By using this website you agree to our Cookie policy

April 2009

Vol. 151 / No. 1273

Alfred Stevens’s images from ‘The Faerie Queene’ rediscovered

By Teresa Sladen

SOME YEARS AGO nine of Alfred Stevens’s long-lost paintings based on Spenser’s The Faerie Queene were found in a shop in Marylebone, London. Once the focal points of a decorative scheme for the drawing room of 11 Kensington Palace Gardens, they had not been seen for almost a century. Painted in 1854–55, their importance lies in the pivotal position they hold within the artist’s œuvre. In 1853 Stevens had been commissioned to copy Raphael’s paintings on the vault of the Stanza della Segnatura for the Italian Court of the Crystal Palace at Sydenham. In 1856, immediately after he had finished the Faerie Queene series, he embarked on his two greatest works, the Wellington Memorial in St Paul’s Cathedral and the decoration of Dorchester House, London (destroyed).

 

Hugh Stannus, his biographer and one-time assistant, claimed that the Faerie Queene paintings contained ‘all the quality of tenderness with decision that characterised the work of the Urbinate’.1 This is questionable; while the debt to Raphael is clear, there is much to distinguish Stevens’s heroines from their Renaissance antecedents. This is especially true of the more warlike characters, Britomart and Radigund, who entirely lack the lyrical grace and sweetness of expression for which Raphael’s work was so greatly prized in the Victorian era. These women, with their hefty limbs and fearless gaze, would look more at home in a Dorset farmyard remembered from Stevens’s youth than in the celestial setting of Raphael’s personifications of Arts and Sciences. They are, in fact, the forerunners of Stevens’s two great statues of Truth and Valour, those Herculean women who sit to either side of the Wellington Memorial calmly crushing two loathsome males, Cowardice and Deceit, beneath their feet.

 

11 Kensington Palace Gardens, now the French ambassador’s residence, is a small mansion designed in the Italian Renaissance style. It was built between 1852 and 1854 by the architect Sidney Smirke for a wealthy Spanish merchant, Cristobal de Murrietta.2 Stevens’s scheme for the drawing room was designed to match the classical style of the house. The decorative framework of this scheme, which included key-patterns, foliage and grotteschi on the ceiling, cornice and pilasters, was painted directly on the plaster, while the figurative elements were painted on canvas attached to the walls and ceiling.

 

When Stannus visited the house in 1890 it was in poor condition, having stood empty for a number of years. He noticed that some of the canvases on the ceiling were beginning to peel off.3 Four years later the house was bought by Robert Perks MP, who spent £10,000 on its extension and redecoration.4 It was almost certainly at this time that the Faerie Queene paintings were stripped from the walls.5 Fortunately, someone – one of the decorators perhaps – had the sense to salvage them. Their rediscovery has made it possible to plot every stage in the development of the scheme from that first proposed by the decorator Leonard Collmann through to the one by Stevens that was finally adopted.

 

Three drawings of different schemes are in the Victoria and Albert Museum, London. Two are watercolours bearing Collmann’s stamp, one of which, a rather dull affair with no figurative work, is probably the first scheme proposed by Collmann’s firm.6 It is possible that the client’s dissatisfaction with it prompted Collmann to ask his friend Alfred Stevens to provide a more lively alternative.

 

Leonard William Collmann (1816–81) played an important role in Steven’s personal and professional life. The two are said to have met as young men when they were studying art in Florence.7 On his return to England Collmann became a pupil of Smirke, setting himself up as a decorator and furniture-maker; Smirke generally used Collmann’s firm to decorate his buildings.8 By the time Stevens returned from Italy in 1842 Collmann was in a position to offer his friend the kind of specialist decorative work to which his skills were particularly suited.   

 

The more sumptuous of the two schemes Stevens designed for 11 Kensington Palace Gardens combined sculpture with polychromatic decoration. In 1855 Stevens was at a turning point in his artistic development; from this time on almost all his decorative schemes involved moulded as well as flat decoration. The sculptural scheme he proposed seems to have been a first step in this direction and, for this reason, is likely to have been the one he favoured. It incorporated a series of statuettes silhouetted against sky-blue circular or rectangular panels, flanked by pink marble pilasters and set against the lemon yellow of the walls.9   

 

Stevens’s second drawing, in pencil, shows the scheme that was actually carried out (Fig.6).10 This drawing is important because it is the only document to show the Faerie Queene paintings in relation to their decorative framework. In Stannus’s book the photograph of the painting of Una shows the trompe l’œil frames that surrounded the paintings (Fig.5).11 Further information on the decorative detail and colour on the ceiling and cornice is supplied by two full-size working drawings in the Victoria and Albert Museum.12 Finally, a photograph of the drawing room taken in 1937 shows the cornice and what remained of the decoration on the ceiling after the paintings on the walls had been removed.13   

 

The paintings on the walls all portrayed female characters drawn from The Faerie Queene: Alma, Briana, Cymoent, Belphoebe, Amoret, Radigund, Serena, Una, Britomart and Mercilla. A roughly drawn plan among Stannus’s papers shows where each painting was placed,14 and also indicates that there were eight oval panels on the ceiling, each containing a single figure. Judging by Stannus’s descriptions of these, one of which is just visible in the 1937 photograph, their subject-matter was not related to The Faerie Queene. 

 

This scheme was less colourful than the sculptural one. The walls and ceiling were painted cream, while the garlands, grotteschi and other decorative details were picked out in muted shades of brown, grey and blue. The figures in the oval panels at the corners of the ceiling were painted in grisaille with darker backgrounds. In contrast to this, the backgrounds of the paintings on the walls were cream, and these figures, with their strong silhouettes and colourful draperies, would have stood out as the brightest elements of the decoration.

 

The overall form of the scheme – with its small rectangular and oval panels painted on canvas and set against a pale background in a framework of lightly drawn foliage and other devices – may be compared to the type of decoration promoted by Robert Adam. Where it differs most obviously, however, is in its employment of single statuesque figures for the panels on the walls instead of some form of mythical scene, and in the substitution of a painted framework of foliate decoration rather than the delicate plasterwork frequently used in the later part of the eighteenth century. And the mood has altered too; the lighthearted decoration of the Georgian era has given way to something more serious.

 

The early years of Queen Victoria’s reign had seen a gradual change in the public’s taste for art; myth and fantasy were being replaced by genre and history painting, including the illustration of literary subjects, and there was a new emphasis on patriotism. This was confirmed in 1842 when those who, like Stevens, took part in the first competition for the decoration of the New Palace of Westminster, were asked to choose subjects from the works of three major British writers: Shakespeare, Spenser and Milton.15 In this context, The Faerie Queene would have been a particularly suitable choice since the tribute paid in the poem to Queen Elizabeth I could easily be transferred to the young Queen Victoria.

 

At least two of the characters from The Faerie Queene painted by Stevens – Mercilla and Belphoebe – were intended by Spenser to represent the Virgin Queen. In the Victorian period, however, it was Una, the heroine of the first book, who was identified with Queen Victoria. In the poem Una, who represents the True Religion, is separated from the Red Cross Knight of Holiness (the Anglican Church) by the wicked magician Archimago. Lost in the wood, she meets the Lion (England), who protects her. In 1839 a £5 coin was minted in which Queen Victoria was represented by the image of Una and the Lion.16

 

However, the patriotic interpretation of The Faerie Queene was unlikely to have been the only thing that prompted Stevens to illustrate this poem. He is known to have acquired a copy of Spenser’s works as an impoverished young man in Italy.17 With little formal education but a liking for books he probably came to know the poem very well.18 Was this why he treated its subject-matter so differently from his fellow artists, depicting a far wider range of female characters than before? It may be argued, of course, that the reason for this had more to do with the nature of his scheme. He was, after all, portraying single figures rather than the scenes in which they played a part. Even so, it is worth noting that the only female characters frequently painted by members of the Royal Academy were Una, Serena and, on occasion, Britomart. Most painters favoured either images of unassailable female virtue, as, for example, Una in the Dark Wood, or those of passive fragility, such as Serena rescued by Sir Calepine from the clutches of the Blatant Beast. Britomart, a woman warrior, could not be represented in this way.

 

In 1833 William Etty, whose lush depictions of naked women often offended Victorian sensibilities, exhibited a painting at the Royal Academy showing Britomart rescuing Amoret from the enchanter Busirane (Fig.8). Clad from head to toe in shiny black armour, Britomart is about to smite the cringing wizard, an image of female domination very unusual at this date. An early drawing by Stevens (Tate, London) suggests that he knew and admired this picture;19 it shows Amoret, bound and half naked, in the same attitude as in Etty’s painting, though reversed. Later he used this image in a design for the decoration of a covered vase (Fig.7). But the figures of Amoret and Britomart that he painted for Kensington Palace Gardens have nothing in common with those in Etty’s painting apart from the vibrant spirit invested in Britomart.

 

The sculptors who chose subjects from The Faerie Queene did not extend the range of characters either. In the late 1840s the development of Parian-ware provided the ideal vehicle for the reproduction of small-scale ornamental sculpture. The first of these sculptures to illustrate a subject from The Faerie Queene was modelled by John Bell in 1847 for Summerley’s Art Manufactures.20 Entitled Purity, it depicted a nude Una seated on the lion. This proved very popular, and the same artist subsequently made two more statuettes of literary heroines, one of Goethe’s Dorothea and the other of Tasso’s Clorinda.21 All these figures were manufactured by the firm of Herbert Minton and shown at the Great Exhibition. Between 1851 and 1852 John Rose & Company produced three miniature sculptural groups portraying scenes from The Faerie Queene; the Vision of the Red Crosse Knight, Sir Calepine rescuing Serena from the Blatant Beast and two of Britomart, one showing the rescue of Amoret and the other her battle with Busirane.22

 

Through his appointment to the Government School of Design and subsequent involvement with a wide variety of manufacturers, Stevens would have come into contact with many of the men making and promoting these products. Minton & Company, for example, produced the ceramic panels for his stoves and fireplaces, several of which were shown at the 1851 Exhibition. But he never made sculpture of this kind, nor do any of the figures in his decorative scheme for Kensington Palace Gardens echo the designs of any of the sculptural groups produced in Parian-ware.

 

At first sight Stevens’s decision to paint ten single figures suggests a pantheon of ‘worthies’ as might have been chosen to decorate a long gallery or hall. But if this were really the case, his choice of heroines rather than heroes was extremely unusual. It seems more likely that, to suit current taste, he was merely adapting the earlier convention whereby human figures were used to personify abstract qualities. This was a formula he had followed before, as, for example, at Deysbrook Hall near Liverpool, where he painted the walls with a series of women representing states of mind such as Joy, Watchfulness and Constancy, etc. He seems to have approached the characterisation of the Faerie Queene figures in a similar way but brought the scheme up to date by giving it an English literary theme.

 

Before embarking on the paintings themselves Stevens made a great many preliminary drawings. He had to rely on variety of outline, form and gesture both to express the character of the women and to give the scheme visual interest. He first made a series of tentative sketches to establish the overall form of each figure (Fig.9);23 here the outlines of Una and Britomart have already taken shape while that of Serena has not been resolved.

 

Una and the lion was an immensely popular image in the Victorian period; examples range from paintings exhibited at the Royal Academy to mass-produced ornamental inkstands. But Stevens’s painting of Una resembles no previous work. The voluminous form of her habit, combined with the way she is grouped with the lion, creates a triangular shape. This makes for a stable composition and imbues the figure with a sense of calm. Later, Stevens worked up the detail in a highly romanticised sketch that, like the previous one, seems to have been wholly drawn from his imagination.24 The same holds good for the final painted version; although he may have made studies of minor details, he seems not to have used a model for the figure as a whole (see Fig.5).

 

In contrast to the calm image of Una, that of Britomart is restless and energetic, shown in the act of donning her helmet and clearly about to jump to her feet. She is strikingly different, both in conception and style, from all the other figures in the series. Nor does she resemble any of the images produced by Stevens’s contemporaries. Even when Britomart began to be adopted as a symbol of emancipated womanhood in the later part of the nineteenth century, artists continued to represent her as a graceful being, not with the rude vitality of Stevens’s image. So what made him represent her in this way? It seems likely that he was drawing on a knowledge of classical art he had gained in Rome, and from the precedent of John Flaxman and other Neo-classical artists who often portrayed woman in violent motion.

 

The Nereid Cymoent plays only a minor part in Spenser’s poem. She is pictured skimming over the waves in a chariot drawn by dolphins in search of her son Marinell who has been grievously wounded by Britomart. Spenser’s description may have reminded Stevens of Raphael’s Galatea in the Villa Farnesina in Rome. Cymoent’s twisted pose, fluttering drapery and pair of reined dolphins echo aspects of Raphael’s painting. Smirke had already used dolphins in the decoration of the exterior of the house;25 Stevens included small dolphin grotteschi in the horizontal bands of decoration beneath the paintings.

 

Two preliminary drawings of Cymoent survive, one in the Ashmolean Museum, Oxford,26 and the other in the Fitzwilliam Museum, Cambridge (Fig.10).27 In the earlier of these the left arm is hidden behind the shoulder while the right one stretches forward to hold the reins. In the second version, similar to that in the final painting, the diagonal nature of the composition is emphasised by the left arm being drawn up behind the head so that the reins of the dolphins are carried up to this point. Both drawings are squared for the purpose of enlargement, and some parts of the second are inscribed with their colours. No further drawings have been found of this figure but there must have been others to work out the drapery and sculptural form of the seat.

 

Stevens made a great many drawings of drapery, usually in red chalk, exploring the way that fabric behaved. One fine example was made for the painting of Amoret (Fig.12).28 In both drawing and painting the figure is passive, all the excitement of the composition lying in the explosive character of the drapery. Of all the Faerie Queene paintings this is the one which most obviously recalls Raphael’s personifications of Arts and Sciences on the vault of the Stanza della Segnatura (Fig.11), but if the drawing of Amoret is compared with that of Raphael’s figure of Justice, it is evident that the artists approached the depiction of drapery in very different ways: Raphael gives it a subservient role while Stevens allows drapery to take on a life of its own, with a more dynamic effect.

 

The extent to which Stevens relied on the use of live models for the Faerie Queene paintings is not entirely clear. He is known to have had a lay figure and he may have used this when the pose – as, for example, in the case of Cymoent – would have been hard for a model to hold.29 In some cases only details – a hand, a tensed foot or the profile of a face – were drawn from life, but at least two of the paintings were closely based on extensive studies of nude, or partially nude, women. One of these is that of Briana.

 

Briana’s unscrupulous lover, Crudor, had refused to marry her unless she gave him a cloak made of human hair. To this end, she employed a henchman to tear out the beards of youths and the tresses of maidens who passed her castle. When devising his painting of Briana, Stevens made use of a drawing of a nude seated woman with draped thighs (Fig.13).30 While the model looks down, seems thoughtful and self-contained, in the painting Briana stares into the distance, pining, perhaps for her vicious lover (Fig.14). A further drawing shows the lower part of a woman’s body entirely swathed in drapery, as it appears in the painting.31 It would seem that Stevens built up his design by combining the upper part of the drawing of the nude model, subtly altered in pose, with the drapery study.

 

The penultimate stage in the design process involved the making of the cartoons. Although none for this project by Stevens survives, two sets of cartoons appear to have been traced by Reuben Townroe from the originals. Townroe was one of the students at the Sheffield School of Design who came under Stevens’s spell while the latter was the chief designer for the ironfounders Hoole & Company in 1850–51. He moved to London in 1859 and thereafter often worked as Stevens’s assistant, but he was not in London when work on 11 Kensington Palace Gardens was in progress, and must have copied Stevens’s cartoons at a later date.32

 

The Faerie Queene figures were painted in Parris’s medium on canvas fixed to the walls.33 This medium was a mixture of mastic varnish, copal and wax which, when combined with pigment, gave the painting a matt surface similar to that of fresco.34 The nine surviving paintings are in fair condition, though they have suffered from some later alterations, as is evident when they are compared with the photographs in Stannus’s book. Unfortunately someone repainted the cream backgrounds and, in the process, damaged the figures’ silhouettes and blotted out such details as fluttering drapery. In addition, sepia paint has been used to outline certain details and darken some of the shadows (Fig.15).

 

Stannus tells us that Stevens used the coloured medium very sparingly so that the white of the priming showed through to constitute the highlights. It is possible to see that it is not just the outlines of the figures that have been changed; in some cases, the highlights have also been repainted. As a result, the ‘logic’ of some of the drapery has been lost and, in the worst cases, it is easier to judge the original quality of the paintings from Stannus’s photographs (Fig.16).

 

Stevens made ten paintings for his Faerie Queene cycle but the one depicting Mercilla, ‘a mayden Queene of high renowne’, has been lost and is now only known from Townroe’s tracing.35 Those of Alma, Serena and Belphoebe survive but a discussion of their attributes would add little to this essay. The image of Radigund, however, cannot be left out as Stevens was more strongly engaged with the design of this figure than that of any of the others in the series (Fig.19), making a great many studies of the model who also appears, similarly posed, in several of his later, more important, works. Radigund was the Queen of the Amazons. In Spenser’s poem, she overcomes the knight Artegall and imprisons him in her castle where he is forced to sit spinning flax until rescued by his lover Britomart. Stevens worked out his imaginative conception of Radigund in a number of preliminary drawings similar in type to those made for Cymoent.36 But the realisation of this monumental female warrior was further aided by a fine study of a model (Fig.17) who sat for him on many other occasions.37

 

Kenneth Towndrow, who had access to a collection of records now lost, says that Stevens ‘did not use the ordinary professional model in his work, always preferring to commandeer the help of friends or members of his household’.38 In this he may have been conforming to a general trend that, from the 1840s onwards, saw a shift away from the use of professional models in favour of people personally known to the artists concerned.39 The vast majority of Stevens’s portraits were made of personal friends; he does not seem to have wished to paint those who were not, in some way, close to him.

 

The drawing of the nude model (Fig.17) is sometimes said to have been a study for the figure of Valour on the Wellington Memorial.40 There is no surprise in this since Radigund’s pose prefigures that of the statue in the later work. In 1857, the year in which Stevens won the competition for the Wellington Memorial, he also entered a competition to design a memorial for the Great Exhibition. The maquette he made, now in the Victoria and Albert Museum, has four seated women representing the Continents round its base. For this he made a series of drawings of the same nude woman in a very similar pose but seen from different viewpoints,41 as well as several drawings that show this model holding the same pose as that of the statue of Valour on the Wellington Memorial (Fig.18).42 In fact, the designs of both Valour and of the figures representing the Continents on the 1851 Exhibition Memorial are closely connected, and are the progeny, I would suggest, of Stevens’s painting of Radigund from the Faerie Queene series.

 

A study of Stevens’s Faerie Queene paintings appears to confirm that he was largely out of step with his contemporaries in Britain. Having spent his formative years in Florence and Rome, he had absorbed too much of the spirit of Neo-classicism either to wish, or to be able, to adapt his style to meet the prevailing taste of his compatriots. Certainly none of the images he painted of the heroines of Spenser’s poem reflect those produced by other English artists in the mid-nineteenth century, be they Royal Academicians, ornamental sculptors or book illustrators. Stevens was, in a sense, both behind and ahead of his time. On the one hand he adhered firmly to the classical tradition at a time when Gothic was being hailed as both the national and the only truly Christian style. On the other, he looked ahead to the doctrine of ‘Unity of Art’, the marriage of painting, sculpture and architecture, so warmly embraced in Britain in the later part of the nineteenth century.  

 

 

I would like to thank Philip Ward-Jackson for his help and encouragement with this

article.

1 H. Stannus: Alfred Stevens and his Work, London 1891, p.15.

 

2 Survey of London: Northern Kensington, London 1973, XXXVII, pp.165–66; London, National Archives, CRES 35/3083.

 

3 Royal Institute of British Architects (hereafter cited as RIBA) Drawings and Archives Collection, London, StH/3/8/1–9.

 

4 National Archives, document cited at note 2 above.

 

5 Ibid. The 1905 Insurance Policy of the house makes no mention of Stevens’s scheme but includes all the other known decorative works. See also London, National Archives, CRES 43/80. A photograph of the drawing room taken in 1937 shows the Faerie Queene paintings had gone by then.

 

6 Department of Prints and Drawings, Victoria and Albert Museum, London (hereafter cited as V. & A.), no.8594.13.

 

7 Obituary of A. Stevens, The Builder 33 (1875), p.405.

 

8 F. Litchfield: Illustrated History of Furniture, London 1922 (7th ed.), p.377.

 

9 V. & A., no.8594.11.

 

10 V. & A., no.D1219–1908.

 

11 Stannus, op. cit. (note 1), pl.XXVIII.

 

12 V. & A., nos.D1220–1908 and D1245–1908.

 

13 Photograph cited at note 5 above.

 

14 Document cited at note 3 above.

 

15 Advertisement, The Art Union 4 (1842), p.93.

 

16 N.K. Farmer: ‘A Moniment Forever More: The Faerie Queene and British Art,

1770–1950’, Princeton University Library Chronicle 52 (1990), p.34.

 

17 J.M. Moore: ‘Notes on Alfred Stevens’, The Art Journal 2 (November 1903), p.342.

 

18 G.C. Eaton: ‘Blandford’s Greatest Man’, The Weekly News of Dorset and Hampshire (13th February 1890).

 

19 Tate, London, N02003.

 

20 G.A. Godden: Minton Pottery and Porcelain of the First Period 1793–1850, London 1968, p.98.

 

21 Ibid., pp.98–99.

 

22 P. Atterbury, ed.: The Parian Phenomenon, Somerset 1989, pp.243–44, figs.812

and 814.

 

23 V. & A., no.E2525–1911.

 

24 V. & A., no.E111–1975v.

25 Dolphins’ heads adorn the cornice of 11 Kensington Palace Gardens.

26 Ashmolean Museum, Oxford, no.1946–315.

 

27 Fitzwilliam Museum, Cambridge, no.2214(2).

 

28 British Museum, London, no.PDB15252.

 

29 The boxwood lay figure Stevens made for himself early in his career was among

the objects auctioned after his death; see K.R. Towndrow: The Works of Alfred Stevens

in the Tate Gallery, London 1950, Appendix, p.130.

 

30 Idem: Alfred Stevens, London 1939, p.111, no.17a.

 

31 Fitzwilliam Museum, Cambridge, no.2229v.

 

32 The cartoons formerly held by Tate, now in the V. & A., have already been identified as the work of Townroe, but those in the RIBA are catalogued as the originals. The two sets of cartoons are identical, and the handwriting on both is that of Townroe.

 

33 Stannus, op. cit. (note 1), p.15.

 

34 L. Carlyle: The Artist’s Assistant, London 2001, p.113.

 

35 RIBA, PB 344 [32] 10.

 

36 RIBA, PB 344 [32] 1; Fitzwilliam Museum, Cambridge, no.2209(2).

 

37 V. & A., no.D1232–1878.

 

38 Towndrow, op. cit. (note 29), p.138.

 

 

39 S. Waller: ‘Realist Quandaries: Posing Professional and Proprietary Models in the

1860s’, Art Bulletin 89/2 (2007), p.241.

 

40 S. Beattie: exh. cat. Alfred Stevens, London (Victoria and Albert Museum) 1975, p.38.

 

41 V. & A., no.D2067–1885.

 

42 Ashmolean Museum, Oxford, no.1962.17.18; British Museum, London, no.1905.0520.53.