Sir Peter Lely 1618–1680
Portrait of a Woman
Oil on canvas
1264 x 1015 mm
Inscribed ‘Dutchess of Cleveland’ in a later hand on stone ledge, centre left
…; in the collection of William, 6th Baron Craven (1738–91) at Coombe Abbey, Warwickshire by June 1769 (as ‘Dutchess of Cleveland … by Sir Peter Lely’); by descent to Cornelia, Countess of Craven, by whom bequeathed to Tate Gallery 1965.
The Stuart Portrait: Status and Legacy, Southampton City Art Gallery, Southampton 2001–2, no.15.
Catalogue of the Pictures at Coombe Abbey Warwickshire, , p.5 (as ‘10. The Duchess of Cleveland P.Lely’); C.H. Collins Baker, Lely and the Stuart Portrait Painters, vol.1, 1912, p.173, reproduced, and vol.2, p.128, no.148 (as ‘Duchess of Portsmouth’); R.B. Beckett, Lely, London 1951, p.58, no.421 (as ‘Duchess of Portsmouth’).
With the restoration of the Stuart monarchy in 1660, Lely became principal portrait painter at the court of Charles II. Highly prolific, he often used other artists to carry out the less individualised areas of his pictures, such as the drapery.
Lely’s debt to the earlier court artist Sir Anthony van Dyck (1599–1641) can be see here in the seated three-quarter-length composition with a richly patterned curtain that billows down behind the sitter to the right, and a stone window opening at the left with an empty terracotta vase on its ledge.1 The sitter rests her left hand on a golden metal container. This may be meant to echo the jar found in earlier religious paintings as a symbol of the reformed New Testament sinner Mary Magdalene. Through the window can be seen a landscape, with a seemingly ruined building and a circular stone tower on a wooded hill, and a stretch of water below.
The sitter cannot be identified with certainty. Although a deep decolletage is seen in many portraits of Restoration court ladies by Lely, it is unusual for him to depict a woman with a breast completely exposed, as here, nipple and all. Lely’s best-known portrait of this type, now in New Haven, dates from the late 1660s, and is thought to be of Diana Kirke (died 1719), who was the mistress of Aubrey de Vere, 20th (and last) Earl of Oxford, before becoming his second wife in 1673.2 It has generally been presumed that this bare-breasted manner of portrayal denotes a mistress, rather than a lady of impeccable public virtue. This presumption still seems plausible, even though the identity of the New Haven portrait is now in doubt. Another portrait by Lely of a bare-breasted woman is known only from the surviving engraving, identified as ‘The Lady Williams’.3
The present painting was conserved at Tate, and an apparently mid-eighteenth-century inscription ‘Dutchess of Cleveland’ was thereby revealed. This enabled the work to be identified in an inventory of the contents of Coombe Abbey taken in June 1769 following the death of the 5th Baron Craven.4 Barbara Villiers, Countess of Castlemaine (1640–1709), was King Charles II’s mistress from around 1660 and was created Duchess of Cleveland in August 1670. But while she was painted by Lely many times and although Lely gave his sitters similar facial characteristics, the present subject cannot, on grounds of age, be identified as her.5
In 1912, the sitter was re-identified, on unknown grounds, as the mistress who supplanted Cleveland in the king’s affections, the French-born, dark-haired Louise de Keroualle (1649–1734).6 Again, however, the resemblance to authentic portraits of Louise is not strong enough to convince.
When the work was acquired from the Craven collection by Tate in 1965, the sitter was yet again re-identified, this time as Margaret Hughes (died 1719), the mistress of Charles II’s cousin, Prince Rupert (1619–82). Margaret was one of the earliest female stage actors, and the first woman to play the role of Desdemona in Shakespeare’s play Othello. At his death in 1682, Rupert had left his goods and chattels to William, 1st Baron and Earl of Craven (1608–1697) in trust jointly for Margaret and his daughter by her, Ruperta. Craven did indeed support the two women financially. Many works of art from Rupert’s own collection and from that of his late mother, Elizabeth of Bohemia (1596–1662) were subsequently recorded with Craven’s descendants at Coombe Abbey. It seems to have been the link with Prince Rupert that prompted Tate to re-identify the sitter – who stretches out her hand to quieten an adoring spaniel – as his mistress. Spaniels were closely identified with the Stuarts, as both Charles I and Charles II were known to be devoted owners of these dogs.7 However, comparison with a documented portrait of Margaret Hughes (private collection) is not conclusive.8
It is not known when the Duchess of Cleveland’s name was painted on the canvas, but similar inscriptions, some erroneous, appear on other works from the Craven collection. They may have been added for the 5th Baron Craven (1705–1769) who, according to a contemporary, ‘had been routing out a thousand curious ancestors that have been nailed up in boxes in a long gallery ever since the Queen of Bohemia’s time ... and are now ranged about the house’.9
While most of the painting is in excellent condition, the bottom left-hand corner had previously suffered damage and overpainting, and the area of the sitter’s hair was rather abraded. At some time, the lady’s blue mantle seemed to have lost a thin top glaze of more intense blue – perhaps made of the costly pigment ultramarine – so that what is now seen is probably the under-layer.
Whoever the subject of this very beautiful painting actually is, the multiple changes of name successively bestowed on her exemplify the difficulties that can arise with early portraits. Where no original lettering, heraldry or other uniquely identifying element has been painted in by the artist, and where the image has not been independently preserved through having been engraved by name, it can be extremely hard to identify a sitter with certainty.