BY RICHARD GREEN

ONE OF THE best-known and most attractive works in the extensive collection of drawings by Paul Sandby in the Royal Library at Windsor Castle, apart from topographical views of the Castle and Windsor Great Park, is the watercolour hitherto titled A lady painting (Fig.2).1 Exactly how and when this entered the library is not certain, but it is first recorded in a group of Sandby drawings lent from Windsor to the Royal Academy’s great exhibition of British art in 1934.2 Since then it has regularly featured in the Sandby literature and in exhibitions, most recently being included in the monographic show held in 2009–10.3 While the watercolour has generally been dated to the 1760s on the basis of comparison with other works by the artist and of costume, the sitter has remained unidentified. Closely related is a drawing by Sandby in red and black chalk of exactly the same size at the Yale Center for British Art, New Haven (Fig.1).4 This depicts the same young woman in a similar setting, although from a viewpoint further to the left, working at the same table but, rather than painting, using a porte-crayon to copy a print or drawing of a head mounted in front of her, perhaps on a vertical flap projecting from her angled worktop. This drawing has also been dated to the 1760s, but again no attempt has been made to identify the sitter.5 And if, as seems likely, both the Windsor watercolour and the New Haven drawing remained in the artist’s possession until his death, we cannot hope for early provenance clues to guide us to her identity.6 Although there are critical, even conflicting, differences between these two works – which will be addressed below – they will initially be scrutinised in tandem.

An important but little documented aspect of Sandby’s career was his activity as a drawing master. Probably instructing the sons of George III,7 and certainly members of the Greville, Harcourt, Williams Wynn and other noble or landed families, he clearly operated at the top end of this particular market. In the two closely related works under discussion there can be little doubt that Sandby has recorded one of his pupils – the young daughter of a wealthy family. The scale of the window behind her suggests a room of generous proportions. She sits at an ingenious painting table with sliding trays, containing shells of colours, which can be pulled out for work in watercolours or left partly closed for drawing; a palette for mixing colours attached to one of the front legs may be swivelled in or out according to the activity. Probably, with the trays fully closed and the adjustable top laid flat, this piece of furniture functioned as a card table or something similar. All of this confirms that the young pupil is on home territory, rather than in Sandby’s studio. More clearly in the watercolour than in the chalk drawing, we see that the lower sash of the window behind her (which would have been subdivided by glazing bars holding small panes) has been fully raised. While, incidentally, indicating that the time of year is summer, as does the lightweight character of the sitter’s clothing and her bergère hat, the open window offers a valuable clue to the location of the room. The view we glimpse through it – in the watercolour – was first identified by the present writer in 2009.8 In 1947 Paul Oppé described this as being ‘over a river to houses’, while more recent commentators have seen the latter as ‘warehouse-type buildings’ or ‘a castle’.9 In fact the Palace, with Archbishop Morton’s gatehouse seen sideways on, the lantern on the roof of the great hall to the left of it and the turreted tower of St Mary’s (the former parish church of Lambeth, now the Garden Museum) to the right. Although these medieval buildings are partly obscured by trees today, their distinctive configuration remains unchanged (Figs.3 and 4).

The most likely candidate for consideration as a sizeable house on the river bank almost opposite Lambeth Palace where Paul Sandby might have been employed in the 1760s to instruct the daughter or daughters of a wealthy family is the old Grosvenor House.10 Formerly known as Peterborough House, it appears as Belgrave House on Richard Horwood’s London map of 1792–99 at the very end of Millbank Row, a continuation of Millbank Street, and marked the then southern extremity of the city. Of seventeenth-century origin, but extensively remodelled in the early 1730s, it had passed from Alexander Davies into the Grosvenor family through the marriage of his daughter Mary in 1677, at the age of twelve, to Sir Thomas Grosvenor, 3rd Baronet. A substantial red-brick house of two main storeys, it was set back from the river by a garden and was skirted by a public pathway. It stood until 1809 when it was demolished, having been vacated by the Grosvenors, presumably in the light of proposals for the building of what was eventually to be the Millbank Penitentiary on land to the south. A painting by Daniel Turner, probably dating from just before its demolition, strikingly illustrates the relationship of Grosvenor House, seen obliquely on its extreme left, to Lambeth Palace in the distance, on the right (Fig.5). A less oblique view of the house, in watercolour, by George Shepherd is dated 1809 (Fig.6), while fully frontal views are provided by an unattributed watercolour perhaps of c.1810 (London Metropolitan Archives) and an etching by Thomas Dale after Thomas Hosmer Shepherd, dated 1821. All three of these corroborate the evidence of Turner’s painting, presenting the house’s river façade in more detail, including its balcony at first-floor level with iron railings. Curiously, however, in the Shepherd watercolour alone the firstfloor windows do not reach down to the level of the balcony to allow easy access to it, although they do fall below the top of the railings. There seems to be a correspondence in this with the window in the New Haven drawing, but not that in the Windsor watercolour, although in the latter Sandby has taken pains to record the red brickwork to the left of the opening.

The old Grosvenor House was the principal London residence of the Grosvenor family from 1719 to 1755. Thereafter, however, it was leased to John Delaval (1728–1808) of the famous Northumberland land-owning family which had seats in that county at Ford and Seaton Delaval, as well as at Doddington in Lincolnshire, and had acquired new wealth from coal-mining, salt-production and bottle-making in Seaton Delaval and nearby Hartley.11While taking the lead in developing these industries, Delaval was elected Member of Parliament for Berwick from 1754 to 1761, from 1765 to 1768 and again from 1780 to 1786, and Grosvenor House would have been extremely convenient for the Palace of Westminster.12 He adopted the middle name Hussey in 1759 on succeeding to the Doddington estate, was created a baronet in 1761 and elevated to the peerage as Baron Delaval in 1783.13 Robert Wilkinson recorded in or around 1819 that after 1755 Grosvenor House was ‘inhabited by Lord Delaval and Mr. Symmons’.14 Joshua Reynolds, in his pocketbooks, noted Sir John Hussey Delaval’s address in the week beginning 2nd April 1764 as Grosvenor House, Millbank;15 while the Delaval papers held at the Northumberland Archives include a visitors’ book for Grosvenor House kept by Lady Hussey Delaval in 1769 and letters written by her daughters Sophia Ann and Elizabeth from that address in 1770 and 1771.16

If we accept that the setting of Sandby’s watercolour and chalk drawing is a room in the old Grosvenor House during the years it was leased to John Delaval, it follows that the sitter who appears in both is almost certainly one of his five daughters, all by his first wife, the widowed Susanna Potter, née Robinson, whom he had married in 1750 – namely, Rhoda (1751–70), Sophia Ann (1755–93), Elizabeth (1757–85), Frances (1759–1839) and Sarah Hussey (1763–1800). In determining which, it is necessary to consider the evidence of the sitter’s costume, datable to around 1760. A vital clue is provided in the watercolour by the leading strings falling from the top of the sitter’s dress down her back, as recently noted by Aileen Ribeiro.17 Leading strings were narrow strips of fabric with the practical purpose of supporting or restraining a child learning to walk but ‘were retained in female dress until about the mid-teens’ as a decorative conceit. Of the five Delaval daughters the only one who can be reasonably matched both with these determinants and with the apparent age of the subject is Rhoda – if we assume she was portrayed by Sandby in around 1763 or 1764, when she would have been twelve or thirteen years old. Although the leading strings do not appear in the New Haven drawing, Ribeiro comments that ‘the back of the dress has a reinforced bodice, possibly to help with either a slight deformity or [. . .] deportment’.18 There is no evidence that Rhoda was deformed, but an aid to deportment would not seem inapprop riate for a young woman of twelve or thirteen at that time.

Rhoda, John Delaval’s first daughter, was born on 18th Feb - ruary 1751 at Seaton Delaval and died, aged only nineteen and probably of consumption, on 7th August 1770 at Doddington, where she was buried in the churchyard. Very little evidence of her brief life survives, apart from a handful of letters among the Delaval papers in the Northumberland Archives. This includes dutiful missives in an elegant hand addressed to her parents. One of them written in French (at the age of eight) to her mother and correspondence in Italian with her tutor at the British Museum demonstrate proficiency in foreign languages.19 Rhoda was the third member of the family to bear this first name. Her paternal grandmother was Rhoda Apreece, who had brought Doddington Hall into the family through her marriage to Captain Francis Blake Delaval. Their eldest daughter (John’s sister) was the second Rhoda, who married Edward Astley and was an amateur artist of considerable standing, having taken lessons from Arthur Pond, starting in 1744.20 Her artistic practice will undoubtedly have set an example for her nieces, including the third Rhoda. Moreover, it is perhaps not too far-fetched to speculate that the painting table seen in the two Sandbys under discussion passed from her to these nieces on her untimely death probably following childbirth in 1757. While it was common practice for members of wealthy families in the eighteenth century to be instructed in the polite art of drawing by teachers such as Sandby, Rhoda Astley exemplifies a particularly strong tradition of artistic activity in the Delaval family. Her younger brother George was actually apprenticed to train under Pond, for £300, in 1754,21 while her younger sister Anne Hussey Delaval, later Lady Stanhope, was portrayed full-length by Reynolds in 1763–64 (Baltimore Museum of Art), holding a porte-crayon and surrounded by other accoutrements of an artist.22

In order to conclude the argument we need to return to the discrepancies between the Windsor watercolour and the New Haven drawing, which have thus far become apparent, hoping to find a way of reconciling them. The major difference concerns the background. In the chalk drawing, the window opening falls considerably lower than in the watercolour – not fully to the floor, but nevertheless indicating that it probably gave access to a balcony, thus locating the scene on the first floor (or at least an upper floor) of the house and tying in with George Shepherd’s view. By contrast, the window in Sandby’s watercolour could not possibly give access to a balcony, its lower edge being more or less at waist height. Perhaps the scene has been moved to a downstairs room. A more sustainable hypothesis, however, taking into account other differences, is that in the watercolour Sandby has created an enhanced and more considered version of the chalk drawing, involving a degree of invention. The painting table has been refined, for example, by making the legs more slender, and the implied stool of the sitter, concealed by her skirt, converted to a chair. Her costume has been elaborated not only by the leading strings referred to above but also by a generous frill to the apron, thus promoting the latter from a protective garment to something both practical and decorative. The room has been furnished with a rug and a somewhat notional checked curtain (suspiciously like those, and chair covers of the same fabric, which appear in Sandby’s drawings of subjects set in the artist’s own house).23 Not least, the watercolour has been finished by including the view across the river to Lambeth Palace – to allow for which Sandby needed to eliminate any balcony railings which might have impeded the prospect – and the normalised window provides an effective frame for this. The glimpse of Lambeth Palace is slightly less oblique than a viewpoint at Grosvenor House might have afforded, but this view could well have been taken, independently, from a position more directly opposite – for the sake of clarity – or indeed from an existing print or drawing. The corollary of this hypothesis is that the New Haven drawing was Sandby’s initial, ad vivum response to the subject, whereas the Windsor watercolour was an elaborated version intended for sale or presentation to his presumed patron, the sitter’s father.

For several reasons, the argument offered here cannot be absolutely conclusive. There is no reference to any of the Delavals in the literature on the artist, while there seems to be nothing in the Delaval papers in the Northumberland Archives documenting payments to Paul Sandby. Moreover, despite the undeniable resemblance of Sandby’s subject to members of the Delaval family, there is no certain portrait of Rhoda Delaval with which the two works under discussion might be compared.24 Because of her early death in 1770 she narrowly missed featuring in the series of full-lengths that Sir John commissioned of himself and his family from the Newcastle upon Tyne artist William Bell from 1770 onwards, remaining at Seaton Delaval.25 Nevertheless, it is hoped that the facts of the matter, as presented here, will be persuasive. The name of Delaval would be a significant addition to the existing, very short list of known patrons of Paul Sandby as a drawing master.

I am indebted to James Birch for information about pictures at Doddington Hall, to Hugh Dixon for insightful comments based on his extensive knowledge of the Delavals, and to Aileen Ribeiro, whose advice on costume has been crucial to my argument. Grateful thanks are offered to them, and for help received from Michelle Goodman, Julie Hawthorn, Paul Lewis, Kate Phillips and Jane Roberts, and to the staff of the Northumberland Archives.

1 Also called A young lady painting; RL14377; see A.P. Oppé: The Drawings of Paul and Thomas Sandby in the Collection of His Majesty the King at Windsor Castle, Oxford and London 1947, p.65, no.259.

2 [W.G. Constable and C. Johnson, eds.]: Commemorative Catalogue of the Exhibition of British Art, Royal Academy of Arts, London, January–March 1934, London 1935, p.150, no.601, as A lady painting a miniature.

3 See J. Bonehill and S. Daniels, eds.: exh. cat. Paul Sandby: Picturing Britain, Nottingham (Castle Museum and Art Gallery) and elsewhere 2009–10, p.172, no.66, as A lady drawing.

 4 B1975.4.1881. The meaning of the letter ‘F’ inscribed on this drawing (lower right) and on others by Paul Sandby is not known.

5 Sandby’s chalk drawing has been included in several exhibitions at the Yale Center for British Art since 1977. See for example [B. Robertson]: exh. cat. The Art of Paul Sandby, New Haven (Yale Center for British Art) 1985, p.53, no.69.

6 For the provenance of the Sandby drawings in the Royal Library, see Oppé, op. cit. (note 1), pp.1–4; and J. Roberts: exh. cat. Views of Windsor: Watercolours by Thomas and Paul Sandby from the Collection of Her Majesty Queen Elizabeth II, Amsterdam (Rijksmuseum) and elsewhere 1995–97, pp.136–38. Only a minute proportion of the Sandby drawings at Windsor can be connected with royal commissions from the artist and none was commissioned by George III, so there is no reason to believe that the work under discussion depicts a member of the royal family. Most of the Windsor figure studies came from the posthumous Paul Sandby sales – either direct, as in the case of the Prince Regent’s purchases from that of 1811, or indirectly. The New Haven drawing was sold from the collection of drawings by Paul and Thomas Sandby ‘formed by William Sandby’ (great-grandson of Paul’s brother Thomas), and left by him to his cousin G.T.A. Peake, father of Hubert Peake, the vendor, Christie’s, London (in the first of two sales), 24th March 1959 (lot 113), bought by Agnew’s, London, from whom purchased by Paul Mellon in 1960. Many of the drawings had passed to William Sandby by family descent, but others came from various sources.

7 Anon.: ‘Paul Sandby’, Arnold’s Magazine of the Fine Arts 1 (1833), p.434; see also K. Sloan: exh. cat. ‘A Noble Art’: Amateur Artists and Drawing Masters c.1600–1800, London (British Museum) 2000, p.140.

8 See R. Green: review of Paul Sandby exhibition, THE BURLINGTON MAGAZINE 151 (2009), p.790.

9 Respectively, Oppé, op. cit. (note 1), p.65; Sloan, op. cit. (note 7), p.232; and G. Waterfield: ‘The scenic route’, RA Magazine 106 (Spring 2010), p.51.

10 For old Grosvenor House, see F.H.W. Sheppard, ed.: Survey of London, XXXIX (The Grosvenor Estate in Mayfair, Part 1 [General History]), London 1977, pp.3, 4, 7, 21, 36 (note) and 43; and XL (The Grosvenor Estate in Mayfair, Part 2 [The Buildings]), London 1980, pp.240 and 242.

11 For the Delavals, see F. Askham (pseudonym of J.E.C. Greenwood): The Gay Delavals, London 1955; and M. Green: The Delavals: A Family History, 2nd ed., Newcastle upon Tyne 2010.

12 John’s younger brother Edward Delaval (1729–1814), a natural philosopher, had a very oblique view of Lambeth Palace, looking upstream, from the riverside terrace of his ‘neat Gothic house in Parliament Place’, immediately adjacent to the Houses of Parliament. This is recorded in an unattributed painting of c.1810 at Doddington Hall.

13 In the Irish peerage; a barony in the peerage of Great Britain followed in 1786.

14 R.Wilkinson: Londina Illustrata, I, London 1819, p.171. If Delaval and Symmons occupied the house concurrently, then the latter was presumably John Symmons of Lanstinan, Pembrokeshire, Member of Parliament for Cardiganshire from 1746 to 1761. In any case, it would seem to be his elder son, also John Symmons, who wrote to his mother from Grosvenor House on 14th November 1785; see F. Jones: ‘Some Slebech Notes’, National Library of Wales Journal 7, part 3 (Summer 1952), p.203.

15 D. Mannings and M. Postle: Sir Joshua Reynolds: A Complete Catalogue of his Paintings, New Haven and London 2000, I, p.165, under no.503.

16 Ashington, Northumberland Archives, 2/DE/43/13 (the visitors’ book, unlocated at the time of writing); 2/DE/39/3/2–7 and 13; and 2/DE/39/6/1–4. The 1st Earl of Grosvenor bought back the lease on Grosvenor House in 1789, by which time the then Lord Delaval’s London home was in Portland Place.

17 Private communication with the author, 30th August 2011.

18 Ditto, 1st September 2011.

19 Ashington, Northumberland Archives, 2/DE/39/3/1; and NRO 429/18/20–25.

20 L. Lippincott: ‘Arthur Pond’s Journal of Receipts and Expenses, 1734–1750’, Walpole Society 54 (1988), pp.260 ff.

21 Idem: Selling Art in Georgian London: The Rise of Arthur Pond, New Haven and London 1983, pp.42 and 96.

22 Mannings and Postle, op. cit. (note 15), I, pp.430–31, no.1689; and II, pl.759.

23 Oppé, op. cit. (note 1), pls.123, 124 and 129; and E.H. Ramsden: ‘The Sandby Brothers in London’, THE BURLINGTON MAGAZINE 89 (1947), p.19, pl.A.

24 The closest contender is an unattributed half-length portrait of a girl at Doddington Hall. Although she holds a rose, or roses, which may allude to the name Rhoda, there is no certainty about the identity of the sitter.

25 William Bell, who probably taught Sir John’s children drawing when they were in Northumberland, exhibited two views of Seaton Delaval at the Royal Academy in 1775, giving his address as ‘At Sir John Delaval’s’, presumably Grosvenor House.