Lot 96
  • 96

John Constable, R.A.

2,000,000 - 3,000,000 USD
Log in to view results
bidding is closed


  • John Constable, R.A.
  • Salisbury Cathedral from the Meadows
  • oil on canvas


Possibly the artist’s sale, London, Fosters, 16 May 1837, lot 37, as ‘Sketch of Salisbury Cathedral, from the Meadows’, to Williams for £6.10;
The Viscounts Hambleden, Hambleden Manor, Buckinghamshire;
By whom sold, London, Christie’s, 10 July 2013, lot 437.  


The following condition report has been provided by Simon Howell of Shepherd Conservation, 11, Berkeley Pl. Wimbledon, SW19 4NN 020 8946 5293, s.howell@shepherdconservation.com, an independent restorer who is not an employee of Sotheby's.The painting was examined using the ambient lighting and with the aid of a hand-held Tungsten-Halogen light. It was also viewed in Ultra-violet light and with the aid of an Infra-Red viewer. An X-radiograph of the painting was made available for examination.The painting is a work in oil paint on a lined canvas, attached to a traditional wooden stretcher.Label : There are two Christies stencils that are now illegible.Paper label on the frame, 'Exposition d... Pottie(r) ...Mballeur ...de tableau Objets d'Art, 14 Rue Gaillon Paris No. 133'The stretcher is a softwood construction with a design that is consistent with the 19th Century. It is notable for its thin section (about 1cm) and its delicately beveled leading edge. All the keys are present and original. It remains in good condition and providing an effective support for the canvas.The artist's canvas has been lined or adhered to a secondary fabric support. The lining canvas is a very fine linen material and the lining technique has employed an animal glue-paste as an adhesive. The lined canvas is attached to the stretcher using iron tacks that now show some signs of rust. The appearance of the lining and the verso of the lining canvas would suggest that it may date to the same period as the stretcher – the later half of the 19th Century. Although of some age, the lining seems to be robust and continues to provide an effective support for the artist's canvas. The process of lining a canvas would have involved some heat and pressure. Fortunately, although Constable's paint is quite soft and resinous, the lining does not seem to have unduly affected the impasto. One can point to certain passages where the highest paint has been slightly compressed, but much of this is due to the natural drying process of the paint. The artist's canvas is a slightly heavier-weight weave, also made of linen. Although the tacking margins have been removed, there are no holes or tears in the original canvas. All four edges of the original canvas retain evidence of a 'barbe' or raised 'lip' of paint where the canvas was folded around the stretcher. It therefore seems that little if any of the painted image was lost when the tacking margins were removed. A traditional method for determining whether canvas has been removed around the edges of a painting is to see if there is 'cusping' or the deformation of the weave that is created during the process of stretching the canvas. Though not visible with the naked eye, it is seen in the X-radiograph as long waves in the weave. The length of the cusping suggests that Constable prepared the canvas on a larger loom before cutting it up into its present format. The artist's technique is remarkable for its thickly impasted ground, followed by thinly applied glazes. A broad and vigorous underpainting in a modulated white pigment can readily be seen in raking light. This is followed by a medium-rich layer that moves through brown to red. The main design layers follow, which are in turn modulated with thin glazes. Throughout, the warm priming is allowed to show through the upper layers of paint. The variations of translucency are central to the overall effect. The X-radiograph reveals clearly the wide sweeps of the brush and a palette knife in applying the paint. The upper left quadrant is remarkably transparent and indicates that the storm from the left was part of the original intention. The right bank of the river is more confused in the X-radiograph and circular marks might indicated that a large tree with foliage was planned for this area as in Guildhall view, or possibly a larger structure for the bridge in the foreground.When considering the condition of this painting, the extent to which this is a quality desired by the artist and what is later unintentional thinning of the paint, is an important consideration. Close examination indicates that as the glaze passed over the tops of the impasto it naturally became thinner, allowing more of the lower ground to show through. In addition, as the glaze has dried, it has separated and formed fissures that have further revealed the lower ground. A final consideration is that as oil paint ages its transparency increases revealing more of the lower layer. Some of the effect must be put at the door of past restorations that have assisted in making the lower ground and priming slightly more prominent than was originally the case. The effect can best be seen in the dark shadow on the right of the river and the definition of the wooden structure in the foreground. Another paint effect worth noting is the use of a bitumen-type paint in the shadows of the left foreground which has 'run' or continued to flow into the surrounding paint.There is very little actual retouching to be found on the painting. Perhaps the only passage is the sky above and to the right of the cathedral where age cracks and some scratches in the surface have been disguised with delicate and accurate touches of pigment.The painting has obviously been restored sensitively and carefully, leaving some of the old varnish in the more susceptible passages such as the shadows of the left-hand trees. The existing varnish is fully saturating the original colours and providing the composition with a pleasing level of gloss.This sketch of Salisbury Cathedral is a beautiful example of Constable's mature technique and its good condition not only reveals the strength of the artist's vision but also his very individual use of his materials.
"This lot is offered for sale subject to Sotheby's Conditions of Business, which are available on request and printed in Sotheby's sale catalogues. The independent reports contained in this document are provided for prospective bidders' information only and without warranty by Sotheby's or the Seller."

Catalogue Note

This Salisbury Cathedral from the Meadows is a rediscovered preparatory work for one of John Constable’s most celebrated masterpieces, now in the Tate, London.  We are grateful to Anne Lyles for preparing the cataloguing for this lot. 

A newly discovered sketch for Constable’s Salisbury Cathedral from the Meadows
by Anne Lyles



Constable is, with Turner, one of Britain’s best loved and most significant landscape painters, a key figure in British Romantic art of the early nineteenth century. A much slower starter than Turner, it was not until the second half of his career that he painted his large-scale, ‘six-foot’ canvases such as The Hay Wain, 1821 (National Gallery, London), The Leaping Horse, 1825 (Royal Academy of Arts, London) or the late Salisbury Cathedral from the Meadows 1831 (Tate, London; formerly on long loan to the National Gallery. fig. 1), paintings which were to establish his contemporary reputation and which are largely responsible for his present day fame.

This oil sketch is one of five preliminary oil sketches which Constable made for Salisbury Cathedral from the Meadows (Tate, London, fig. 1) which is perhaps the greatest of his late masterpieces, and which was only recently acquired for the nation after a major fund-raising campaign. The recent emergence of this oil sketch from the Hambleden collection, where it was hitherto completely unknown to scholars, reveals its key role in establishing the dramatic and beautiful chiaroscuro of the final picture; the striking light effects on the cathedral in the completed picture, with its majestic spire piercing the stormy sky like a needle, are derived chiefly from the present study. Moreover, it also reveals Constable’s development of the composition, notably at the right where the familiar shape of Harnham Ridge now comes into view. It is thus one of the most exciting and important additions to the master’s oeuvre to have emerged in recent decades.  



From the mid to late 1820s, Constable had begun to diversify his subject-matter. The large exhibition pictures, or six-footers, with which he had first established his reputation in the period 1819-1825, had all shown scenes on the River Stour in Suffolk near Flatford Mill close to East Bergholt where he was born, in 1776, and where he had spent his childhood years. However from about 1825 he started to select views of inland Suffolk scenery for his large exhibition canvases, and then views of other locations which he knew well through family or friends such as Brighton and of course Salisbury. 

Constable’s connection with the city of Salisbury first arose, and was then nourished, through two important friendships, with Bishop John Fisher and with his nephew Archdeacon (also John) Fisher, both of whom – but especially the latter – were to become important patrons and supporters of the artist. Constable made frequent visits to the city to stay with one or other of them, between 1811 and 1829, at their respective residences - the Bishop’s Palace and Leadenhall – in the Cathedral Close. In the early 1820s the Bishop commissioned from Constable a mid-sized painting of the Cathedral from the grounds of the Bishop’s Palace (Victoria & Albert Museum, London), and in subsequent years Constable produced a number of variants of this composition.  

However it was Archdeacon John Fisher who, in 1829, encouraged Constable to paint the larger, more ambitious – and, of course, unique - version of Salisbury Cathedral as a distraction from the grief the artist was suffering after the death of his wife Maria in 1828. The final painting shows the Cathedral, viewed from the North-West across the River Nadder, dramatically set against a backdrop of billowing storm clouds. A streak of lightning is visible above the Cathedral roof whilst an over-arching rainbow, which ends at Leadenhall, Fisher’s house in the cathedral close, indicates that the storm is about to pass. The painting was exhibited by Constable at the Royal Academy in 1831 and then again at the British Institution in 1833, at the Birmingham Society of Arts in 1834 and at the Worcester institution in 1836. However – like most of his major late pictures – it failed to find a buyer (perhaps in this case because Constable was loath to part with it). It remained in his studio – where he continued to retouch it – until his death in 1837.  


Constable’s working process in the second half of his career

In the first half of his career when – as yet unmarried – he was still making regular visits to the family home in East Bergholt in Suffolk where he had first been inspired to paint the local scenery, Constable generally produced highly naturalistic paintings. His practice in these earlier years was to make copious pencil drawings or oil sketches on the spot, which he would then use in the preparation of his exhibition pictures which thus remain more or less faithful to the original scenes. Sometimes he executed these paintings of Suffolk subjects in his London studio but, especially between about 1814 and 1816, he would often work on them partly in the open air; that is in front of the scenes themselves.  

However, once Constable had finally committed himself in 1816 to marrying his long term sweetheart Maria Bicknell, he became permanently based in London and, distanced from his native scenes, he began to develop a method of recreating his Suffolk material at one remove, working entirely in the studio. Unlike Turner, Constable did not have a good visual memory, and he rarely managed to paint his later studio landscapes without having recourse to a variety of studies and sketches. Whilst he could (and often did) turn for inspiration to earlier plein air sketches when planning his new subjects, these rarely supplied enough detail for him to complete a large-scale, often complex composition.

So, in the second half of his career, Constable began to depend increasingly on compositional studies, in pencil or in oils, to try out ideas for his larger landscapes. The number of studies he might make in connection with a given subject varies, depending on the degree of difficulty he experienced when attempting to realise its composition. His preparatory studies also vary greatly in scale – they might be small, mid-size or even painted on the same, ‘full’ scale (usually about six foot in width) as the intended final canvas. Sometimes they seem to concentrate on compositional concerns, at other times they focus on lighting effects and tonal contrasts. Indeed, in later life Constable talked frequently about the importance of chiaroscuro (or light and shade) in serving to animate a landscape, especially in the treatise he wrote to accompany the group of landscape mezzotints he published between 1830 and 1832, English Landscape.


Early ideas for Salisbury Cathedral from the Meadows, 1831

After many years striving for professional recognition, in February 1829 Constable was finally elected a full Royal Academician. He was, however, still grieving for his wife, Maria, who had died the previous November, and his closest friend the younger John Fisher encouraged him to apply himself ever more rigidly to his profession as a way of distracting himself from his loss.

The first large landscape Constable exhibited as an Academician, in the spring of 1829, featured the ruins of Hadleigh Castle dramatically sited by the estuary of the River Thames (Hadleigh Castle ,Yale Center for British Art). Given its antiquarian and ‘picturesque’ echoes, this was in many ways a more conventional subject that those he had adopted for previous six-foot landscapes. The fact that it was well received by the critics may indeed have encouraged him to select another subject with antiquarian overtones – the large Salisbury – for his next main Academy exhibit. In all events, Constable had certainly decided on the Salisbury composition by August 1829 when Fisher wrote to him that ‘I am quite sure the “Church under a cloud” is the best subject you can take’.1

Constable stayed with Fisher at Leadenhall in Salisbury twice in 1829, the first time for about three weeks in July, and then again for a period of up to two weeks the following November. It was during these visits – but also perhaps between and certainly after those visits when in his London studio – that Constable started working on his ideas for a large composition of the Cathedral. His early drafts, mainly in pencil, show him exploring the best viewpoints for his subject. It is clear that from a very early stage he decided to represent the Cathedral from a westerly direction, no doubt as it was less hemmed in by other buildings from this aspect but also as it could be represented from this side across the city’s distinctive water meadows. He made one drawing of the Cathedral from due west.2 However, nearly all his other preliminary pencil sketches for the subject are taken either from the North-West at a point near Fisherton Mill, or from the South-West near Harnham – and indeed one of the two drawings he made from the south-west may have suggested to him the idea of incorporating a rainbow in the final picture (Private Collection).3 Eventually he settled on a view from the north-west, as shown in the present painting.


The role of this sketch in the evolution of Salisbury Cathedral from the Meadows, 1831

Having established the ideal viewpoint for his composition, Constable then needed to decide the exact viewing distance for the composition for it, that is to work out much of the picture space might be taken up by the cathedral itself, with its striking silhouette, and how much by the surrounding landscape. Furthermore, given that by August he had settled on presenting the cathedral under stormy conditions – or ‘under a cloud’ as Fisher described it – Constable also needed to consider the balance of light and shade. Altogether there are eight directly related preparatory sketches for the final painting, three in pencil, and five in oils. The Hambleden study is one of these latter five, and played a key role in establishing the critical chiaroscuro of the picture as well as providing an important detail in the architecture of the cathedral’s west front.

The three preparatory drawings relating to the composition for Salisbury Cathedral from the Meadows were all made by Constable in a sketchbook measuring approximately 9 by 13 inches (22 by 33 cm.) which he used in 1829. Like most of his sketchbooks, this book has now been dismantled and the three pencil sketches in question are now in different collections: one is in the Fitzwilliam Museum, Cambridge (fig. 2); another is at the Yale Center for British Art; and the third is in the Lady Lever Art Gallery, Port Sunlight.4 The first two of these, one (in the Fitzwilliam) showing the cathedral from close to and the other (at the Yale Center) from a greater distance, both appear to have been made in the open air, either during the July or the November visit. The third drawing, in the Lady Lever Art Gallery, is a conflation of these two earlier drawings, and thus likely to have been made by Constable as a compositional sketch, in the studio, perhaps even after he’d returned to London towards the end of the year.

The three smallest of the five compositional oil studies Constable made for the final picture are almost direct transcripts of these three pencil drawings. Two of them, taken from the two plein air drawings in Cambridge and New Haven, are very small - indeed at approx. 7 ½ by 11 inches (19 by 28 cm.) they are even smaller than the original pencil drawings - rather dark and rapidly executed on paper prepared with a reddish-brown ground. One (fig. 3) is now in a private collection, the other is in the Clark Art Institute, Williamstown.5 Constable seems to have painted them chiefly to experiment with the position of the horizon lines, and it is noteworthy that both of their related pencil drawings have ruled lines in the area of the horizon. He may also have wanted to investigate how a stormy sky might affect the illumination of the cathedral (especially its spire) – and in the oil sketch in a private collection, most of the building is shown plunged into deep shade.

The third small oil sketch, now in Tate Britain, is very closely based on the pencil drawing in the Lady Lever Art Gallery.6 In fact the latter is actually squared (with an interesting diagonal grid) for transfer, and this sketch (on canvas) was evidently made from it as it bears a similar diagonal grid underneath the paintwork. This latter oil sketch is usually regarded as the one from which Constable then worked up the full size sketch in the Guildhall Art Gallery (fig. 4) – from which the final painting derives – and certainly the Tate and Guilldhall sketches are very close compositionally. If this is the case, where then does the Hambleden study fit into the sequence of sketches, and what was its exact role?

Constable always found it difficult to project his ideas for landscape on a large scale, and the fact that Salisbury Cathedral from the Meadows has so many related studies is an indication of how much work he put into its conception, as well perhaps as the challenge he had with its realisation. Once he had a firm idea for a new large picture destined for exhibition, he would try to work on it over the winter, and then send it in to the Academy the following spring. It is clear from Constable’s correspondence with Fisher that he intended the large Salisbury to be his main exhibit in 1830, but in the event he sent in another picture that year, Helmingham Dell (Nelson-Atkins Museum of Art, Kansas City), instead.7

This raises the intriguing question as to how far Constable had progressed with work on the large Salisbury by the winter of 1829-30 and, indeed, how many of the smaller preliminary oil sketches – not to mention the full-scale sketch and the picture itself - he had even painted (or started painting) by that date. What does seem likely is that he made the two small rapidly executed oil sketches on paper (Private Collection and Clark Art Institute) at an early stage in the gestation of the picture, perhaps in the winter of 1829 when the composition was still fresh in his mind and he was anxious to try out his ideas. However, once Constable had decided on the more close up view as his best model for the picture - that is to say the sketch in a private collection - how soon afterwards did he proceed to paint the Hambleden study or the small Tate sketch, and indeed which of the latter two is he likely to have painted first?

It was pointed out by Constable scholars Leslie Parris and Ian Fleming-Williams in 1991 that from a stylistic point of view the small Tate oil sketch seems very different from the other two small oils similarly based on pencil drawings from the 1829 sketchbook, which tends to suggest a lapse in time between their execution.8 Meanwhile, the Hambleden study is actually quite close to the oil sketch in a private collection (fig. 3). Both have a low horizon and both show the cathedral sited close to the centre of the composition. However the Hambleden study reveals Constable experimenting with extending the composition at the sides, especially on the right where the familiar shape of Harnham Ridge has now come into view. Whilst it lacks the figurative detail of the foreground fisherman in the sketch in a private collection, it does include a lone figure – perhaps a shepherd - on the opposite side of the River Nadder. The Hambleden study also shows Constable paying particularly close attention to the light and shade in the composition, especially the way the stormy sky affects the illumination of the cathedral. For unlike the sketch in a private collection where most of the building is in shade, here by contrast the cathedral is dramatically lit by some of the sun’s rays emerging through gaps in the clouds.

 It is conceivable that Constable painted the Hambleden study in the winter of 1829-30, in which case it might precede the small Tate sketch by as much as a year. In whichever year it was painted, it certainly came in very useful for Constable when he was refining some of the details in the final painting over the winter of 1830-31. For the striking light effects on the cathedral building in the latter are derived chiefly from the Hambleden study (all the other preparatory sketches show the cathedral building more or less in shadow). Moreover, the dramatic stormy sky in the full-scale sketch in the Guildhall (fig. 4) also derives more closely from the Hambleden study than the other sketches. Indeed the cluster of black storm clouds in the full-scale sketch to the right of the cathedral spire was once closer in appearance to the formation seen in the Hambleden picture until Constable decided to knock them back in the former by overpainting parts of them in white.

Furthermore, Constable may have depended on the Hambleden oil for helping provide some of the architectural detail of the cathedral’s west front in the finished picture. In 2002 John Gage pointed out that in the full-scale sketch Constable mis-represented the structure of the West front, compressing it so that the flanking pinnacles seem to run directly into the gable of the nave, whereas in reality they are separated from it by a broad crenelated parapet.9 Gage argued that, to correct the error in the final painting, Constable must have had to refer to an antiquarian publication such as, for example the Wiltshire volume of John Britton’s Beauties of England and Wales, 1814, a copy of which Constable owned.10 However, the Hambleden sketch proves that Constable was well aware of the West front’s true architectural structure, as here it is represented correctly – even if not in any great detail. Constable must have decided not to waste time delineating it with any particular accuracy in the full size sketch when he knew he could find details of the structure elsewhere, either in more detailed earlier drawings of his own, or indeed in publications such as that by Britton.


Provenance and condition

The earliest sale of this painting, as with many of the sketches which Constable retained until his death, is uncertain, however a good deal may be inferred from the records of his posthumous sale at Fosters in 1837 (A Catalogue of the Valuable Finished Works, Studies and Sketches, of John Constable, Esq. R.A, Foster, 15-16 May 1838). The large Tate Salisbury (lot 79 in the Foster sale) achieved £ 110.5, whilst the full-scale sketch for the picture in the Guildhall is generally identified with lot 37 (‘Sketch of Salisbury Cathedral from the Meadows’) which achieved £6-10 (which was a fairly average price for the large studio sketches). However, no less than seven lots (or works in mixed lots) in the Foster’s sale present themselves as potential candidates for this picture, described simply as ‘Salisbury Cathedral’, or ‘Salisbury from the Meadows’ or simply ‘Salisbury Meadows’ (nos. 10, 12, 13, 14, 23, 34 and 50). Indeed some of these works sold for higher prices than the large Salisbury sketch in the Guildhall. Lot 50, for example, Salisbury Meadows, which might be identifiable with this painting (even though described as ‘painted from nature’), sold for as much as £ 35-14-0.The date at which the painting entered the collection of the Viscounts Hambleden is uncertain, however the most prolific collectors of that family were W.H. Smith (1792-1865), founder of the eponymous Book sellers, and his son W.F.D. Smith (1868-1928), the Second Viscount. They amassed a fine collection of Old Masters, including a masterpiece by Claude Lorraine, as well as several exceptional British old masters, including the present painting. At the time of the dispersal of the contents of Hambleden Manor at Christies in 2013, it is interesting to note that this painting was overlooked, as was, the Claude (which was first offered at South Kensington as a copy after Claude, before being withdrawn and selling for over £5m at Christies King Street later in the year). While the Claude was presented in a relatively straight forward condition, the present work by Constable was heavily retouched with a dark and opaque pigment which probably dated to the late 19th or early 20th century, in a misguided attempt to ‘finish’ the painting, thus depriving it of its lively, sketchy quality. This intervention should not surprise us, for the same fate befell the large Guildhall sketch, at some stage at the end of the nineteenth century, before being cleaned and restored in 1951. Thankfully the retouchings on the present painting were readily soluble in the course of its recent cleaning, and Constable’s original and brilliant conception has been once again revealed. 

Anne Lyles is the former Curator of 18th and 19th Century British Art at Tate Britain.


1. R.B.Beckett (ed), John Constable’s Correspondence: VI: the Fishers, Ipswich 1968; letter from John Fisher to John Constable, 9 August 1829, pp.250-1.

2. Clark Art Institute, Williamstown; see Graham Reynolds, The Later Paintings and Drawings of John Constable, New Haven and London, 1984, no.29.20, plate 719.

3. His pencil views from the north-west are discussed below. But for those from the south-west ( from near Harnham) see Reynolds 1984, no.29.91, plate 720 ( Salisbury Cathedral seen from the River, Victoria & Albert Museum) and R 29.28, plate 726 ( Salisbury Cathedral from the Meadows, private collection).

4. The Fitzwilliam drawing is Reynolds 1984, no.29.43; that at the Yale Center for British Art is R. 29.42; and the one at the Lady Lever is R.29.13.

5. For the oil sketch in a Private Collection ( in the Hart collection in 1991), see Reynolds 1984, no 31.4; and for that in the Clark Art Institute, R.31.3.

6. The oil sketch in Tate Britain is Reynolds 1984, no. 31.5. See also Constable: the Great Landscapes, ed. A.Lyles, 2006, exh cat., Tate Publishing, pp.180-3 for a fuller discussion of how this relates to the full-scale sketch in the Guildhall (R. 31.2) and the final exhibition canvas in Tate Britain (R.31.1).

7. See letter from Fisher to Constable cited note 1 and also R.B.Beckett 1968, letter from Fisher to Constable 3 Sept 1829, pp.252-3. Helmingham Dell, 1830, is Reynolds 1984, no.30.1 (plate 770).

8. Constable, exh. cat., Tate Gallery 1991, p. 363.

9. Gage in Constable: Le Choix de Lucian Freud, exh cat., Paris, Grand Palais, 2002, pp.232-4.

10. A copy was in the sale of Constable’s effects at Foster’s sale rooms, 22 Dec 1838; see L.Parris, C. Shields and I. Fleming-Williams, John Constable: Further Documents and Correspondence, London and Ipswich, 1975, p.46.

11. John Constable, Salisbury Cathedral from the Meadows, Tate, London, ref. no. N01814.