace (1818–90) of Paris and London by 1874, probably after 1870;² by inheritance to Lady Amélie Julie Charlotte Wallace (née Castelnau) (1819–97) in 1890 when recorded at no. 2 rue Laffitte, Paris;³ inherited from her by Sir John Edward Arthur Murray Scott (1843–1912), who bequeathed NG 2967 (among other paintings all then in Paris) to the National Gallery,⁴ which was advised of the bequest in 1914 and accepted it that year, agreeing to pay French Mutation Duty of 7780 francs.⁵

Exhibitions

Paris 1874, Palais de la Présidence du Corps Législatif, Exposition au profit de la colonisation de l’Algérie par les Alsaciens‐Lorrains (931) (as ‘Portrait en pied d’une princesse et de ses enfants’); Paris 1888, Exposition de l’Art Français sous Louis XIV et sous Louis XV au profit de l’Oeuvre de l’Hospitalité de nuit (27); London 1956–76, Lancaster House on long‐term loan to the Government Art Collection; Montreal 1981, Largillierre and the French 17th Century Portrait (7).

Technical Notes

The condition is generally good apart from some minor wear, particularly in the older child’s hair. What appear as vertical staining or wear, mostly in the sea and the sky, are traces of an earlier composition (see below). There has been some fading in the red lake of the older child’s cloak.

The primary support consists of two pieces of plain‐weave canvas joined vertically some 41 cm from the right‐hand edge. This support was relined sometime before the Gallery’s acquisition of the painting, probably during the nineteenth century. The stretcher, which bears a circular stamp MOMPER/PARIS is also probably nineteenth century.⁶ NG 2967 was last restored in 1956.

Cross‐sections have shown that the painting has a double ground consisting of a rich mid‐brown lower layer and a mid‐pink upper layer. Double grounds are usual in French seventeenth‐century paintings, but typically consist of orange, red or brown ochreous first layers with a second priming of grey, grey‐brown or light brown over them. The particular composition of lead white mixed with crystalline mineral red [page 245][page 246]iron oxide for the lower layer, and fine red earth mixed with a large proportion of lead white for the upper layer, as found in NG 2967, may, however, be typical of paintings by Mignard.⁷ It may be the way in which the ground was applied, probably with a spatula, that has caused the vertical streaks visible on the surface of the painting, but from the infra‐red photograph (fig.1) itseems more likely that there were once three columns in the background and a small ketch to the left of them.

The intense blue of the marquise’s cloak is painted in ultramarine of the highest quality, while smalt has been used in the sea and azuritein the boy’s breeches –in both cases mixed with other pigments. The glaze on the younger child’s red drapery contains a cochineal‐based red lake pigment. The paint medium has been identified as walnut oil.⁸

Pentimenti are visible to the position of the younger child’s right foot, at the top of his quiver and to the outline of the marquise’s right shoulder. It seems from the X‐radiograph that the marquise’s hand holding the locket was originally painted slightly higher and 2–3 cm to the left, and that it was not holding any pearls. The X‐radiograph also shows that the lace of the marquise’s dress was originally lower and revealed her right breast.

Discussion

It is rare for the Gallery’s Trustees to overrule the Director over the acquisition of a painting. This, however, is what occurred in the case of NG 2967. Following his visit to Paris with J.P.H. Heseltine in 1914 to see the painting (and the others in Murray Scott’s bequest), the then Director of the National Gallery, Charles Holroyd, noted it as ‘a large cold picture dull in composition & colour but well painted; a fair example of this dull painter. I recommend that it be declined. If accepted, it could only be hung downstairs.’⁹ Holroyd’s attitude may have been influenced by the fact that, as far as he was aware, the portrait’s subjects were unknown. The description of the picture in the letter from Murray Scott’s executors’ solicitors to the Gallery, itself quoting from Murray Scott’s will, dated 26 October 1900, called it ‘Portrait of a Queen with sin (sic!) and Cupid by Mignard’.¹⁰

Murray Scott’s description is surprising, given that the sitters had been identified by Mantz in 1874 (by reference to the abbé de Monville’s biography of Mignard, first published in Paris in 1730) as Madame de Seignelay and her two sons.¹¹ The painting had also been so published by M.J.J. Guiffrey in 1875, as well as being exhibited as such in Paris in 1888 (see Exhibitions).¹² However, the question of the mythological guises in which Madame de Seignelay and her companions appeared seemed not then to have been settled. De Monville, on whom Mantz had relied, had been quite clear: ‘[Mignard] painted among others Madame de Seignelay and her two sons full length in the same picture; Madame de Seignelay, who has the younger of them by her as Cupid, is shown as Thetis with all the attributes of the sovereign queen of the seas; and her elder son is painted as Achilles: the sea forms the picture’s background.’¹³ However, on the occasion of the picture’s exhibition in Paris in 1888, the marquise and her elder son were said to be in the roles of Amphitrite and Mars. When NG 2967 first appeared in a National Gallery catalogue, its compiler – presumably Holroyd – identified the marquise as Amphitrite and the younger child as Cupid, but called the elder child ‘a boy who stands as a Roman Warrior’.¹⁴

The current mythological identification of the figures, in accordance with de Monville’s description, was first adopted by the Gallery in 1929¹⁵ and has not been questioned since, but only in 1946 were the two children in the portrait acknowledged by the Gallery as sons of the marquise.¹⁶ Since de Monville claimed that his biography of Mignard was based on the recollections of the painter’s cherished daughter, Catherine, who was then still alive, it seems reasonable to rely on it, at least so far as the identification of the principal sitter as Madame de Seignelay is concerned.¹⁷ However, in one respect de Monville’s biography is inaccurate, because the marquise had five sons, not two, and it is unclear which two are represented.¹⁸ This also raises the question of whether de Monville correctly identified the sitter’s mythological guises, but before considering this, some account of the marquise herself seems appropriate.

Catherine‐Thérèse de Matignon‐Thorigny, marquise de Lonrai (1662–99), was the younger daughter of Henri de Goyon‐Matignon, comte de Thorigny, and Marie‐Françoise Le Tellier. Henri de Goyon‐Matignon (1633–82) became governor of Cherbourg and, like other members of his family, followed a military career. In 1679 Catherine‐Thérèse married, as his second wife, Jean‐Baptiste‐Antoine Colbert (1651–90), marquis de Seignelay and eldest son of the great Colbert (1619–85).¹⁹ After learning the principles of sailing in 1670, Seignelay toured Europe the following year, visiting Provence to learn about the Levant trade, Italy to study the fine arts, and Holland and England to look at shipyards. In 1676 he was appointed Minister of the Marine, and in 1683, on the death of his father, he became Secretary of State at the Marine. As such he developed the French fleet to the height of its power under Louis XIV. He also, however, saw active service, directing operations in naval battles with the English as well as the naval bombardments of Genoa and Algiers.²⁰ As Secretary of State at the Marine, Seignelay was in constant rivalry with the marquis de Louvois (1641–91), Secretary of State for War and the most powerful of Louis XIV’s ministers after the death of Colbert. This rivalry, known as the ‘debate between the Earth (i.e. Louvois) and Seas (i.e. Seignelay)’ also found expression in Seignelay and Louvois competing to patronise Mignard, and it was presumably with Seignelay’s encouragement that in 1684 the artist painted for Louis XIV at Versailles Neptune offers the Empire of the Sea to the King (Compiègne, Musée National du Château; fig. 2), as an allegorical homage from Seignelay to his monarch.²¹

As well as being an admirer of Mignard,²² Seignelay was a collector of old masters on a lavish scale.²³ Indeed, his whole style of living was distinguished by its magnificence,²⁴ which seems to have suited Catherine‐Thérèse well. According to [page 248]Saint‐Simon she was a tall, good‐looking woman whose excessive pride was sustained by that of her husband and by ‘his opulence, his magnificence, his authority in the [King’s] council and his position’.²⁵ When Seignelay died in 1690, she was, according to Madame de Sévigné, inconsolable,²⁶ but, and this also according to Saint‐Simon, ‘she was consumed with desire for a rank and another name although she had several children’.²⁷ Jilted by the duc de Luxembourg after all the arrangements for her wedding to him had been made, Catherine‐Thérèse found another man to give her the rank she felt her due and, it seems, a better home than Luxembourg could have done.²⁸ This was Charles de Lorraine‐Armagnac, comte de Marsan and Grand Ecuyer de France, whom she married quietly in Paris in 1696.²⁹ She died three years later.

Catherine‐Thérèse had had five sons by her marriage to Seignelay: Marie‐Jean Baptiste (1683–1712), Paul‐Edouard (1686–1756), Louis‐Henri‐Charles (1687?–1705?), Charles‐Eléonore (d.1740) and Théodore‐Alexandre (1690?–95?).³⁰ The three sons who survived into adulthood all followed military careers rather than, as might have been expected, careers in the navy.³¹

Portraits in allegorical, mythological or historical guise were common in both the seventeenth and the eighteenth centuries. Mignard, for example, painted various ladies of the court as goddesses for the decoration of the Petits Appartements du Roi at Versailles.³² This fashion in portraiture presumably required a degree of learning on the part of painters, who were no doubt happy enough to display it. However, neither the choice of Amphitrite nor that of Thetis seems entirely appropriate to the marquise de Seignelay – although portraitists did not always make appropriate choices in this regard.³³ Certainly, Amphitrite, as the wife of Poseidon, was goddess of the sea and as such would clearly have alluded to the position of the marquis as Minister of the Marine. It is also true that, given the frequent identification of the Greek god Poseidon with the Roman god Neptune, Catherine‐Thérèse’s portrayal as Amphitrite would have fitted neatly with the message of Mignard’s earlier Neptune offers the Empire of the Sea to the King at Versailles. However, if Catherine‐Thérèse was Amphitrite, her son by Seignelay/Poseidon would have been Triton, who, besides living at the bottom of the sea, was always shown with the lower body of a fish.

On the other hand, if some ancient authors were followed, Thetis would have been a distinctly unhappy choice. Thetis was a granddaughter of Poseidon and so a marine divinity. She married the mortal Peleus and bore him Achilles, who would become the Greek hero of the Trojan war. However, this marriage was against her wishes (Iliad, XVIII:432–3), and Peleus, unlike the marquis de Seignelay, survived the death of his son Achilles. More disturbing, a number of ancient writers state that Thetis destroyed by fire six children she had by Peleus, and it was only his intervention that prevented her from doing the same with the seventh, Achilles. On the other hand, in support of de Monville’s identification of the figure, according to Homer, Thetis showed Achilles great tenderness and it was she who procured for him from Hephaestus, god of fire, ‘shield and helmet, and goodly greaves fitted with ankle‐pieces, and corselet’ (which, save for the greaves, appear in NG 2967).³⁴ However, more relevant perhaps than Homer was Thetis et Pelée, a play by Bernard Le Bovier de Fontenelle with music by Colasse,³⁵ published by the Académie Royale de Musique in 1689 and so presumably well known in court circles. In this there is no suggestion of Thetis’ infanticide, and, far from being the unwilling partner of the mortal Peleus, she loves him with a constancy (which he reciprocates) that persuades Jupiter to restore him to her.³⁶ It seems perfectly possible that it was this very recent characterisation of Thetis which the marquise and/or Mignard had in mind when NG 2967 was being painted. Moreover, it is to Hephaestus, whose forge was Mount Etna, that the smoking volcano in the background probably alludes. Jean‐Claude Boyer has pointed out that the volcano may also allude to a particular naval battle: either that of 10 February 1675 by the volcanic island of Stromboli, in which a French fleet of nine ships under Duquesne vanquished a Spanish fleet nearly five times its size, or the celebrated battle of 22 April 1676 at Agosta opposite Mount Etna, which saw the death of the Dutch admiral Ruyter.³⁷ It was during these years that Seignelay was signing naval dispatches under the supervision of his father, Colbert, so NG 2967 may well have been intended to recall the politics of Colbert and Seignelay as well as memories of the men themselves. The volcano with its double cone as depicted by Mignard seems not accurately to represent Stromboli, Etna or, for that matter, Vesuvius, although it is likely that eruptions have changed their shapes. However, the depiction of the volcano may have been based on a misunderstanding of Seignelay’s own fascinated account of Vesuvius written during his trip to Italy in 1671, in which he wrote of the volcano’s inner crater being pierced at its summit by two large holes vomiting thick and sulphurous smoke.³⁸ Mignard has painted the volcano with two summits belching smoke.

Another mythological identification of the figures which would be consistent with the volcano being identified as the forge of Hephaestus, and more consistent with the younger child being identified as Cupid, would be to see the marquise and the elder child as Venus and Aeneas. In legend, Venus was the mother of both Cupid and Aeneas and her standard attributes included the scallop shell, on which the marquise’s right foot is resting here, and strings of pearls, which in NG 2967 are prominently displayed by, and offered to, the marquise. The oval locket(?) held by the marquise is faced with a giant pearl (of anthropomorphic shape?), reinforcing this connection.³⁹ Furthermore, according to Virgil’s Aeneid (VIII:370–85) Venus went to the forge of Vulcan (Hephaestus) to procure arms for Aeneas. Paintings of Venus presenting arms to Aeneas usually portrayed Aeneas as a mature war‐rior,⁴⁰ but this can be ignored given the ages of the marquise’s children. However, had the pearls and scallop shell, not to mention Cupid, led the marquise’s contemporaries to identify her with Venus, this would also have led to some awkward, and presumably unintended, conclusions, because anyone following this line of mythologically based interpretation to its conclusion would have deduced that the marquise (Venus) while married to Seignelay (Vulcan) had had an adulterous [page 249] relationship with his great rival, Louvois (Mars).⁴¹ One must conclude, therefore, that de Monville’s reading of the picture is correct, but that Mignard either imperfectly realised the project, or more probably that he deliberately, and flatteringly, gave the marquise certain of Venus’ attributes while, as the X‐radiograph shows, covering the naked breast which might have suggested her identification with that goddess. The fact that she is represented with reeds and coral in her hair – neither usual attributes of Venus – as well as pearls, and that she uses Venus’ attribute of the scallop shell as a foot‐rest, may be meant to show the marquise de Seignelay’s superiority to the goddess of beauty (fig. 3).

By the time de Monville came to write his biography of Mignard,⁴² there had arisen another circumstance which, with hindsight, must have made the identification of the sitters as Thetis and Achilles seem additionally appropriate, namely the early death of the marquise’s eldest son, Marie‐Jean Baptiste. Thetis told Achilles that he would either have a long but inglorious life or one that was short but of imperishable renown (Iliad, 410–16).⁴³ Marie‐Jean Baptiste saw significant active service in the years 1702–11, rising to the rank of brigadier by the age of 25, only to die four years later. Present at the battles of Friedlingen (1702) and Malplaquet (1709) among others, wounded at the second battle of Hochstaedt, [page 250] [page 251] his life was indeed short. It must be admitted, however, that his death from apoplexy lacked glory, and his renown was not elevated by Saint‐Simon’s comment shortly afterwards that ‘although extremely fat, he excelled in dancing’.⁴⁴

If the identification of the marquise as Thetis and the child at the left as Achilles is not entirely satisfactory, it is nevertheless surely correct. And given the ages of her children, ‘Achilles’ must be her eldest son Marie‐Jean Baptiste, nine years old in 1691 when the picture was painted. There can be no doubt also that the winged child with his quiver is correctly identified as Cupid, and if he was intended to represent any of Catherine‐Thérèse’s children, then the youngest, Théodore‐Alexandre, seems the most likely. This would provide a certain symmetry between oldest and youngest, and whereas the eldest looks towards the locket(?) held by the marquise, which presumably contains a portrait of Seignelay,⁴⁵ the youngest, who may have been born after his father’s death and in any event could not have known him in any meaningful sense, looks towards his mother. It is not, however, clear why the marquise had herself portrayed with only two of her children.

The cup that Cupid offers to Thetis is similar to a shell cup with a gilded silver base in the Palazzo Pitti (fig. 4).⁴⁶ Itself an object of rarity, and possibly one of the ‘thousand other things which marked the fine powers of discernment’ of Seignelay,⁴⁷ it holds pearls symbolising the wealth of the sea, and coral, which according to Ripa had properties protective against evil.⁴⁸ At the bottom of the picture (fig.5) are various seashells, some of them imaginary, but those which can be identified include examples from New Zealand, the Caribbean and the Pacific Ocean.⁴⁹ The shells may therefore have a threefold symbolic function: they refer to the marine nature of Thetis; like the shell cup they point to the wealth and rarity of Seignelay’s collections; and they reinforce the global nature of Colbert’s and Seignelay’s marine ambitions.

One other point integral to NG 2967 is the extensive use of ultramarine of the highest quality (see Technical Notes). This may be seen as a reference to the wealth, power and progeny of the marquise herself.

The pose of the marquise may have been derived from that of the central figure in Jacopo Zucchi’s Reign of Amphitrite, one of the four known versions of which– that now in Lviv (fig. 6) – was probably at the Villa Medici during Mignard’s sojourn in Rome in the years 1635–57.⁵⁰ The cross‐legged stance of the marquise’s elder son may have been derived from an altar relief now in the Museo Chiaramonti in the Vatican, but which Mignard may have seen in the Giustiniani collection in Rome, and which had also been engraved in 1602 (fig. 7).⁵¹ As has been pointed out, in his choice of intense colours, his facial types and his placing of large figures close to the picture plane, Mignard was possibly influenced by Guido Reni, whose works he would have known both in Rome and from the series of The Labours of Hercules then in the Grands Appartements of Louis XIV at Versailles.⁵²

NG 2967 is clearly dated 1691. It is possible that it was started before Seignelay’s death on 5 November 1690, since the marquise is not shown in mourning dress, and that a [page 252] paired painting showing Seignelay with the other children was projected. But Seignelay would almost certainly have had himself portrayed with his heir, Marie‐Jean Baptiste, who, however, is shown in NG 2967 with his mother. It is therefore safer to assume that the marquise herself commissioned the portrait soon after her first husband’s death, and that once it had been decided to portray her as Thetis, mourning dress was deemed inappropriate. Regrettably, no record of the commission and of the picture’s early history is known,⁵³ though a painting attributed to François de Troy and datable c. 1712 (fig. 8) seems to be derived from NG 2967,⁵⁴ suggesting that the latter remained in the area of Paris or Versailles at least until then. If NG 2967 was once at the hôtel Colbert, rue Neuve des Petits‐Champs, it was presumably removed by 1720 when the duc d’Orléans established his stables there.⁵⁵

General References

L’abbé de Monville 1730, p. 148. (1731 edn, p. 123); Guiffrey 1874–5, p. 143; Le Brun‐Dalbanne, ‘Pierre Mignard sa famille et quelques‐uns de ses tableaux’, Mémoires de la Société Académique d’Agriculture des Sciences, Arts et Belles‐Lettres du Département de l’Aube, vol. 14, 3rd series, 1877, pp. 215–368, no. 290; Davies 1946, p. 69; Davies 1957, p. 157; Nikolenko 1983, pp. 94–5; Wright 1985b, p. 122; Nantes and Toulouse 1997–8, p. 226.

Notes