Introduction

Nathaniel Armstrong Wells was the son of Nathaniel Wells, a Welsh country gentleman born in the West Indies and Britain’s first Black sheriff (of Monmouthshire), an exceptional figure in his time (Rainsbury 2020; Evans 2002; Draper 2013: 19; Smith 2013: 63). A comprehensive portrait of Nathaniel Armstrong Wells himself is difficult to form due to the limited biographical material available. Beyond the recording of his baptism in Wales in 1806, his marriage in February 1844 to Georgiana Lucy Price (d. 1889), a British national born in Saint Germain in Paris, and his death two years later in 1846, the only other mention of him is in 1807 in the diary of Joseph Farington, where he is described as ‘dark as his father’ (1924: 110). As a result, it is the book he published in 1846, The Picturesque Antiquities of Spain, that provides the most insight into his life. There are no drafts, diaries, or journals, and none of the ‘letters’ published in the book, or any other letters for that matter, have been located. I was able to track down thirty-four of Wells’s original drawings uncatalogued at the Tate Gallery, all of which are of an architectural theme, comprising thirteen views of Caen and one of Rouen in France, as well as twelve views of Seville, seven of Toledo, and one of Burgos in Spain. These drawings are complemented by twenty-nine published views: seven of Burgos, two of Madrid, seven of Seville, nine of Toledo, three of Valladolid, and one of Guadamur Castle, near Toledo. After deducting the eight Tate drawings that were also engraved and published, the total amounts to fifty-five views. Of the forty-one drawings of Spain, all but three depict examples of medieval architecture, with a clear emphasis on views of Seville and Toledo, which underscores Wells’s significant contribution to the historiography of the architecture of these cities.

Before analysing Wells’s Spanish sketches, it is worth considering whether the antecedents and inspiration for his Spanish sojourns may be located in France. It is assumed that Wells lived in Caen, a city associated with his wife and son. The 13 drawings of the city and one of Rouen at the Tate would appear to confirm that he spent time in Normandy, just as popular artists such as Samuel Prout, Richard Parkes Bonnington, and David Roberts had in previous decades. Wells would certainly have been familiar with their picturesque urban views, maybe even an engraving by Roberts, published in The Gallery of Modern British Artists in 1834, of the 13th-century church of St Pierre in Caen seen from the south, across a canal that bordered the church at that time (Figure 1). The low vantage point thrusts the Gothic pinnacles into the sky, adding sublimity to the scene as the medieval spire dematerialises into the fading light. Wells’s view of the same scene is from a higher vantage point, with dark pencil shading articulating the delicacy of the carved stonework while emphasising its solidity (Figure 2). It is the earthbound complement to Roberts’s ethereal vision.

Figure 1

Caen Cathedral, David Roberts, engraving. From The Gallery of Modern British Artists (1834: 24).

Figure 2

St Pierre, n.d. Nathaniel Armstrong Wells, graphite on paper, 11.3 × 18.0 cm. Tate, transferred from the National Gallery 1955, X84513.

A further French connection is evident in the fact that both journeys presented in the book begin in France. In 1844, before setting out on the tour of Castile, Wells may have been perusing the second volume of the Spanish artist Genaro Perez Villaamil’s España Artistica y Monumental, published in Paris that same year. This almost exclusively represents the same buildings in the same four cities as Wells’s Picturesque Antiquities of Spain. Villaamil’s three-volume lithographic survey of Spanish architecture was itself a response to Roberts’s Spanish engravings and lithographs of the 1830s. Roberts famously influenced Villaamil when they met in Seville in 1833 (Hopkins 2021), creating a thread that links the three men. Furthermore, Villaamil was sketching in Rouen in August 1842, making it possible to situate Wells, Roberts, and Villaamil sketching in both Rouen and Seville at different times.

A side-by-side comparison shows that Wells’s draughtsmanship is equal to, if not superior to, that of his more renowned peers. Initial assumptions that the degree of precision in Wells’s sketches, and curious deformations in others, stemmed from his use of optical devices gave way to a degree of uncertainty about where his ability to draw freehand ended and where his aided drawing began, if at all. At the same time, these drawings contain minor errors and distortions, which allow an appreciation of his natural ability as his errors originated in his use of vanishing points when setting up a perspective view. Overall, his drawings are reliable depictions of the architecture they represent.

Wells’s book is very much a product of its time. It combines a lively travelogue with observations on politics, local customs, and fine art, with the ubiquitous references of his time to Cervantes’ Don Quixote and Byron’s Childe Harold’s Pilgrimage (1846: 100, 441). What distinguishes it from those of his British peers is its focus on architectural descriptions of lesser known examples of the antiquities of four one-time capital cities of Spanish kingdoms, which display exemplars of the country’s diverse architectural heritage. This is evident in the book’s full title: The Picturesque Antiquities of Spain; Described in a Series of Letters, with Illustrations, Representing Moorish Palaces, Cathedrals, and Other Monuments of Art, Contained in the Cities of Burgos, Valladolid, Toledo and Seville. All but three of the buildings illustrated in the book date to between c. 1000 and c. 1500, i.e., from the Islamic architecture of Toledo to the florid Isabelino Gothic of Valladolid. Historiographically, this choice reflects the broadening interest in Britain in the diversity of Spanish medieval architecture, which had hitherto been overlooked, since much of it was considered inferior to purer expressions of Gothic or Islamic medieval architecture. This is evident in Henry Swinburne’s Travels through Spain in the Years 1775 and 1776 (1779), a key text in the dissemination of Spanish medieval architecture in Britain. To please the classical scholar, it is profusely illustrated with Spain’s Roman remains but is groundbreaking in its descriptions and engravings after Swinburne’s own drawings of the Nasrid Alhambra in Granada and the Umayyad mosque in Córdoba, drawings that are among the earliest representations of these edifices by a British amateur. The absence of depictions of Seville, and Swinburne’s description of the Alcázar as ‘a pasticcio of Saracenic, Conventual, and Grecian architecture’ (1779: 253) shows that the Alcázar was considered, in part, to be in imitation of the Alhambra. In fairness, Swinburne did admire the Alcázar’s principal façade, calling it ‘a piece of as good Morisco work as any I had yet seen’. Nevertheless, a dedicatory panel of Gothic script across the top makes clear to the visitor that it was constructed by and for Pedro I in 1356, over a hundred years after Fernando III took Seville.

The only other Spanish building deemed on a par with the Alhambra was Philip II’s austere Renaissance monastery of El Escorial (1562–84), a masterwork primarily by the hand of Juan de Herrera, stylistically very different from the Alhambra, but nonetheless similarly regarded as the apogee of an empire (Howarth 2007: 157, 171). Wells describes Herrera’s Lonja (1572), or Exchange, in Seville, now known as the Archivo de Indias, as ‘so perfect as to bid defiance to criticism. It might have been built by Vitruvius’ (1846: 435). It was only after the Peninsular War (1807–14), as Spain became more accessible, that Seville Cathedral began to be regarded as worthy of its peers among Europe’s great cathedrals. Until then, it had been criticised for the almost total absence of decoration in the vaulting, but that same lofty simplicity came to be praised for not diminishing the effect of the space on the visitor. This admiration is reflected in Roberts’s numerous depictions of it, but more so in Wells’s two interior views, in one of which his choice of perspective renders the cathedral almost unrecognisable in its austerity.

Precedents for Wells’s study of Seville and Toledo are few, and in British historiography even fewer. In the 18th century, Seville’s religious and secular architecture was rarely represented, often only illustrated from pre-existing engravings, such as William Jacob’s Travels in the South of Spain (1811), whose plates were predominantly copies of engravings by Pedro Tortolero acquired in Spain. The only British historian prior to Wells to write positively of Toledo was Samuel Widdrington-Cook, whose father had an English copy of Alexandre de Laborde’s Voyage historique et pittoresque en Espagne, published between 1806 and 1820, in his library. Widdrington-Cook recognised the city’s architectural heritage represented ‘almost every style, Roman, Moorish, middle age, Jewish, Gothic, Classic and Modern’ (Cook 1834: 144). Not only do Wells’s drawings address the earlier lack of attention, but they also reflect the growing British interest in Spanish architecture at that time. What spurred his interest is unknown, but in omitting Spain’s Islamic and Renaissance architecture to centre on its varied and complex medieval architecture, Wells’s book marks, visually at least, the beginning of the re-evaluation of Spanish medieval architecture, and introduces buildings of novel interest to the Victorian gentleman scholar.

French sources further illuminate the visual precedents for Wells’s work, particularly for the architecture of Valladolid, of which no engravings in English texts existed at the time (Azofra 2018). Early in the 19th century, Laborde published engravings of the four cities in Wells’s book as part of Voyage historique et pittoresque en Espagne. Notable examples are the engravings of the Isabelino façade of San Pablo and the cloister of the college of San Gregorio, both in Valladolid and also later depicted by Wells (Laborde 1806–1820: pls. 30, 31). Those who could afford them would have acquired Laborde’s engravings, so Wells could be seen to be popularising them in a more accessible format. Another significant visual source, contemporaneous with Wells’s travels, is the first volume of Villaamil’s España Artística y Monumental (1842). It includes two views of the cloister of the college of San Gregorio in Valladolid, which Wells also depicted (Villaamil 1842: 22, 24). Villaamil’s visual survey of Spain influenced the discourse on the nationalistic value of Spanish architecture, with his preference for buildings materially distinguished from central European examples by their yeserías (plasterwork), azulejos (wall tiles), sebka (interlacing brickwork), and artesonados (carpentry ceilings). These architectural and constructional elements, often associated with Islamic architecture, led Villaamil to focus disproportionately on Burgos, Seville, and Toledo across his three volumes. Wells ought not to be considered a copyist, as his visit to Seville predates the publication of the earliest of Villaamil’s work, and the gestation of Wells’s publication overlaps with his French and Spanish contemporaries.

What is more, Wells’s historical interpretations of buildings lean towards those of contemporary Spanish historians, particularly in his interpretation of Islamic and Gothic forms. His conclusions, however, were limited by the language of the period, notably the problematic use of the term ‘Moorish’. When describing the architecture of southern Spain, the term ‘Moorish’ was indiscriminately used for both the Islamic architecture of Granada and the Jewish or Christian architecture of Seville and Toledo. Despite this limitation, Wells excelled in describing the materiality and construction of Spanish medieval architecture, especially demonstrating uncommon sensitivity by including the craftsperson as integral to the architectural form and appearance of the style. In doing so, he elucidated his belief that the solidity of the buildings, complemented by decorative elements such as spoliated columns, yeserías, azulejos, sebka, and artesonados, surpassed all other architectural traditions. Theoretically, Wells synthesised rather than advanced contemporary thinking on medieval Spanish architecture, albeit doing so intuitively in his descriptions, which appear closer to those of contemporary Spanish historians. Interestingly, when describing the 13th-century tomb of Fernán Gudiel ‘Alguacil’ (1278) in Toledo cathedral, Wells used the term ‘Moorish’ to denote not just style but authorship. He wrote that it was sculpted by a ‘Moorish’ artisan (1846: 128–9) and was coeval with the Gothic structure, and did not predate it. This interpretation aligns with Amador de los Ríos’s description of it:

[A] niche of Arabic taste … The tomb is adorned with a horseshoe arch. … On the sides of the cornice, with which it terminates, are two animal figures, resembling bears, which reveal at once the Saracen origin of this work in a Christian temple. (1845: 87)

In contrast, the tomb is unrecognisable in Richard Ford’s description: ‘In [the chapel of] San Eugenio are some remains of the old mosque, with Cufic inscriptions, and an arch and tomb of elaborate Tarkish work’ (by ‘Tarkish work’ Ford means elaborate plasterwork) (Ford 1845: 844; italics in original). However, it is possible that he based his account on second-hand information rather than direct observation. All three descriptions of the tomb were published in 1845 or 1846, with Wells’s interpretation more closely mirroring that of Amador de los Ríos, who led contemporary thought on the history of Spanish architecture. Ford, incidentally, favourably reviewed Wells’s book in the Quarterly Review for his ability to introduce remarkable insights in an accessible literary style, and desired Wells revisit Spain on a tour of lesser-known places ([Ford] 1846–47).

Wells’s appreciation of yeserías and artesonados is apparent in his drawings of the Tránsito Synagogue (1357) in Toledo, which he considered the finest example of the medieval architecture associated with Andalusia outside the region and ‘the first ecclesiastical edifice of its style recorded as having set the example of an open area, destitute of columns and arcades’ (1846: 200). His drawing (Figure 3) notably omits the altarpiece presumed to have been removed after Villaamil’s earlier depiction, before 1840 (Figure 4), possibly due to the disentailment of church property that took place 1836–37, known as the Desamortización de Mendizábal. The combination of the elevated vantage point in Wells’s drawing, its slightly upwards inclination, and its forced perspective contribute to making the artesonado the focal point. At the time, these carpentry ceilings were considered the outstanding feature of medieval Spanish architecture. When visiting the Alcázar in Seville in 1842, Elizabeth Grosvenor considered the ceilings to be the most refined in Europe, ‘indeed all other ceilings, whether in France, England, or Italy, appear coarse and common in comparison’ (1842, vol. 1: 38). While Villaamil’s lithograph is the earliest publication of an artesonado, Wells’s drawing of the ceiling of the Cuarto del Príncipe in Seville predates it. Similarly, his drawing of the Tránsito Synagogue in Toledo was created slightly later. This overlap demonstrates that both artists were drawing artesonados in the same years. Wells’s view omits the fixtures, fittings, and apertures in the walls, with the exception of the azulejo-covered benches. This choice emphasises the airy, column-less hall he admired so much. Despite the overall quality of the drawing, and aside from the omission of later architectural features, the benches extend too far forward and should terminate in line with the first roof beam.

Figure 3

Interior, Toledo, 1844. Nathaniel Armstrong Wells, graphite on paper, 21.0 × 16.0 cm. Tate, transferred from the National Gallery 1955, X84492.

Figure 4

Sinagoga Mayor de Toledo, Genaro Perez Villaamil, lithograph (Villaamil 1842: 44).

Wells studied the upper frieze of yeserías in the synagogue in another drawing (Figure 5), which displays great skill in capturing the intricate depths of the plasterwork and incomparable refinement in the transcription of the Hebrew script, particularly when seen side by side with a contemporary engraving by José Amador de los Ríos (Figure 6).

Figure 5

Interior, Toledo (detail), 1844. Nathaniel Armstrong Wells (1806–1846), graphite on paper, 13.5 × 20.0 cm. Tate, transferred from the National Gallery 1955, X84493.

Figure 6

Ajimez de San Benito, José Amador de los Ríos (Amador de los Ríos 1845: 243).

When describing the tomb of Fernán Gudiel in Toledo Cathedral, Wells appreciated how the craftsperson’s skill was integral to the stylistic architectural expression. This critical insight is also evident in his descriptions of other Christian or Jewish buildings decorated in the Islamic tradition, such as the 14th-century palace of Pedro I (1356) at the Alcázar in Seville, although it is not so clearly articulated. It was not until 1859 that Amador de los Ríos appropriated the word ‘Mudéjar’ to describe the Alcázar in Seville, the Tránsito Synagogue, and other examples of medieval architecture that displayed constructional and decorative elements associated with Islamic architecture. The term ‘Mudéjares’ had previously been used to describe Muslim converts who remained in the new Christian kingdoms, and Amador de los Ríos argued that their construction skills and craftsmanship imbued Spanish medieval architecture with its Islamic appearance across various regions. Amador de los Rios’s radical contribution to Spanish architectural discourse was to frame the term ‘Mudéjar’ as the result of the way builders, joiners and craftspeople employed brick, plaster, wood and ceramic in the construction of a building, as opposed to the visual signifiers of an architectural style as such. This skill-based argument formed a coherent narrative identity for medieval Spanish architecture that was flexible enough to include regional variations. Notwithstanding its immediate acceptance, its accuracy and usefulness were contested at the time by Pedro de Madrazo, who argued that omitting to define its stylistic qualities was problematic. He also emphasised that the craftsmen who were supposed to have given rise to the style were not all Muslim converts but included men whose antecedents reached back to the earlier pre-Islamic Christian populace called Mozarabes, who had simply adapted their skills to the new Islamic style (Amador de los Ríos and de Madrazo 1859). There are nationalistic undercurrents to both historians’ arguments.

The term ‘Mudéjar’ has been recently re-evaluated by Juan Carlos Ruiz Souza (2010), who gives examples of many medieval buildings that are commonly classified as Mudéjar despite having no tangible Islamic stylistic influence. The term is also used for stone buildings in the Spanish colonies (i.e., after 1492) that continue to replicate many of the same stylistic elements, so, as a signifier the term has become all-encompassing and, for some historians, redundant, since it was a 19th-century construct. The argument about Mudéjar is a complex one, outside the scope of this article, but what is relevant here is Wells’s acknowledgement of the contribution of the craftsperson to the stylistic appearance of the building.

Seville Cathedral

Wells’s four drawings of Seville Cathedral are among his most instantly identifiable subjects, as three of them resemble views published by his contemporaries. Wells likely had prior knowledge of the Giralda (Figure 7) before he came to Seville, if not through the famous oil painting by David Roberts, created in Seville in 1833 and exhibited in London, then perhaps through an almost identical engraving by John Frederick Lewis (1804–1876) (Figure 8) in a work on the poet Byron (Finden 1834).

Figure 7

Giralda, Seville, c. 1841, Nathaniel Armstrong Wells (1806–1846), graphite on paper, 19.3 × 12.8 cm. Tate, transferred from the National Gallery 1955, X84485.

Figure 8

Seville, The Giralda, John Frederick Lewis, engraving (Finden 1834: pl. 8).

Wells’s drawing of the Giralda is beautiful and airy, with the surface texture of the sebka panels that decorate the 12th-century Islamic body seemingly picked out from the sky — only the shadows in the windows and Renaissance belfry give it depth. There is something unavoidably awkward in drawing a tower, made more difficult here by Wells’s proximity and eye-level point of view. While perspective should make the tower naturally taper towards the top, most artists draw the body of the tower in elevation, with proportions close to those of a measured drawing with no foreshortening or distortion. How then does the artist resolve the transition from the body in elevation and the marked perspective in the belfry? Handling this is never easy, and Wells has drawn the perspective of the cornice receding on the sides of the belfry, which causes the illusion that the belfry is leaning outwards towards the viewer. Lewis avoids this issue by drawing the Giralda from a slightly higher point of view, with straight sides from its base to the urns that crown it, omitting the projecting cornices in profile.

There are many interesting details in the construction of this drawing. To the right, facing the tower, is the outer wall of the Patio de los Naranjos, the ancient sahn of the mosque, which houses the Biblioteca Colombina, and facing it on the left is the side elevation of the Archbishop’s palace. The wall of the Patio de los Naranjos has a vanishing point at eye level in the centre of the Giralda, while the vanishing point of the Archbishop’s palace lies farther to the right. This indicates that Wells was aware that these buildings are not parallel but converge slightly toward the tower. The perspective line along the base of the Archbishop’s palace wall is higher than the others, as the ground rises ever so slightly towards the corner of the palace. As this is scarcely perceptible to the naked eye, it raised the possibility that Wells had made use of a camera obscura or camera lucida, but this can be discounted in looking at his errors. The most noticeable error is the angle of the window sills of the palace, which does not correspond with the perspective of the drawing as they ought to appear horizontal since they are near eye-level. This drawing, therefore, is a carefully studied composition that combines very subtle responses to almost imperceptible topography and urban space yet conceals a simple error in the position of the vanishing point of the windows.

The iconic view of the Giralda from the Patio de los Naranjos (Figure 9) allows viewers to compare Islamic, Gothic, and Renaissance architecture within a single composition. It was drawn and widely reproduced by many artists, including Roberts (1836: 121), and Philibert-Joseph Girault de Prangey (1836–39: pl. 1) (Figure 10), as well as five other European artists. What is difficult to explain in Wells’s view is the clear vertical curvature in the body of the tower. At first glance, this appeared to result from drawing the Giralda through a lens, but on closer inspection, these parallel deformations do not correspond to the converging or diverging distortions that result from using a camera obscura (Fiorentini 2008: 212). Girault de Prangey was also suspected of employing optical devices in the construction of his drawings, but the results were inconclusive. Like Wells, some but not all of Girault de Prangey’s drawings show underlying construction lines (Caumont 2019: 177–78). It is usually the delicacy of Wells’s drawing that is admired, but here, with singular confidence and a thick pencil, he has emphasised the solidity of the roughhewn stonework of the north door, unfinished after 300 years, when many other artists hid this unattractive stonework behind a judiciously positioned tree.

Figure 9

Giralda (or Bell Tower), Cathedral of Seville, c. 1841, Nathaniel Armstrong Wells (1806–1846), graphite on paper, 19.3 × 12.8 cm. Tate, transferred from the National Gallery 1955, X8448.

Figure 10

La tour de la Giralda, Girault de Prangey (1804–1892), lithograph (Girault de Prangey 1836–1839: pl. 1).

Wells’s interior views of the cathedral are more interesting as they are less familiar and more personal and revealing. In his writing, he contrasts the more florid French Gothic of Burgos Cathedral with the sombre austerity of Seville: Burgos ‘is elegant, but deficient in grandeur; beautiful, but wanting in majesty. The stern and grand simplicity of the one [Seville], thrown into the scales against the light, airy, and diminutive, though graceful beauty of the other [Burgos]’ (1846: 50–51). William Stirling-Maxwell, writer of the Annals of the Artists of Spain (1848), described his impression of the interior of Seville Cathedral to his sister:

I got accustomed to the dim religious light that filled [the cathedral]. The immense height of the groined vaults above — the breadth of the five great aisles which retreat further than the eye could reach within the forest of pillars — the number & skilful disposition of the painted windows the splendour of the high altar surrounded by screens of wrought iron of the finest design and completely gilt through the bars of which the morning sun comes streaming from the opposite windows — would … have been worthy of all admiration — but together were absolutely overpowering. (Brigstock 2015: 354)

Wells’s interior of Seville Cathedral that was later published (Figure 11) exaggerates its majestic grandeur atmospherically in a mysterious penumbra (1846: 352). Where earlier artists like Roberts had depicted the central nave looking towards the high altar (Roberts 1838: 34), the novelty in Wells’s view lies in looking away from the altar (hidden behind the wrought-iron screen on the left) and instead across the interior towards the crossing, simultaneously drawing attention upwards to the decorative ribbed vaults at the cathedral’s heart. The result is a view that is almost a collage of architectural elements that is difficult to decipher spatially but evokes the splendour of Stirling-Maxwell’s description. His drawing predates the publication of a similar view by Villaamil in the second volume of España Artística y Monumental of 1844 (Figure 12), which is taken from a similar position, but its stronger perspective looking down an aisle ‘further than the eye could reach’ is spatially more coherent but makes a less interesting image overall.

Figure 11

Interior of Seville Cathedral, c. 1841, Nathaniel Armstrong Wells, graphite on paper, 21.0 × 13.5 cm. Tate, transferred from the National Gallery 1955, X84482.

Figure 12

Interior de la nave lateral de la Catedral de Sevilla, Genaro Perez Villaamil, lithograph (Villaamil 1844: 60).

The stern simplicity of Seville Cathedral is better represented in another view of the interior (Figure 13), in which the cathedral is barely recognisable. Wells chose a point of view in the south aisle, south of the choir, looking northwest through the forest of pillars. This composition lacks an obvious focal point, but it emphasises the effect of the colossal piers within the diaphanous space. The vaulting in this western end of the church is unadorned, the columns are simple yet majestic, and the quality of light is diffused and soft. The centre point of the arches of these side aisles rises to 26 metres above the floor, and the perspective of the cornice of the pillars, which appears as little more than a band around the colossal piers, conforms to two vanishing points, external to the drawing, elevated above eye level, approximately the height of the oculus on the distant wall. However, the vanishing points of the bases of the pillars are not the same as those of the cornice, but lower. This deliberate construction reduces, or effectively eliminates, the expected perspective and increases the sense of the volume of the space.

Figure 13

Interior of Seville Cathedral, c. 1841, Nathaniel Armstrong Wells, graphite on paper, 21.0 × 13.5 cm. Tate, transferred from the National Gallery 1955, X84483.

The Façade of the Palace of Pedro I

Wells’s view of the Patio de la Monteria is framed by an arch cut through the wall of the 11th-century Islamic Alcázar fortress (Figure 14). The monumental façade of the palace of Pedro I is partially hidden to the left, while a plain wall directly ahead makes an odd focal point. In the seven drawings of the palace, Wells in effect takes the viewer on a guided tour of its lower floor that terminates in the Cuarto del Principe, situated behind this blank wall. Wells omitted a window that had been cut through the outer wall of the palace, as detailed on the 1759 plan of the Alcázar by Sebastián de Van Der Borcht (AGP plan No. 4,581). This window also appears in Girault de Prangey’s 1839 view (drawn in 1833) (Figure 15), and Edward King Tenison’s photograph (1852: 12). Wells’s omission may have been influenced by his understanding of the inward-looking arrangement of the palace when it was originally built. His intention seems to have been to present the 14th-century façade as authentically as he imagined it, without later alterations.

Figure 14

Patio del Alcázar, Seville, Court of the Palace, c. 1841, Nathaniel Armstrong Wells, graphite on paper, 21.0 × 13.3 cm. Tate, transferred from the National Gallery 1955, X84481.

Figure 15

Façade de l’Alcazar, Girault de Prangey, lithograph (Girault de Prangey 1836–39: pl. 3).

The view is taken at an angle similar to that of Girault de Prangey but from a greater distance, thus revealing more of the roof layout. This perspective supports Antonio Almagro Gorbea’s hypothesis that there were three pyramidal roofs corresponding to three rooms on the first floor of the palace, whose artesonados were never recorded. He based this on their appearance in an engraving by Louis Meunier in 1668, held at the Biblioteca Nacional de España in Madrid, as well as archaeological evidence obtained from inspecting the supporting walls (González Chávez 2004: 91; Almagro Gorbea 2015: 98–99). These three roofs also appear in Wells’s published view of 1846, as he had drawn them in 1841. By 1846, the roofs had been rebuilt, raised higher, and modified with pitched roofs replacing the two flanking pyramids. The supporting walls were decorated with brickwork to complement the architecture of the façade, contributing to the palace’s orientalisation (González Chávez 2004: 70; Almagro Gorbea 2015: 95, 97). In the view of the façade from under the arcade on the right (Figure 16), Wells included the outline of the frame of one of the windows omitted in the previously described view. This window, like the roof, was also orientalised with an ajimez (bifora), clearly depicted in an undated view (MLG, ref. 06304) (Figure 17).

Figure 16

Patio (Court) del Alcázar, Seville, c. 1841, Nathaniel Armstrong Wells, graphite on paper, 13.5 × 19.5 cm. Tate, transferred from the National Gallery 1955, X84489.

Figure 17

Puerta de los Reyes y patio del Alcázar, c. 1834–66, unknown artist, pencil on paper, 62.4 × 40.5 cm. Museo Lázaro Galdiano. ref. 06304.

It is difficult to measure the impact of Wells’s book, but his description of the façade as ‘the Arab ornamental portion’ (1846: 325–6, 343) seems to be echoed by William Edward Baxter (1825–1890), who came to Seville in 1850 and described the façade as ‘a truly Arab portal, ornamented with rich tracery’ (Baxter 1852: 88) — this is an error, as woven through this tracery are the symbols of Castile, but with both writers describing the façade as such, it may be one instance of Picturesque Antiquities of Spain influencing later travellers.

Patio de las Doncellas

Having passed through the façade of the palace and along a hallway cut through the medieval fabric of the palace, the 19th-century visitor entered the Patio de Doncellas midway along its longer elevation through a ‘magnificent embroidered arch’ (1846: 326) (Figure 18). By standing at a point just before entering the patio, Wells was able to include the decoration on the inner surface of the arch in order to develop the theme of structural solidity and surface decoration. When this drawing is compared to the published version (Figure 19) that was engraved by William Francis Starling (1805–1878), his finesse with a pencil becomes more apparent. In fact, the quality of the engravings was criticised at the time. The Spectator judged that Starling

has executed [the engravings] in a coarse, flimsy, unartistlike manner, as though cheapness rather than excellence had been aimed at. They are a combination of etching, aquatint, and machine-work — the commonest means at the engraver’s disposal. The extreme beauty of the original drawings (judging of them from the wood-cuts) merited a better treatment. (‘Wells’s Antiquities of Spain’ 1846)

Figure 18

Patio, Alcázar, Seville, c. 1841, Nathaniel Armstrong Wells, graphite on paper, 21.0 × 13.3 cm. Tate, transferred from the National Gallery 1955, X84480.

Figure 19

Great Court of the Alcázar, Seville, Nathaniel Armstrong Wells, engraving (Wells 1846: 326).

In the published view, figures standing in the patio were introduced and dressed in robes of an imaginary bygone time, making this image one of the earliest Orientalist representations of the palace. Until now, the people included in a view were invariably dressed in contemporary if sometimes picturesque fashion.

Arco de los Pavones

In his view of the Salón del Techo de Felipe II, Wells combines a beautifully detailed drawing of the yeserías that incorporate birds of regal and eternal significance with a vista that looks through the Salón de los Embajadores to the Patio de las Doncellas, from where the visitor has just come (Figure 20) (1846: 328). The deeply coffered ceiling continues the tradition of artesonados but in a markedly different language; as the name of the room implies, it is Hapsburg rather than Trastámara and dates to around 1589. The third row of coffers runs along the centre of the ceiling, approximately six metres above the artist’s head, something unapparent in the careful construction of the drawing and impossible to see when looking through the arches towards the patio.

Figure 20

Pillars of Black Marble, Alcázar, c. 1841, Nathaniel Armstrong Wells, graphite on paper, 21.0 × 13.5 cm. Tate, transferred from the National Gallery 1955, X84490.

There are some slight differences between the yeserías in the picture and those present today, notably the scale of the cornice of muqarnas (projecting tessellated brackets) and the frame around the arch. The current cornice is the same as that in the Patio de las Muñecas, which was installed in the 1840s. While that mould may have been taken from here, it is also plausible that when the coffered ceiling in this room underwent restoration in 1845, a mould from elsewhere was used to make a new cornice here as well as at the other patio. This drawing raises questions about the extent of the restoration work carried out in the 1840s, and it may depict elements of the original plasterwork (Chávez González 2004: 100, 127; Cómez 1996: 97).

Patio de las Muñecas

For Wells, the beauty of the Patio de las Muñecas stemmed from the contrasting hues of the spoliated marble columns and the irregular carving of their capitals:

A space, of about twenty feet by thirty, in which ten small pillars, placed at corresponding but unequal distances, enclose a smaller quadrangle, and support, over a series of different sized arches, the upper walls, has furnished materials to the artist for the attainment of one of the most successful results in architecture. The Alhambra has nothing equal to it. (Wells 1846: 330)

His enthusiasm for it can be read in contrast to the restrained neoclassicism of Piercefield House (1792), an early work by John Soane that was Wells’s childhood home. His view of the patio (Figure 21) is centred on a column of dark marble and its ‘ill-proportioned capital’ that enchanted him so much (1846: 331). Its oblique sightline crosses through two doorways, purposefully contrasting the structural massiveness of the building with the delicacy of its surface decoration. Wells presumed the columns and capitals had been spoliated from the Roman city of Itálica, nine kilometres north of Seville, but the capitals are actually Islamic, from different periods, with some possibly from Córdoba (Bermúdez Cano 2004). The spatial dynamism of Wells’s view is markedly different to one by Girault de Prangey (Figure 22), who stood about a metre in front and looked directly across the patio towards the Salon de Embajadores. As a result, Girault de Prangey’s view recedes in planes like a stage set, and any sense of solidity is lost.

Figure 21

Court of Dolls, Alcázar, Seville, c. 1841, Nathaniel Armstrong Wells, graphite, watercolour and ink on paper 27.8 × 20.5 cm. Tate, transferred from the National Gallery 1955, X84486.

Figure 22

Patio de las Muñecas, Joseph-Philibert Girault de Prangey, lithograph (Girault de Prangey 1839: pl. 5).

Wells’s other view of the same patio (Figure 23) allows an invaluable glimpse of the appearance of the upper-floor gallery before its restoration, or transformation, in 1843 (Figure 24), and resolves the doubts of many historians and architects (Almagro Gorbea 2015: 109, 110; Chávez González 2004: 60, 63, 73). The original medieval courtyard consisted only of the lower-floor gallery around the patio, and the 10 marble pillars supported a low roof. The first floor in this part of the palace dates to around 1500, and its appearance can be reconstructed from Wells’s drawing, his written description (1846: 334), and the archival research of María del Rosario Chávez González. The first-floor gallery consisted of 10 wooden columns with Corinthian capitals, positioned directly above those of the patio below, connected by a wrought-iron railing and wooden balustrade. This upper gallery was also covered with a carpentry ceiling, whose geometries should be imagined as robustly Renaissance rather than gracefully medieval. The smooth surface above the 14th-century arches and below the later gallery is an entresol drawn with windows in a view by John Gardner Wilkinson (Sorowka 2022), which Wells has omitted, just as he omitted those in the Patio de la Monteria.

Figure 23

Alcázar, Seville, c. 1841, Nathaniel Armstrong Wells, graphite on paper, 13.5 × 19.5 cm. Tate, transferred from the National Gallery 1955, X84488.

Figure 24

Patio de las Muñecas, photograph taken by the author in 2021. Everything above the ground-floor arch directly ahead is an 1843 invention, as is the ajimez seen through the door.

As the drawing is constructed as a one-point perspective with a vanishing point directly ahead, the columns should recede in four corresponding planes, but they do not. The column in the foreground, in the middle of the picture, is on the same plane as the column on the left that appears further away, which should not in fact be visible. It is only visible due to the unnaturally wide cone of vision.

Cuarto del Príncipe

The viewer has been escorted by Wells through the principal spaces of the lower floor of the 14th-century palace, with the exception of the Salon de los Embajadores. The private chamber called the Cuarto del Príncipe is located between the interior Patio de las Muñecas and the exterior Patio de la Montaría. In his view (Figure 25), light from an unseen window in the opposite alcove falls from one of the undrawn windows in his depiction of the main façade (Figures 8, 9). By looking up at the three-dimensional quality of the artesonados, the aesthetic is fully realised, and this is the earliest drawing of a room in the palace that includes its medieval carpentry ceiling. Its novelty makes it regrettable that the publisher of Wells’s book, Richard Bentley, did not choose to include it — their contract stipulated that the choice of illustrations rested with Bentley rather than Wells (BL Add. MS. 46614).

Figure 25

Galería del Alcázar, Seville, c. 1841, Nathaniel Armstrong Wells, graphite on paper, 21.0 × 13.5 cm. Tate, transferred from the National Gallery 1955, X84491.

Wells wrote that ‘this room has been selected in modern times, as being the best in the palace, for the experiment of restoring the ceiling’ (1846: 332–3), which had been carried out in 1834. While it was true that it was an experiment, this ceiling was chosen for the simple reason that it was in such a neglected state that without immediate intervention, it may have been permanently lost. In light of this, when Wells described it as the most strikingly beautiful ceiling, it was certainly the most complete at the time: its geometries stood out more dramatically and were more easily decipherable thanks to its recent repainting.

Although it may appear as if the viewer is outside the space looking in, the layout of the room is symmetrical along both axes; Wells positioned himself in the corner of one alcove, looking diagonally across the space towards the opposite corner in the opposite alcove. Of all his drawings, this one adheres most closely to a single-point perspective, as all lines meet at a single point at eye level. However, as Wells draws with such an exceptionally wide cone of vision, it inevitably leads to distortion, as can be seen in the curvature of the foreground arch. This room does not display the customary azulejos, but Wells was very clear that this was the only room in the palace not decorated with them. The ones admired there today were not installed until 1894 (Pleguezuelo 2015: 229).

Two Fountains in the Gardens of the Alcázar

For the 19th-century traveller, the geometric formality of the gardens of the Alcázar stood in contrast to the English picturesque landscape of which Wells’s childhood home was an outstanding example. Elizabeth Vassall-Fox (c. 1771–1845) was delighted by the Alcázar’s enclosed gardens when she visited them in 1803, and remarked how fatiguing English landscapes were in comparison (Holland 1910: 58). In her early life, Vassall-Fox received an annual income of seven thousand pounds from her family’s sugar plantations in Jamaica, where she may have been born. This connection establishes another link between the legacy of slavery in the West Indies and British people in Spain.

The fountains drawn by Wells are situated in separate gardens to the south of the 14th-century palace (Figure 26). Their medieval origins are uncertain, but it is known that their current layout was established in the 16th century, with architectural interventions from the late 16th and early 17th century shaping their present appearance (Plaza 2015: 57–58; Marín Fidalgo 2015: 107–8). The fountain on the left, in the Jardin de Troya, features a scalloped basin of Islamic origin. It would not have been raised on a pedestal originally, but was placed so when installed in the garden in the 18th century (Manzano 2013: 192). The fountain on the right is in the adjacent Jardin de la Danza and dates to the end of the 16th century. The azulejos are finely depicted but have not survived to the present. They merit further investigation to determine whether they conform to designs from the time of the garden’s inception or from a later restoration (Riego Ruiz 2017: 112).

Figure 26

Fountains at the Alcázar, Seville, c. 1841, Nathaniel Armstrong Wells, graphite and ink on paper, 9.1 × 18.5 cm. Tate, transferred from the National Gallery 1955, X84480.

Conclusion

Wells’s untimely death in 1846, the year his book was published, cut short any further contribution to the architectural history of Spain that Ford had hoped for. Had he lived and continued his scholarly appraisal of Spain’s lesser-known buildings and styles, how might Wells have responded to Amador de los Rios’s clear articulation of the role of the medieval craftsperson in the development of a national style one decade later? He might have considered that Amador de los Rios had substantiated ideas that he himself had been contemplating for many years. Wells had intuited an architecture that was considered sufficiently outside the recognised Islamic or Gothic styles not to merit close study, and his insightful book redressed that oversight. If we had only his book, it would not be possible to grasp his thoughts so clearly; it is his unpublished drawings that allow a fuller interpretation of them. What transpires from these drawings is his disinterest in British writing on the subject, and his consonance with that of Laborde, Girault de Prangey, Villaamil, and Amador de los Rios. In this light, Wells can be seen as presenting new imagery into English-language architectural historiographical discourse in response to earlier French publications, the architectural tastes of Spanish artists, and the contemporary theories of Spanish historians.

In this limited survey of Wells’s work, only a few of his drawings resemble the paradigmatic views Seville or Toledo of his time. The majority reflect his profound interest in architecture; they are unique and unmatched documents that have withstood the test of time. His artistic ability is exceptional when compared to his contemporaries, and his oeuvre is characterised by an awareness of refined spatial relationships combined with a sensitivity to the solidity and superficial decoration so admirable in Andalusian medieval architecture. Although there are errors in the construction of some drawings, they do not detract to any significant degree. Such errors debunk the initial conjecture that he had used an optical device, and affirm his natural talent. A criticism of his work could be his exclusion of windows, which he may have assumed were later insertions, or the ‘missing’ altar in the Tránsito synagogue. These omissions introduce an element of uncertainty into some of his drawings, but further research could help resolve these ambiguities. As historical documents, Wells’s drawings are astutely transcribed from life and are, for the most part, accurate. They illuminate hitherto unseen aspects of the architecture of Seville and Toledo and serve as a reliable, truthful resource that suggests new directions for investigation and future restoration projects.

Competing Interests

The author has no competing interests to declare.