Paintings departing from country house collections to which they have historically belonged may be subject to a version of what the art historian David Carrier has dubbed ‘museum skepticism’ – the longstanding anxiety surrounding the removal of art works from their traditional contexts (churches, public buildings, palaces) to other settings which seem, from this perspective, far less appropriate and may actually be destructive. As Carrier outlines, ‘According to these museum skeptics . . . museums preserve old objects, but fail to preserve the works of art constituted by those objects’.¹ In its removal from the context of the country house, with its historic lighting and hanging schemes, rich architectural environment and decoration (and lack, generally speaking, of interpretative materials), to the probably artificially lit, relatively austere and, generally speaking, didactic setting of a museum or gallery, the object supposedly loses its authentic character and spirit. The Old Master or classic British painting becomes a fish out of water and all the museum may be able to do is feebly to recreate its original setting with approximations of wall colours and coverings, the tokenistic introduction of a console table or some ornamental objects of roughly the right kind.

Carrier offers an extended commentary on this point, and is sceptical about the sceptics. He emphasises the value as well as the distinctiveness and potency of the kind of museum spaces which emerged at the end of the eighteenth century (and which remain normative today). A complementary point would be to consider how those traditional environments which the museum sceptics have so cherished and sought to defend are themselves rather less static, stable and organic than the sceptics might insist. The recent history of Reynolds’s portrait of the 5th Earl, largely painted in 1769, delivered to Castle Howard in 1775 and installed there ever since – and, thanks to an arrangement made with HM Government in 2016 by which the painting has been accepted in lieu of tax and joins the national collection of British art at Tate while remaining, under normal circumstances, on display at the house, where it will stay – offers an intriguing case study in this regard (fig. 1). The arrangements of 2016 may be the most dramatic change in its ownership status in the picture’s history, breaking the descent of the painting through the Howard family and moving the work from private to public ownership, but nor may it be right to cast this break in the sceptical terms of a shift from the natural to the artificial, the proper to the inappropriate. The history of the painting, in terms of its genesis, and the longer history of physical display and reception which has shaped its appearance and status is more complex than that. The technical examination and conservation of the painting in 2018, prior to the picture’s redisplay for the first time in the context of the national collection of British art shown at Tate Britain, has been an opportunity to reassess this history.

The modern literature confidently dates the painting to 1769.² We know that there were a series of sittings with Reynolds in that year, between May and December, the last on 30 December. It seems highly unlikely that Reynolds put down his brush on that last date and declared the painting finished, because the painting and its frame were not paid for until September 1775, after which the framed picture was sent on to Castle Howard. At that point, some mishap occurred which meant that the painting was returned to the studio for lining and then repaired by the artist. We know of this from Reynolds’s letter to Carlisle on 2 November 1775:

I am extremely sorry for the accident which has happened to the Whole-length Picture. The circumstance of the Bar getting loose never happened to me before. The damage that is done can only be reme[died] by lining the Picture; for which it must be sent back again to Town, lining the picture is pasting it on another strong canvass, and if it were full of Holes will not be perceiv’d. The Painter being luckily still alive he can restore it to its original –

I am sorry of the disappointment which it occasions. The best way I think will be to send it rolled up. Tho it cannot be returned in that manner when it is lined³

Technical examination has revealed what seems certainly to be evidence of a tear to the canvas caused by this mysteriously detached stretcher bar, as well an early impact to the canvas which, in tandem with subsequent drying problems, has caused a large circular crack pattern in the area of Carlisle's beloved dog Rover, disrupting the image (fig. 2). A cynical reader might detect a certain archness in the tone of Reynolds’s letter. To judge from a letter sent by George Selywn to Carlisle on 17 October 1775, Carlisle (who lost fortunes through gambling at this time) had been chased by the painter for payment and had been reluctant to receive the pictures: ‘The debt to Sir Jos[hua] Reynolds does not weigh much with me; we know that it was contracted long ago, and as to appearances I think they are better served by the pictures being in your own house than in his’.⁴ The same letter indicates that the painting may have only recently been completed or, rather, that the question of payment had been put directly on that basis only relatively recently and rather to Carlisle’s surprise: ‘You ask’, wrote Selwyn, ‘if he had any orders to finish these pictures? None that I know of, but perhaps that he took that for granted’.⁵

Looking in the other direction, we can note that there were earlier sittings with the adult Lord Carlisle, dating to 1767 (Reynolds had already painted him as a boy, in 1757–59). Ellis Waterhouse in 1985 concluded that these ‘probably came to nothing’.⁶ They are now taken as resulting in the head and shoulders portrait (in a private collection; fig. 3).⁷ The turn of the head and the treatment of the features are close to the full-length portrait, although the sitter is shown in a silk fur-lined jacket, typical Grand Tour costume of the type seen most influentially in the portraiture of Pompeo Batoni. Carlisle had, in fact, discussed being painted by Batoni on his Grand Tour, which commenced with his departure to Paris in the summer of 1767 and continued through to March 1769, when he was back in England (and sitting soon after for Reynolds again). These were important years for both sitter and painter. After a bout of ugly internecine conflict within the London art world, from which Reynolds deftly stayed largely aloof, the Royal Academy was founded in December 1768 and he was anointed as its founding President, receiving his knighthood in April 1769. As Lord Carlisle set out on his Grand Tour in 1767, he was expecting his elevation to the prestigious Order of the Thistle.⁸ After somewhat protracted arrangements about the delivery of the insignia, involving at one point to Carlisle’s disappointment the recently retired Lord Clive (‘it might as well be sent by a Chelsea pensioner!’), he was invested into the order by the King of Sardinia in Turin on 27 February 1768.⁹ He had been painted by J.-B. Greuze in Paris in 1767 (now in the Buccleuch estates, Selkirk) but was disappointed with the result, which he considered excessively sober and bourgeois (‘Creuze [sic] has made me look like a Common-Councilman’).¹⁰ Already, in January 1768, Selwyn, to whom Carlisle had sent his portrait, was writing expectantly of a more obviously glamorous likeness, one which would reflect his newly elevated status: ‘If you sit to Pompeio [Pompeo Batoni] I shall hope to have a better, and with your Order’.¹¹

Carlisle and Reynolds were friends, so Reynolds was perhaps bound to make special efforts with this work. But he was with this ambitious picture also actively affirming his professional supremacy in the British art world at this point, signalled most forcefully by his contributions to the first exhibition of the Royal Academy, held in the spring of 1769 just as Carlisle sat to him.¹² The pale blond light, palatial architectural setting, steep, low perspective, the glitter of gold and deep green, evoke in a determined manner the examples of Veronese and Titian. And, above all, there is in Carlisle’s gesture, famously seen already in Reynolds’s breakthrough painting, his first full-length portrait, of Admiral Keppel in 1753 (National Maritime Museum) and famously, inevitably, associated with the gesture seen in the canonical classical sculpture of Apollo Belvedere (fig. 4). Recent research has suggested that the pose was mediated by seventeenth-century sculpture, where the hand is positioned lower and pointing (so unlike what is seen in the Apollo).¹³ The important point is that this gesture would have been associated with social power and authority.¹⁴ Such would have suited a young man of elite status, fresh from the Grand Tour, donning his new robes and seeking a rather more dynamic and determinedly aristocratic image than Greuze had been able to provide for him in 1767. The investment of painter and subject in this work is clear not only in the grand (arguably grandiose) treatment of the sitter but also in the frame, which is almost certainly original to the painting and therefore the frame which was paid for in September 1775.¹⁵ This is of a Carlo Maratta form, popular in the late eighteenth century and frequently used for Reynolds’s paintings, and in its structure and joinery has been well preserved and apparently untouched.¹⁶ Aside from the quality of the original carving, most striking is the subtle use of gilding techniques. Although partly obscured by restorations, it is notable that different surface finishes in the gilding of the different decorative elements of the frame were utilised to great effect. The frame’s front hollow was burnished to an almost mirror finish, which reflects light onto the portrait.¹⁷ This is contrasted with the matte oil gilding of the leaf and shield ornament to create interesting textural and optical differences. As an ensemble, picture and frame were designed to sit among an exceptional collection of Old Masters at Castle Howard. As an aristocratic visitor to Castle Howard noted in 1796: ‘although surrounded by the best works of the most pre-eminent masters, [it] stands the test of a situation so trying’.¹⁸

The head and shoulders of 1767 and the full length dated to 1769 have been treated as distinct works but there is a question of what closer relationship they might have enjoyed. X-radiography of the latter picture at Tate has revealed several significant changes in the painting which together suggest that the original was differently conceived (fig. 5). The setting was originally a landscape, or at least dominated by landscape features, with open sky running behind the figure where now there is a wall. Carlisle stood on the ground rather than descending steps, his feet positioned much more like those of Keppel in the earlier portrait. To the left, an ornamental urn originally topped the balustrade where now can be seen the Order of the Thistle hat, and the pilasters above the sitter’s arm had an Ionic capital. To the bottom right, where the fluffy black form of Rover now sits, there are indications of a dog’s rear legs, hinting at the form of a dog jumping up on his master’s leg. Curved pillar bases and possibly a statue were also here, in front of the arched gateway. And there is a suggestion of an acorn finial at the bottom of the balustrade. What this adds up to is a different kind of portrait, showing Lord Carlisle not as a statesman but as a gentleman in nature, pawed by a favourite pet.

Reynolds’s full-length male portraits only relatively infrequently feature dogs but they were a stock feature of the portraiture of Batoni, and the position of the original dog’s legs appears close to that seen in the portrait of the Honourable George Venables Vernon by Reynolds’s rival, Thomas Gainsborough (Southampton Art Gallery), which had been exhibited at the Society of Artists in London in 1767.¹⁹ So the painting as first conceived showed Carlisle as a country gentleman, still a figure of authority but of a different order, situated within a more open landscape setting but with indications of architecture, parkland rather than palace.²⁰ If the image which Reynolds finalised is rather formal, stately portraiture, conceived, painted and framed with the intention of being imposing, the figure of Carlisle’s beloved dog Rover is rather incongruous. The front edges of the stone steps show through Rover’s form, which has become more transparent with age. The x-radiograph confirms that he was an addition, introduced certainly after the architectural setting of the steps had been finished. Rover had died in Paris in 1768, run over by a carriage, having survived an accident in Florence earlier that year. Reynolds’s painting was, then, posthumous, painted over a resolved architectural setting and over an earlier dog.

Had the first version of the full-length even been on the cards in 1767, giving rise to the sittings and to the head-and-shoulders portrait? This would probably have been executed directly onto the canvas, with no preparatory studies.²¹ If so, might that surviving portrait be properly considered as a study towards the full-length portrait? Had this larger painting already been conceived at that earlier date, using the head-and-shoulders portrait as a reference image? The idea of showing Carlisle in his Thistle robes was already in play at the beginning of 1768, so it would be odd for a new portrait on a grand full-length format to be commenced without that aim after that date. For his part, Carlisle was well aware that once a portrait likeness was established (as had been the case with Reynolds, in 1767), the conception of a portrait could be overhauled quite drastically. Writing of that disappointing Greuze portrait to Selwyn, he noted: ‘if the face is like, there will not be much trouble in altering the dress’.²²

This remains speculation but it is clear that the painting had a more complex genesis than we have hitherto realised, even if that genesis was restricted to the period of the sittings of 1769 or, as seems more likely, it occupied a more extended period from 1767 to perhaps even 1775, with the painting undergoing several significant revisions over that time.²³ In addition, there are indications that the painting was suffering from visible intrinsic deterioration even within the 5th Earl’s lifetime. James Northcote, one of Reynolds’s pupils, stated that he had seen the portrait ‘at Jackson’s’, reporting that it was ‘in a shameful state; it was cracked and faded from the varnishes and nostrums [quack medicine] which Sir Joshua had made use of’.²⁴ When Reynolds was chasing Carlisle for payment in 1775, the Earl’s friend Selwyn quipped to him: ‘I hope that if he puts any false colours on your conduct towards him, that they will be like all his other colours, of a very short duration’.²⁵ The strict accuracy of Northcote’s comment about fading might be questioned, though, for the flesh tones of the face, which in other paintings by Reynolds have been subject to fading,²⁶ appear well preserved, as too the deep green of the robes. Perhaps rather than fading, Northcote was noticing discolouration of the oil medium, which would have been heightened by resin additives. However, the painting does display the extensive cracking which is characteristic of Reynolds’s work, occurring as the paint dried (fig. 6).²⁷

Whatever damage might have been done in 1775 and whatever the appearance of the painting in the early nineteenth century, there were certainly further changes as the century progressed. In the earlier nineteenth century the painting hung in the Orleans Room, apparently over the fireplace which is where it appears in Ellen Best’s 1832 watercolour (fig. 7). It is notable that there is cracking and degradation of the bottom strip of painting; similarly, the frame was dirty with loose dust and a heavily engrained soiling layer, particularly on the lower member, surely both as a result of this position. The paint was also found to be light-damaged, surely reflecting its many years on display at Castle Howard on walls facing the windows, in its various room settings, in the Drawing Room in the later nineteenth and early twentieth century, in the Tapestry Room at mid-century and most recently in the Turquoise Drawing Room (fig. 8).

There are no records of restoration treatments at Castle Howard but these certainly took place. The upper varnish in place until the most recent treatment of 2018 probably dates from the period when it was lined by the London picture restorer George Morrill in 1857–65, as can be established from the punch in the current stretcher (fig. 9).²⁸ This lining treatment would have involved removing the earlier lining canvas and attaching a new one using an adhesive made from animal glue paste. To bond the two canvases, the paint surface was pressed with a hot iron, which despite protective methods used by the liners frequently caused damage to the paint surface. Close examination of the portrait reveals characteristic squashed and melted impasto and moating.²⁹ Some areas of the upper varnish are tinted and pigment particles can be seen lying in age features, showing that the pigmented resin is certainly not original. There is evidence of earlier restoration too. More than a single campaign of overpaint is present, as indicated by patches and networks of strokes of varying darkness detectable under ultraviolet light.³⁰ Early overpaint and varnish residues were present before the lining and have been embedded in the paint surface due to excess heat and pressure softening the paint (fig. 10).

A sceptical reviewer for The Athenaeum in 1876 wrote disparagingly of the Reynolds paintings at Castle Howard, including Omai, which he judged ‘can hardly have been in its present condition when it left the hands of Sir Joshua in 1775’, Countess Carlisle in what ‘is now a white dress, under another which was originally almost puce’ and the 5th Earl, although the complaint then was not about condition but the choice of pose, ‘that attitude which Sir Joshua employed when he intended to be at once picturesque, stately, and graceful . . . an instance of the failure of vaulting ambition, due to early impressions of the Apollo Belvedere’.³¹ Notably, that reviewer commented on the mismatch between the colouring of the hands and face of Omai, something which is still clearly visible in the present appearance of the portrait of Lord Carlisle (fig. 11).³²

Removal to Tate Britain for display means another context, another display aesthetic. This has raised practical issues around the appearance of the frame and the painting, which would be exposed under the more emphatic and even lighting of a modern gallery, with its sheer painted walls and relatively spacious hang. The overall discolouration and unevenness of gloss, particularly in the background areas, where matte overpaint contrasted with glossy varnish presented a problem for the painting to be exhibited alongside other Tate pictures. The upper varnish layer was carefully removed from the whole surface, revealing that the passages of paint below were in varying states of preservation and appearance. The lower dark areas of the dog and the foreground and staircase were disrupted by tiny cracks, only properly visible under the microscope which on a normal scale served to scatter light and make the paint appear unsaturated.³³ The sky at the top right and the area of wall behind the sitter’s head are affected by wide and extensive drying cracks. These had been heavily overpainted with thick strokes of oil paint matched to the khaki colour of the darkened varnish. This final area was hard to analyse and microscope examination was not sufficient to identify with certainty which of the eroded layers could be original paint applied by Reynolds and then damaged by cleaning, and which were overpaint. In some places all the varnish residues and overpaint were missing and a pale crumbly surface was revealed, appearing eaten away by past cleaning methods (fig. 12). The impossibility of cleaning this area further meant that varnish removal was stopped at this point, despite substantial residues of earlier varnish remaining over the portrait. Its appearance overall was much improved and the cool values of the blue areas reinstated, so it was decided that the residues could remain and any unification of different passages be done in the inpainting stage. The opaque and crudely applied overpaint was removed as far as possible, in all cases revealing cracks or damages far smaller in size than the overpaint and which in many cases would not require inpainting. The notion of retaining a ‘patina’ on a painting such as this is problematic, as of course the darkened material on the paint surface is from different sources, few of which are original. However, when paint has suffered such extensive cracking, lining damage and erosion, there is not much to gain by cleaning it completely and the decision to hold back seems justified.

A similar degree of tact has been applied to the treatment of the frame. This had also been restored more than once and cross-sectional analysis revealed that the level of intervention in the decorative scheme was different depending on its position in the frame’s profile.³⁴ In this vein, the treatment aimed to extend the frame’s life history rather than fundamentally altering it. Therefore, where possible traditional material and methods were used, which corresponded to those utilised in the frame’s original construction. It was also considered important not to alter the frame’s slender profile or hide its original mitre-and-keyed joinery: therefore applying a build-up (wooden addition) to the back of the frame was resisted, which might have been considered in a normal situation when the frame was just being displayed at Tate. The treatment began with stabilisation.³⁵ Next came cleaning. This followed a tailored approach, where the cleaning system used depended on the type of dirt that was being removed and the gilding beneath, which was to be retained.³⁶ With the frame clean and stable, the distracting losses in the decorative scheme could be considered and filled appropriately.³⁷

These technical decisions relate to the new, amphibious status of the painting: both a museum picture, belonging to the national collection, and a picture for display in its historical stately home. But as should be clear even from this sketchy outline, the history of the painting to this point is already far more complex. The picture we see now has mutated and changed appearance more or less dramatically over the last two and a half centuries, during its years in Reynolds’s studio and repeatedly as soon as it had been sent off to Castle Howard. And once at the house, it has been moved and recontextualised repeatedly, positioned in rivalry with Old Masters, aligned with dynastic portraiture, paired with Omai as a masterwork by Reynolds, or a mixture of these. While the considerable merit of maintaining Castle Howard as the primary home of the painting is recognised in the arrangements made in 2016, it is not the case that the stately home context is somehow immutable, any more than any individual painting (least of all a painting by Reynolds) can ever be expected to be.