Some underdrawing is visible to the naked eye in places where the paint has become more transparent over time, but more of an extensive underdrawing executed in a dry material is revealed by infrared reflectography (fig. 2). The drawing material has been identified using SEM‐EDX as an alloy of lead and tin, and is therefore metalpoint of a similar composition to that Raphael employed for the underdrawing of the Madonna of the Pinks (NG6596); in the sample, from the unpainted edge just below Saint John, the drawing is clearly on top of the priming (fig. 3).

The infrared reflectogram also reveals ruled lines that bisect the composition horizontally and vertically. Similar ruled lines are found in four of the smaller paintings by Raphael in the National Gallery (Saint Catherine of Alexandria [NG168], An Allegory (‘Vision of a Knight’) [NG213], Procession to Calvary [NG2919] and Saint John the Baptist preaching [NG6480]), and also in the Small Cowper Madonna (about 1505–6) in the National Gallery of Art, Washington, DC.⁷

Most clearly visible in the infrared images is detailed underdrawing for the figures, which includes locks of hair and individual drapery folds,⁸ confidently executed with very few revisions and closely followed in the painting. The concentric curves that describe the Christ Child’s forehead, temple, eyebrows, nose and knuckles are also found in other drawings and underdrawings by the artist. Delicate parallel hatching in the underdrawing indicates shadows, for example around the shoulder of the Baptist. However, new infrared reflectograms have revealed that this underdrawing was the second attempt at placing the figures on the panel. A fainter underdrawing can be made out, with the Virgin, Christ Child and upper part of the Baptist posed exactly as they were when finally painted, but positioned slightly to the right of their final positions (fig. 2 and fig. 4). The lower part of the Baptist’s body in this first underdrawing was, however, different. It is difficult to make out exactly what was planned, but underdrawing for a toe at the bottom of the painting near the middle shows that Raphael first tried to have the Baptist sitting with at least one leg horizontal. As discussed under ‘Reputation’ below, it seems highly likely that there was a Raphael Virgin and Child with Saint John that resembled the Garvagh Madonna but introduced the changes which are visible in the engraving of 1642 signed ‘G.M.V.’ (fig. 15), in which the young Saint John the Baptist is seated. There are some parallels with the pen and ink verso of a drawing in Lille (454), which could suggest the Virgin Mary sitting in a position akin to that in the Alba Madonna, but on balance it seems more likely that we are looking at a variation for the figure of Saint John the Baptist. A seated or kneeling position was also considered in Lille 437. It is possible to show that the second underdrawing for the figures (apart from the Baptist’s legs) exactly matches the first underdrawing but shifted to the left, suggesting that a cartoon was used to ensure that the successful parts of the composition remained unchanged, although there is no evidence of any pouncing or tracing. In the first position, the Virgin Mary was not set against a pier of masonry as she is now, but instead against a continuous landscape. The underdrawing for the landscape was quite detailed and included rough shapes indicating bushes and clouds, as well as the hills and buildings, which were varied during execution (unlike the figure group where the drawing was followed closely). These changes are analysed under ‘Subject and description’ below.

Straight lines for features such as the Baptist’s cross, the edges of the parapet, the column and the bench were incised using a straight edge after the priming had been applied and are visible as dark lines in the X‐radiograph. The windows in the building in the distant landscape at the left were depicted by incising into the wet paint.

Most scholars agree that the Garvagh Madonna, as it came to be known,¹⁴ was painted by Raphael in the years immediately following his arrival in Rome. Its colour and figure types can be compared to the frescoes in the Stanza della Segnatura (1508/9–11), and this range of dates is also supported by analysis of related preparatory drawings (see ‘Drawings’ below) that once formed part of the so‐called ‘pink sketchbook’ that Raphael used in exactly these years and in which one finds drawings associated with the School of Athens as well as with the ceiling roundel that depicts Poetry (fig. 6).¹⁵ The position of the Christ Child’s legs in the Garvagh Madonna are almost identical to those of one of the putti who accompanies Poetry. All these connections led Carol Plazzotta to propose a date of about 1509–10.

Jürg Meyer zur Capellen dated the painting to about 1510.¹⁶ Konrad Oberhuber insisted with almost comical confidence in his clairvoyance on a slightly later dating, ‘senza dubbio nell’estate del 1511’ (‘without doubt in the summer of 1511’), and observed stylistic parallels with the face of Temperance in the fresco of the Cardinal Virtues on the Justice wall of the Stanza della Segnatura (fig. 7), and with the putto with a torch who stands beside Temperance on the balustrade in that fresco.¹⁷

An even later dating has been advanced in recent years by Silvia Ginzburg.¹⁸ Based on an analysis of one of the related drawings in the Musée des Beaux‐Arts, Lille (fig. 8), where she highlights the relationships with slightly later paintings, as well as a possible connection with one of the ignudi painted by Michelangelo on the Sistine Chapel ceiling in 1511–12, she argues for a date in or after late 1512.¹⁹ She detects similarities with the use of colour and the formal preferences in Raphael’s frescoes in the Stanza di Eliodoro, and tentatively links the picture with a payment to Raphael of 20 ducats in May 1514 by an agent of Alfonso d’Este I (1476–1534), Duke of Ferrara, a connection which coincides neatly with arguments for the picture’s early provenance; for this see ‘Previous owners and provenance’ and ‘Reputation’ below.²⁰ The picture could not, however, date from as late as 1514, and even pushing it into 1512 is to credit some of the available evidence over other more persuasive considerations in the dating of the picture.²¹ A discussion of these now follows.

In addition to the links with other works suggested by study of the pink sketchbook, it is instructive to compare the picture with its most closely related panel paintings, such as the Alba Madonna, which was originally on panel but has now been transferred to canvas (fig. 9). This painting, which was cleaned in 2003–4 prior to being shown with the Garvagh Madonna in 2004–5, is the only other picture in which one again finds both a blue and green turban, threaded with gold, and a similarly sensuous rendering of fur in the Baptist’s camel‐skin garment. The Alba Madonna also shares the rich pastel palette and the hazy depth of the landscape. The beautiful colour balance is very subtly achieved, with the blues and pinks of the Virgin’s drapery being picked up in the landscape. The quality of the light, and the way it is seen to fall, is extraordinary and develops the increasingly refined approach to depicting atmosphere first encountered in the Saint Catherine (NG168). The pose of the child is also very clearly related to the figure of Christ in the Garvagh Madonna. The idealised beauty of the Virgin Mary in the Alba Madonna can be compared with the figure of Sappho in the Parnassus in the Stanza della Segnatura, and although earlier dates have been proposed, these aspects of the Alba Madonna point to a date of about 1509–11.

On balance, a dating for the Garvagh Madonna in the period around 1509–11 seems clear, but this writer is more comfortable with a dating during the later phase of painting the Stanza della Segnatura, and so in the years 1510–11.²² It should be acknowledged that this can become a circular argument as none of the dates of these comparative works are secure. There is, however, a very pronounced stylistic divide between the later works executed in Florence and a series of responses to Raphael’s move to Rome that might suggest the Garvagh Madonna was not executed immediately upon arrival, but after he had absorbed some of these new experiences (for this, see ‘Style’ below).

Despite being datable to Raphael’s early Roman years, the Garvagh Madonna remains deeply indebted to the artist’s experience in Florence, and to his investigation of the theme of the Virgin and Child in the years 1507–8. As discussed under NG6596 (the Madonna of the Pinks), Raphael demonstrates his awareness of Leonardo da Vinci’s Benois Madonna (about 1481–2; State Hermitage Museum, St Petersburg; see NG6596, fig. 7) and he further developed the idea of a central carnation in this picture, where it becomes the linking motif between the three figures. The paired windows behind the Virgin also recall Leonardo’s Madonna of the Carnation (about 1475; Alte Pinakothek, Munich), while David Ekserdjian has suggested that the pose of the Christ Child also originated with Leonardo, pointing to its similarity to the Christ Child in the Madonna of the Violets (about 1498–1500; current location unknown) by Marco d’Oggiono, formerly in the De Navarro collection, which follows a lost composition by Leonardo that was, perhaps, also known to Raphael.²⁶

Raphael had experimented in Florence with pyramidal figure arrangements set in a landscape – the Madonna of the Meadow (1506; Kunsthistorisches Museum, Vienna) is the most successful example – and this spatial solution also originated with Leonardo. It was only here in the Garvagh Madonna, painted a few years later, that Raphael tried placing a compositional pyramid of this type in an interior setting ²⁷ The only non‐landscape composition by Raphael that anticipates this group is his half‐length allegory of Charity in the Baglioni Entombment predella of 1507 (fig. 5). Here, in the Garvagh Madonna, the curious way in which the Virgin sits on the bench helped Raphael to construct his picture. Her complex pose, with one leg underneath her, makes the grouping somehow more informal, and even sets up resonances with the Madonna of Humility type (where the Virgin sits on a rock or on the ground).²⁸ It connects the picture with Raphael’s parallel interest in Michelangelo’s Doni Tondo (1505–6; Le Gallerie degli Uffizi, Florence), a picture that he also referred to when designing the Baglioni Entombment (Galleria Borghese, Rome). The surface geometry of the picture – which we know to have been carefully planned and based on a central plumb line – is then punctuated by lyrical curving rhythms in which the graceful oval of the Virgin’s head is echoed in the S‐shaped bend of her neck, and in the stem of the carnation. As Plazzotta has observed, the Garvagh Madonna ‘represents the height of formal purity’ – quite distinct from its starting point in either Leonardo or Michelangelo – and the ideal that Raphael created fascinated later painters such as Sassoferrato (Giovanni Battista Salvi called Il Sassoferrato; 1609–1685) and Jean‐Auguste‐Dominique Ingres (1780–1867).²⁹

The less saturated palette of the Garvagh Madonna compared to earlier works has already been mentioned, but it should not be passed over as though it were an obvious or inevitable development. From Friedrich Wilhelm Basilius von Ramdohr (1757–1822) onwards it has been related to Raphael’s contemporaneous involvement in painting frescoes in the Stanza della Segnatura where a comparable ‘scale of tints’ was deployed.³⁰ The comparison was most fully developed by Konrad Oberhuber, while Sylvia Ferino Pagden described the landscape as ‘veneto‐giorgionesco’.³¹ This analysis of Raphael’s style is usually applied to his portraiture from about 1511 and linked to his contact with Sebastiano Luciani (later called del Piombo, about 1485–1547), who arrived in Rome in August 1511. On balance, however, the soft lighting of the landscape seems more reminiscent of Raphael’s Florentine experience and the direction of travel in his art that was stimulated by his interest in the landscapes depicted by Fra Bartolommeo (1472?–1517), combined with witnessing the early morning light in and around Rome.

As discussed under ‘Dating’ above, the Garvagh Madonna is universally accepted as a Roman work and it also shows the artist’s responses to his arrival in Urbe. The pink sketchbook contains studies after antique sculpture, which may have a link to one small but crucial detail in the composition that requires special explanation. The way that the Virgin holds a bunch of her drapery in her right hand may have developed from Raphael’s interest during this period in the Virgin lifting a veil over the sleeping child, but is suggestive here of looking at Roman coins or statues. Raphael may even have had a specific model in mind: the so‐called Mazarin Venus (Getty Museum, Los Angeles), which was apparently discovered in Rome about 1509–10. The statue was copied (fig. 10) in an engraving by Giovanni Antonio da Brescia (active 1490–1519); it was also recorded in drawings by Raphael’s workshop and may have entered the collection of Raphael’s important patron, Agostino Chigi.³² The right hand of the statue is a modern repair (and was already missing when the statue was discovered), but the left hand holds a knot of drapery, too; the pose was sometimes understood to have suggested that her right hand held the other end of the drapery with which the figure is presumed to be covering herself, inviting artists, such as the later restorer of the statue, to invent her right hand and imagine how it held the other end of her drapery. One might also consider whether the child’s pose can be linked to the very well‐known third‐century Roman copy of a Hellenistic prototype of the Seated Boy with an Egyptian Goose (Gallerie degli Uffizi, Florence), which Raphael could have seen in Florence (where it had served as a source for Domenico Ghirlandaio and Leonardo, among others). The way in which the child supports himself against the Virgin’s stomach, and the position of his legs, is at least generically reminiscent of this much‐studied model.³³ Perhaps there is no direct connection – and Raphael’s drawings frequently demonstrate his astonishing ability to record or recall on demand children in multiple spontaneous poses – but it may be that here and in the Alba Madonna antique models assumed a new importance to him once he had arrived in Rome. The Virgin’s orientalising headdress has also been related to Roman fashions of the 1510s, and the landscape – which has been described as ‘a Tiberine landscape in the suburbs of Rome’³⁴ – is suggestive of the different ways in which arriving in Rome could change the art of a young artist.

The confidence of the underdrawing in the Garvagh Madonna suggests that the design had been fully worked out, at least for the figures, before Raphael started to draw on the panel. It is nevertheless the case that no drawings can be precisely connected with the painting. Instead, there are a series of sheets that study related motifs, but which are also linked – in an entirely typical fashion – to the genesis of other compositions that Raphael developed in his early years in Rome. The following group of drawings has frequently been discussed, for example by Catherine Monbeig Goguel, Barbara Brejon de Lavergnée, Carol Plazzotta, Jürg Meyer zur Capellen and Paul Joannides.³⁵

The drawing that is most often connected to the Garvagh Madonna is now in the Musée des Beaux‐Arts, Lille (fig. 11).³⁶ In the rapid, investigative metalpoint drawing on the pink‐prepared recto of this sheet, Raphael developed ideas for a Virgin and Child composition in which the Virgin would be shown sitting, with her thighs parallel to the picture plane, and her feet tucked behind her. The Christ Child sits on the Virgin’s lap, looking up from an open book and towards his mother. Raphael studied this group five times and paid particular attention to the child’s pose. The figure of Saint John the Baptist has not yet been introduced, the book would eventually be replaced by a carnation in the Garvagh Madonna, and the studies are in the opposite sense (i.e. reversed) from the finished picture. Nevertheless, there are strong connections between the pose of the child and his mother in the painting and in this drawing. Plazzotta has noted links between the schematic drawing of facial features, circles and arcs, and the underdrawing discussed above.³⁷ The sheet also has a study of the head and wings of an angel (possibly related to a figure in the ceiling roundel of Poetry in the Stanza della Segnatura) and studies of an architectural elevation, which have been associated with the planning of the background of the Disputa (although they could also relate to the School of Athens, or possibly to Raphael’s ideas for the church of S. Eligio degli Orefici in Rome).³⁸ The drawing has been cut down at the top and at the left, and some of the designs are partial as a result.

A brush and ink drawing on the verso of the same sheet develops the composition further (fig. 12). The Christ Child has been rotated through 90 degrees and, instead of concentrating on his mother and a book, he plays with the Virgin’s veil. The format as a small independent painting has been more fully established by the addition of framing lines that suggest the proportions of the Garvagh Madonna as painted. Raphael would eventually reverse the Virgin’s pose and move the child from the recto to the other side of his mother’s lap; he then added the figure of Saint John the Baptist, and made the act of passing a carnation between the two children the central pivot of the composition.

Another drawing in the Musée des Beaux‐Arts, Lille shows the Virgin seated on a block with her knees to the right (fig. 8).³⁹ Although slight, and probably studied from life, it concentrates on the fall of light on the Virgin’s draperies. Her headdress is comparable to that in the Garvagh Madonna. Neither the book nor a carnation is shown, and Saint John the Baptist does not appear. There is some disagreement over whether the left hand of a figure studied in the upper part of the sheet was also intended for the Virgin as suggested by Joannides, and Ginzburg has now attempted to relate this detail to one of Michelangelo’s ignudi on the Sistine Chapel ceiling as part of an argument for a later dating of the Garvagh Madonna (see under ‘Dating’ above).

A fourth, much quicker, study, which introduces the figure of Saint John the Baptist, is found on the left edge of a sheet where Raphael subsequently explored a solution for a reclining Christ Child (fig. 13), which looks ahead to the Madonna di Loreto (about 1511–12; Musée Condé, Chantilly), as well as a figure grouping (again with Saint John the Baptist) related to the apparently contemporaneous development of the Alba Madonna.⁴⁰ Below this, in the study connected to the Garvagh Madonna, the Baptist looks up at his cousin, and is included in the Virgin’s embrace as in the final painting; Saint John was also studied a second time further to the right.

A fifth sheet from the pink sketchbook, again in the Musée des Beaux‐Arts, Lille, studies the upper part of a male figure in profile in the School of Athens (fig. 14).⁴¹ To the right of this figure there is a quick study of a grouping of the Virgin and Child with Saint John the Baptist. The interaction of the Christ Child with the Baptist, who is again shown in profile, links this study with the Garvagh Madonna.

One further sheet from the pink sketchbook showing the Heads of the Virgin and Child is in the British Museum, London (inv. 1866,0714.79), and is variously related to the Mackintosh Madonna (NG2069) or the Garvagh Madonna.⁴² The angled gaze of the Virgin and the brightly lit description of her youthful features are contrasted with the sketchier treatment of her hair and of the Christ Child. Both this study and the painting are lit from the left, and silverpoint, as has been identified here,⁴³ was very well‐suited to such a study of the radiant face of the Virgin.

The metalpoint drawings referred to above are part of a group of 11 small‐scale studies that Fischel identified as coming from a single sketchbook (the ‘pink sketchbook’). Whether or not the drawings were ever bound together is debatable since there are no traces of binding holes or of more wear on the right side of the sheets, which is sometimes indicative of page turning. The sheets are now dispersed, but similar numbers added in pen and ink on several of them (probably dating from the sixteenth century) demonstrate that at least some of these drawings remained together during their later history. The sketchbook was apparently used by Raphael in his early years in Rome, and the compositions studied therein are usually dated 1509–11. As seen above, they included studies for the ceiling and walls of the Stanza della Segnatura, for the Alba Madonna, Mackintosh Madonna and Garvagh Madonna, as well as drawings after the antique.

As discussed under ‘Underdrawing’ above, Raphael drew onto his primed support using a metalpoint confirmed to be an alloy of lead and tin, as also used in the Madonna of the Pinks (NG6596). A metalpoint of this type seems to have been quite common at the time and has been found in a number of drawings by Italian artists of the late fifteenth and early sixteenth centuries, including in the faint study at the bottom right edge of Raphael’s Head of a Woman: and four faint studies of heads in the margin by a childish hand (British Museum, inv. 1895,0915.611). The larger drawing of the woman’s head on this sheet is instead in silverpoint, as also identified on the sheet from the pink sketchbook mentioned above, showing the heads of the Virgin and Child (British Museum, London, inv. 1866,0714.79).⁴⁴

The number of copies of the Garvagh Madonna, drawn, painted and engraved, attests to the fame of the composition, and possible knowledge of the composition by artists active in Ferrara has been brought into discussions of the early whereabouts of the picture (discussed under ‘Previous owners and provenance’ below). It should be acknowledged, however, that Raphael’s compositions became very famous, and that knowledge of them spread far, wide and fast. In the comments that followed, apparent knowledge of the composition can be identified in north, central and eastern Italy, as well as in the work of artists who worked in Rome. In other words, the reputation of the picture does not clearly point to a particular location in the sixteenth century.

In addition, looking at the ripples across Italy that the copies and versions listed below imply, it is probable there has been a historic conflation of two compositions: the Garvagh Madonna and a related grouping of the Virgin and Child with Saint John. It has not been observed before, but it seems highly likely to this writer that there was a Raphael Virgin and Child with Saint John which resembled the Garvagh Madonna but introduced the changes that are visible in the engraving of 1642 signed ‘G.M.V.’ (fig. 15): namely, that the Christ Child has his left arm around the Virgin’s neck, that she holds his foot with her left hand, and that the young Saint John the Baptist is seated. Each of these modifications is intensely Raphaelesque, and an inscription on this engraving states that Raphael had painted such a composition (‘Raphael Sanctius pinx.’).

It is this composition, not the Garvagh Madonna, that lies behind a number of the apparent copies after Raphael, including those that underpin the arguments made for an early Ferrarese provenance. On the basis of reflections of the composition in the work of Benvenuto Tisi, called Il Garofalo (about 1481–1559), Giovanni Battista Benvenuti, called L’Ortolano (about 1487–after 1527), Ludovico Mazzolino (active 1504–about 1528) and especially Zenone Veronese (1484–1542), Alessandro Ballarin has suggested that the Garvagh Madonna arrived in Ferrara in about 1510.⁴⁵ The principal comparison for Ballarin is a picture by Zenone, formerly in Blaise Castle, Bristol, which is signed and dated 1529, along with a second composition in the Muzeum Narodowe, Warsaw.⁴⁶ Although these compositions, and details such as how the Virgin holds a knot of drapery, do suggest some connection to the Garvagh Madonna or, as I would argue, to a related composition, it is certainly not enough of an argument to create an early Ferrarese provenance for the National Gallery picture. Instead, the various pictures which develop the composition recorded by Zenone suggest that there might have been a variant by Raphael or his workshop which included features that are not found in the Garvagh Madonna, as listed above. That such a variant was known by the 1520s in northern Italy is not exceptional and similar arguments apply to the weaker links drawn with Garofalo, L’Ortolano and Mazzolino.⁴⁷ Moreover, and as discussed by Carol Plazzotta under NG2069, we know that Garofalo and other northern Italian artists visited Rome at an early date, so one should not give too much weight to the connection.

David Ekserdjian discusses the work attributed to Zenone in Warsaw as probably the work of Raffaellino del Colle,⁴⁸ and points to versions of this composition by Girolamo del Pacchia, Ventura Salimbeni, Marco d’Oggiono and Bagnacavallo.⁴⁹ It has been suggested that Francesco Mazzola, known as Parmigianino, was responding to the Garvagh Madonna in his Mystic Marriage of Saint Catherine with Saint James the Lesser (about 1527; Louvre, Paris), and Federico Barocci apparently copied the Christ Child in an early drawing.⁵⁰ As well as recognising that these artists were geographically dispersed across the Italian peninsula, it is important to consider carefully whether the Garvagh Madonna or the related composition of the Virgin and Child with Saint John lie behind, for instance, Perino del Vaga’s tondo of The Holy Family (about 1540; Princely Collections, Liechtenstein). The copies listed below are those that directly follow the Garvagh Madonna, not the alternative model reflected in the engraving of 1642.⁵¹

Sassoferrato represents a later moment in the reputation of Raphael’s Madonna; he copied Raphael’s work throughout his career and with evident impact on his own developing style. Of Sassoferrato’s copies that are known, the most important is a canvas in the Musée du Louvre, Paris (MI 629; usually dated 1625–50).⁵² He omitted the carnation and varied the landscape in this otherwise faithful copy. Sassoferrato presumably knew the original in Rome (where he is recorded from at least 1641), and its palette must have had a particular appeal to him. A further copy attributed to Sassoferrato (sold at Fausti in Milan, 12 May 1913, lot 40) is now in the Davis Museum at Wellesley College, Wellesley, MA and further attests to his interest in this composition.⁵³

Painted copies

A painted copy on panel (29.5 × 22.3 cm), said to date from the seventeenth century, is in the collection of National Galleries Scotland (NG1854).⁵⁴ It eliminates the arched tops of the central pier and places a curtain at the top of it.

A number of pictures were listed as copies after the Garvagh Madonna in the nineteenth century that appear from the descriptions given to have had very little direct connection with Raphael’s original picture. Other copies are so vaguely referenced as to be effectively untraceable. Johann David Passavant recorded painted copies in the following collections in 1839:⁵⁵

• 1. Bergamo, Accademia Carrara.
• 2. Urbino, Staccoli collection. This picture was further described by Crowe and Cavalcaselle and attributed by them to Palmerini.⁵⁶
• 3. Milan, Silva collection.
• 4. Rome, Fesch collection. The figures were said to be full‐length.
• 5. Rome, Campana collection. Two copies were recorded in Eastlake’s diary in 1856.⁵⁷
• 6. A picture formerly in Spain and subsequently in a collection in Germany.

A reference to a picture mentioned by Luigi Pungileoni was listed by Passavant in 1839, but removed in 1860, presumably because Passavant had by then eliminated any connection.⁵⁸

Friedrich Heinrich von der Hagen, Ignaz von Wessenberg⁵⁹ and Georg Nagler⁶⁰ described a version of the Garvagh Madonna in the Camuccini collection, Rome, in 1817, 1827 and 1836 respectively, in the latter case with reference to the original being in the collection of Lord Garvagh.

All of the following undated copies after the Garvagh Madonna are recorded and illustrated at the Raphael Research Resource, and photographs are in the National Gallery Archive, London.⁶¹

• 1. Sale, Christie’s, London, 28 March 1952, lot 26, painted copy, inscribed on verso of photograph ‘E.S. Moss, 15 × 12’ (anon.).
• 2. Sale, Sotheby’s, London, 7 October 1964, lot 53, painted copy, 16¼ × 13¼ in. (= 41.3 × 33.7 cm), (anon.).
• 3. Sale, Christie’s, 14 February 1969, lot 25, painted copy, 14½ × 12 in. (= 36.8 × 30.5 cm), (anon.).
• 4. A painted copy (dimensions unknown) was said to be in Rome and was inscribed on the reverse of a photograph: ‘Canvas. Said to be same size as ours. / Owner. / Diego Pettinelli / Via Eleonora Pimentel 2 / Rome 196’ (anon.).⁶²
• 5. Sale, Christie’s, London, 3 March 1972, lot 81, 16 × 14 in. (40.6 × 35.6 cm), (anon.). This appears to be the same picture that had previously been sold at Sotheby’s, London, 21 July 1965, lot 79.
• 6. Sale, Sotheby’s, New York, 9 December 1988, lot 140, medium and dimensions unknown (anon.).
• 7. Sotheby’s, Milan, 1 June 2004, lot 122, oil on canvas, 65 × 50 cm (optimistically presented as Follower of Jacopo Pontormo, nineteenth‐century (?) copy). Annotation on the label of an ektachrome.⁶³ The lot did not sell.

Drawn copies (as far as possible in chronological order)

• 1. Florence, Uffizi, inv. GDSU 11530F, black chalk, 12.4 × 7.2 cm (a partial copy of the Christ Child in an album of drawings (Libro A) by Federico Barocci [about 1533–1612]). This is on the same sheet as a 1:1 copy after the Madonna of the Pinks (NG6596) identified by Luca Baroni with the suggestion that Barocci might have known the composition in Urbino, and the connection to the Garvagh Madonna was made by Ekserdjian.⁶⁴
• 2. Writing in 1935 Arthur Popham described two unidentified brush and wash drawings (ex‐Woodburn) that copied the hands and a section of drapery from the Garvagh Madonna as ‘almost contemporary’ to the original painting.⁶⁵ I have not been able to trace these drawings.
• 3. Edinburgh, National Galleries Scotland, inv. D 4823 II F.III, pencil, subsequently squared in brown ink, 33.4 × 26.4 cm (a careful copy by Richard Cooper [1701–1764]). This was apparently based on his tracing on oiled paper, also in National Galleries Scotland (inv. D 4823 I F.LIV; 39.4 × 31.3 cm), which is inscribed: ‘Traced from Raphael’s picture. Rome’. Cooper was principally an engraver, who is known to have spent an extended period in Italy, and the meticulous nature of the squared drawing might suggest preparation for an engraved copy.⁶⁶
• 4. Formerly San Francisco (anon., eighteenth century?). This drawn copy (omitting the architecture) was said to be ‘With Russell Hartely / San Francisco, 1963’ on the verso of a photograph in the Gallery’s dossier for NG744.⁶⁷
• 5. Windsor, Royal Collection, RCIN 850884, about 1853–76, albumen print, 27.7 × 22.3 cm (sheet). An old photograph of an anonymous drawing, which was in the collection of the late August Grahl (1791–1868) in Dresden when the photograph was acquired (before 1876).⁶⁸ The drawing appears to be a relatively faithful, later, copy of the Garvagh Madonna. It is identical with a drawing now in the Whitworth Art Gallery Manchester, inv. D.1930.25, pen and brown ink, and black chalk, on beige paper, 28.8 × 22.8 cm.
• 6. Greenwich, National Maritime Museum, inv. PAI6766, gouache, 43.6 × 33.0 cm (sheet) (anon., probably nineteenth century).

Engravings (as far as possible in chronological order)

• 1. London, British Museum, inv. U,5.107, 29.0 × 22.5 cm. The composition was engraved, in reverse, by Bartolommeo Passarotti (1529–1592), who inscribed his initials: ‘B.P.’. This engraving, which omits the setting and landscape background, can probably be dated 1550–90.⁶⁹
• 2. Windsor, Royal Collection, RCIN 850882, about 1600–40, 20.2 × 16.3 cm (sheet). A small engraving attributed to Girolamo Imperiale (d. 1639?).⁷⁰ The composition is reversed.
• 3. Windsor, Royal Collection, RCIN 850886, 24.7 × 20.3 cm (sheet), 23.7 × 20.0 cm (image). A variant (fig. 15), which eliminated the architecture and introduced a curtain on the right, was engraved by an artist who signed and dated his print: ‘G.M.V. del. et sculp. Romae 1642’. This version has been discussed above.⁷¹
• 4. Passavant refers to an oval engraving by ‘F. Andriot, chez Vallet, au Buste de Louis XIV’; I have not seen this print.⁷²
• 5. An engraving by Girolamo Carattoni (1749–about 1809).⁷³
• 6. An anonymous engraving, said to be based on a 1:1 tracing from the original painting, was published in Séroux d’Agincourt 1823, III, p. 172, IV, pl. CLXXXIV. Francesco Longhena attributed this engraving to Alessandro Mochetti (1760–1812).⁷⁴ A different engraving illustrated the Italian edition of 1829.⁷⁵
• 7. Windsor, Royal Collection, RCIN 850879, 84.4 × 60.8 cm (sheet), 39.5 × 31.0 cm (plate‐mark) (Augustin Bridoux [1813–1892], 1855).⁷⁶
• 8. Windsor, Royal Collection, RCIN 850883, 19.3 × 15.1 cm (sheet), 54.5 × 44.2 cm (plate‐mark). A partial engraving, which reproduces the heads of the Virgin and the Infant Christ in the National Gallery painting (Joseph Alois Drda [1783–1883], about 1803–73).⁷⁷
• 9. An unsigned engraving was also reproduced in Müntz 1881, p. 395.
• 10. An anonymous German print of 1884.⁷⁸

Before and after Cardinal Aldobrandini

The picture has an Aldobrandini provenance in the seventeenth and eighteenth centuries but how it arrived in Ferrara and entered the collection of Cardinal Pietro Aldobrandini (1571–1621), where it is recorded with some certainty from 1603, is unclear. Three possibilities for its earlier provenance have been suggested: that the painting was acquired by Duke Alfonso I d’Este (1476–1534), perhaps directly from the artist; that it came to Ferrara with Margherita Gonzaga (1564–1618) when she married Alfonso II d’Este (1533–1597) in 1579; or that it entered Este collections via Alfonso II’s sister, Lucrezia d’Este (1535–1598), and so might have an earlier provenance from Urbino.

As discussed under ‘Reputation’ above, on the basis of supposed reflections of the composition in the work of Garofalo, L’Ortolano, Mazzolino and especially Zenone Veronese, Ballarin has suggested that the Garvagh Madonna arrived in Ferrara in about 1510.⁷⁹ This could suggest that it was acquired by Duke Alfonso I d’Este, and the argument has now been developed by Silvia Ginzburg who has tentatively linked the picture with a payment of 20 ducats that was authorised by Alfonso in May 1514.⁸⁰ But, as discussed under ‘Dating’ above, the latter argument is not persuasive, and the claimed impact of the picture on northern Italian artists is insufficient to create an early Ferrarese provenance for the picture.

An alternative explanation is that the picture entered the Este collections when Margherita Gonzaga married Alfonso II d’Este in 1579. Georg Gronau and later Cecil Clough published an inventory of Margherita’s capeletta (small chapel) in Ferrara in 1586–8 and listed its contents.⁸¹ Clough proposed that the ‘madonna de Rafaello da Urbino’ included in this inventory might be the Garvagh Madonna, but the description is incomplete (and Saint John the Baptist is not mentioned). Alfonso died without a legitimate heir and, as discussed below, his and other Este collections passed to Cardinal Pietro Aldobrandini, but tracing this picture to Margherita is not straightforward.

The third and perhaps most plausible possibility is that the picture can be identified as one of the four Madonnas by Raphael that entered the Aldobrandini collection with the possessions of Lucrezia d’Este. She married Francesco Maria II della Rovere, Duke of Urbino (1549–1631), in 1570 and received a gift of pictures, jewels and silver from the city of Urbino and from her father‐in‐law Guidobaldo della Rovere (1514–1574) the following year. These included several Madonnas attributed to Raphael, which were subsequently listed in an inventory of her possessions made in Ferrara in 1592; none of these mention any additional figures or offer any further description that might narrow down their identification.⁸² She had returned to her native Ferrara in 1576, and a separation from her husband was formalised by the Pope in 1578. Her most precious possessions were housed (like those of Margherita Gonzaga) in a small oratory, and pride of place was given to a Madonna by Raphael that was described as being mounted in a gold reliquary made of ebony, cornelians and lapis lazuli.⁸³ This was one of four Madonnas attributed to Raphael but otherwise imprecisely described in Lucrezia’s inventory of 1592. When Lucrezia died in 1598 she left most of her property to Cardinal Pietro Aldobrandini, cutting out her direct family and with them a Ferrarese future for her inheritance. Duke Alfonso II d’Este, who died without having produced an heir in 1597, had designated his young cousin Cesare d’Este (1562–1628) to succeed him. Lucrezia, the late duke’s sister, met with the papal legate to Ferrara, Cardinal Pietro Aldobrandini, and agreed that the Duchy of Ferrara and its possessions should revert to the papacy. She appointed Aldobrandini as her residuary legatee, and as a result her part of the extraordinarily rich Este collections passed to him on her death, most of which was transported to Rome.⁸⁴ While it is possible that the Garvagh Madonna was the ‘madonna di Raffaello, ch’è antica di casa nostra’ (‘a Madonna by Raphael, which is historically of our house’) which the Duke of Urbino tried to reclaim in 1598,⁸⁵ it has also been suggested that this was instead the Small Cowper Madonna now in Washington.⁸⁶ The picture’s description in the related correspondence is too imprecise to allow one to be sure that it was either of these Madonnas that the duke was prioritising over any of the other pictures attributed to Raphael in Lucrezia’s inventory.⁸⁷ In any event, his plaintive entreaties were destined to be in vain given that Cardinal Aldobrandini was the nephew of Pope Clement VIII Aldobrandini (1536–1605; Pope 1592–1605) and had been sent to secure Ferrara as his legate.

As discussed immediately above it is likely that four Madonnas entered the Cardinal’s collection with Lucrezia d’Este’s inheritance; the Garvagh Madonna and the Madonna del Passeggio (about 1516, National Galleries Scotland, Edinburgh) were certainly in his collection a few years later.⁸⁸ In some cases, a description of a picture is clear enough at every turn to allow one to go through its provenance at speed, but in the case of the Garvagh Madonna it is necessary to set out each step in detail as the correct identification is not always self‐evident. The first certain reference is found in Cardinal Pietro’s inventory of 1603 (drawn up by his maggiordomo, Girolamo Agucchi, 1555–1605, brother of the more famous Giovanni Battista Agucchi, 1570–1632), which lists ten works by Raphael.⁸⁹ The Garvagh Madonna is described as follows: ‘3. Una Madonna con Christo et S. Giovanni, in Quadro piccolo, di Raffaele sudetto’ (‘A Madonna with Christ and Saint John, in a small picture, by the above‐mentioned Raphael’).⁹⁰ This identification as No. 3 accords with the inscription on the back of the picture (‘N.ᵒ 3 · D · RAFFAEL · D· VR·^NO’), which when compared to those on other works in the Agucchi inventory of 1603 proves that this inscription was added at that date or at least with reference to that inventory.⁹¹ The 1603 inventory is unusual for the absence of any indication of the rooms in which objects were kept, and it has been suggested that in 1603 the collection was awaiting a permanent home. Within a few years, however, it was apparently installed in Rome in Cardinal Pietro’s villa at Monte Magnanapoli on the slopes of the Quirinal Hill.⁹²

Following the death of Cardinal Pietro his possessions passed relatively swiftly to his sister, Olimpia Aldobrandini senior (1567–1637).⁹³ To consolidate family property she had married a relative, Gianfrancesco Aldobrandini (1545–1601), and the couple had 12 children before his death in 1601.⁹⁴ In 1626 Olimpia senior gave Cardinal Pietro’s villa at Monte Magnanapoli and its collection of 190 paintings (including the Garvagh Madonna, which was again identified as No. 3) to her son Cardinal Ippolito Aldobrandini (1596–1638), and he assumed greater importance in the family inheritance when his elder brother Giorgio Aldobrandini (1591–1637) died in 1637 just a few weeks after his mother.⁹⁵ Cardinal Ippolito, who was subject to an entail of 1611 that protected Aldobrandini properties from sale, inherited the rest of his mother’s and brother’s possessions, and when he died in 1638 the picture was described as follows: ‘Una Madonna con Christo e San Giovanni in quadro piccolo come si dice di Rafaele sudetto in tavola palmi uno e mezzo, cornice negra’ (‘A Madonna with Christ and Saint John, in a small picture, as said by the above‐mentioned Raphael, on panel one and a half palmi, black frame’).⁹⁶

Cardinal Ippolito’s will established that his 14-year‐old niece, Olimpia Aldobrandini junior (1623–1681, also known as Olimpia Aldobrandini Borghese Pamphili, Princess of Meldola, Sarsina and Rossano), would inherit as the last descendant and sole heir of the Aldobrandini family (fig. 16), and it anticipated the issue of her imminent future marriage by creating an important precedent: that her property should pass to her second‐born male heir (recognising that the first‐born would very probably be the heir to another family and another name).⁹⁷ In fact, marital politics were already afoot when Cardinal Ippolito died of a fever in his palace on the Corso on 19 July 1638. Realising her importance to the Aldobrandini inheritance, she had been plucked from the monastery (SS. Sisto e Domenico) to which she had been sent in 1637 following the death of her father and the remarriage of her mother. Cardinal Ippolito brought her and some young female companions to live with him in the Palazzo Aldobrandini al Corso (now the Palazzo Doria Pamphili), and he had started to negotiate a power‐match with Paolo Borghese (1624–1646). Concerned by this potentially powerful alliance, and inimical to Cardinal Ippolito, Pope Urban VIII Barberini (1568–1644) tried instead to marry the young Olimpia to his nephew Taddeo Barberini (1603–1647), or failing that to send her back to monastic life and (after Ippolito’s death) declare the Aldobrandini line extinct. The pressure for the young Olimpia to renounce her inheritance and return to the monastery included Cardinal Francesco Barberini sending the mother of her closest companion to try to persuade her to adopt the veil, and subsequently an attempt to bring her into the presence of the Pope, whose instruction she could not have resisted; this latter manoeuvre was deflected by sending a message to the Pope on 24 July that she was unable to attend due to a fever. The following day – less than a week after her uncle’s death – she was sufficiently recovered to marry Paolo Borghese and to preserve the Aldobrandini inheritance, ‘con dispiacere dei Barberini, i quali la disideravano in casa loro’ (‘with the displeasure of the Barberini who wanted her in their house’).⁹⁸

Despite the picture being recorded in Olimpia Aldobrandini junior’s collection before 1665 and again after her death, it is often presumed to have remained in Borghese ownership (or perhaps custody) after Donna Olimpia remarried Camillo Pamphili in 1647 (following Paolo Borghese’s death the previous year).⁹⁹ This analysis hinges around identifying the present picture as the ‘Vergine, con Christo, e San Giouannino, […] di Raffaelle’ (‘Virgin with Christ, and Saint John, … by Raphael’), which Iacomo Manilli described as being at the Villa Borghese in 1650.¹⁰⁰ While possible, Sara Tarissi de Jacobis has now demonstrated – in part on the basis of understanding the role of the Aldobrandini fideicommissum, and in part on a comparison of what was described where – that if a picture was identified by Manilli then it probably does not have an Aldobrandini provenance.¹⁰¹ Be that as it may, Donna Olimpia’s inheritance was divided in 1682 between the sons of her first (Borghese) and second (Pamphili) marriages, and it seems highly likely that the picture appears in the 1693 inventory of her eldest son, Giovanni Battista Borghese (1639–1717), 2nd Prince of Sulmona (and specifically in the Palazzo Borghese in Campo Marzio, where it was certainly described in 1787); but the evidence points to the picture always being kept apart as an Aldobrandini heirloom.¹⁰²

The picture is next recorded in the collection of Paolo Maria Borghese (1733–1792), who took the title of Prince Aldobrandini (this codification of the inheritance of Aldobrandini property by the second‐born son is further discussed below). It had probably passed to Giovanni Battista’s son Marcantonio III Borghese (1660–1729), and subsequently to his son Camillo I Borghese (1693–1763), before appearing in Paolo Maria Borghese’s apartments. At some point earlier in the eighteenth century the picture was also traced by Richard Cooper on a visit to Rome; see ‘Copies and versions’ above. It was described in Prince Aldobrandini’s mezzanine apartments in the Palazzo Borghese in 1787 by Friedrich Wilhelm Basilius von Ramdohr: [in the margin: Part of the palace [which is] of Prince Aldobrandini] In that palace, in which now also lives Prince Aldobrandini, uncle of Prince Borghese, are some rooms, in which there is still a selection of paintings … [in the margin: A Holy Family by Raphael, of his middle period.] A Holy Family by Raphael. A most precious small painting of his middle period. The composition is very good. The Christ is beautiful, and the Saint John true, only the head of the Madonna compared with the others, is less beautiful. The disegno is most delicate. One notices from the colouring that the master had painted a lot al fresco at that time. The colours are not very much rubbed.¹⁰³

The resurrection of the Aldobrandini family name followed an agreement of 1767 whereby the title (and ex‐Aldobrandini possessions) would pass to the second son of the Borghese family (with the first son maintaining the Borghese title).¹⁰⁴ Paolo Maria was the great‐grandson of Olimpia Aldobrandini, and the possessions had apparently been kept apart from the rest of the Borghese collection over the intervening century.¹⁰⁵ They did not appear in later Borghese inventories as a result of this division.

Von Ramdohr’s and Mariano Vasi’s identifications were confirmed in 1823 when the Garvagh Madonna was engraved in Jean Baptiste Séroux d’Agincourt’s (1730–1814) posthumous Histoire de l’art pars les monuments as ‘à Rome, au palais Borghèse, dans l’appartement du prince Aldobrandini’ (‘in Rome, at the Borghese palace, in the apartment of Prince Aldobrandini’).¹⁰⁶

By that date, however, Paolo Maria had died without fathering an heir and it appears that his younger brother, Giovanni Battista Borghese Aldobrandini (d. 1802), inherited the Aldobrandini title and inheritance. Only after his death did the title revert to the recognised route when it passed to Francesco Paolo Borghese (1776–1839), the second son of Prince Marcantonio IV Borghese (1730–1800) who was, however, a minor when Paolo Maria Borghese Aldobrandini died in 1792. Following Napoleon’s occupation of Italy (and especially during the short‐lived first Cisalpine Republic, 1797–9), families fearing for the future ownership of their possessions increasingly looked to sell these to dealers who were ready to profit from the political uncertainty. The Garvagh Madonna was apparently purchased from Giovanni Battista Borghese Aldobrandini by the artist and picture‐dealer, Alexander Day (1773–1841).¹⁰⁷ Day acted with the Camuccini brothers, Pietro and Vincenzo (1760–1833; 1771–1844), and in partnership with other British speculators and collectors, notably the 2nd Baron Northwick (1769–1859), acquiring numerous pictures in this period (see also Saint Catherine of Alexandria [NG168], Madonna of the Pinks [NG6596]).¹⁰⁸ In a note of 2 October 1799 Pietro Camuccini listed pictures that had been jointly acquired and were now in Day’s hands, including: ‘Un Quadretto di Raffaelle d’Aldobrandini la Madonna il Bambino e S. Gio. Tavola’ (‘A small picture by Raphael from the Aldobrandini, the Virgin, Child and Saint John. Panel.’).¹⁰⁹ In 1800 a selection of these were sent to London; they arrived by October, and from 2 February to 7 May 1801 they were displayed at the London gallery of the painter Henry Tresham (1751–1814) at 20 Lower Brook Street for sale by private contract.¹¹⁰ The Garvagh Madonna was one of these, as was Saint Catherine of Alexandria (NG168). The former was valued at 1,500 guineas and was sold to George Canning, 1st Baron Garvagh (1778–1840), in 1818 (as Passavant noted, the search for a buyer took some time). Before this sale was completed the picture was lent in 1816 to the Royal Mews Gallery, London, and after Lord Garvagh’s purchase it was lent by him to an exhibition at the British Institution in 1819.¹¹¹ It was subsequently described in his collection by Passavant in 1833.¹¹² The painting has at least two Garvagh seals on its reverse,¹¹³ and an address at 41 Hertford Street, London.¹¹⁴

‘The Garvagh Raphael’¹¹⁵

George Canning, 1st Baron Garvagh, was an Anglo‐Irish politician, and served as MP for Sligo (1806–12) and Petersfield (1812–20). He was a cousin of George Canning (1770–1827), who was Foreign Secretary, President of the Board of Control and, at the very end of his life, Prime Minister. Garvagh married Charlotte Isabella Rosabella Bonham (known as Rosabelle; 1804–1891) in July 1824. After his death the 1st Baron’s possessions passed (as per the terms of his will of 1836) for the duration of her life to his widow, and were then to pass to one of his sons.¹¹⁶ The picture was lent after the 1st Baron’s death to the British Institution in 1845 and was described by Gustav Waagen in the collection of Rosabelle Garvagh in 1854.¹¹⁷

The 15 years of negotiations to acquire the picture for the National Gallery, at one of the highest prices ever paid for any painting in the period,¹¹⁸ and after the question had been twice referred to Cabinet, is a vignette of art politics and society in the mid‐nineteenth century. In July 1850 the Trustees of the National Gallery were made aware that the painting owned by the Garvagh family might become available for purchase and they wrote to the Treasury to request whether they might spend up to £2,000 on its acquisition.¹¹⁹ No opinion was forthcoming on the grounds that the National Gallery had not made a specific, argued recommendation for the purchase; and the proposed acquisition lapsed at that time.¹²⁰ A direct approach to the family was made in March 1851 by Lord Monteagle, apparently offering £1,000 subsequently increased to £1,500,¹²¹ but this offer was rejected by the 2nd Baron Garvagh (1826–1871).¹²² A year later the 2nd Baron approached the Trustees who ‘took into consideration an offer made to them for sale of the Aldobrandini Raffaelle, the property of Lord Garvagh’ on 8 March 1852.¹²³ The Keeper, Thomas Uwins (1782–1857), was instructed to write to Lord Garvagh offering £3,000, and his letter was written that day.¹²⁴ The 1st Baron’s widow, the Lady Dowager Garvagh, replied on 17 March on behalf of her son. She rejected the Gallery’s offer and returned Uwins’s letter, adding that her son had not understood the need for the late Lord Garvagh’s executors – of which she was one – to agree to any sale.¹²⁵

The Lady Dowager Garvagh lived as a widow for 46 years, and her tenacity emerges strongly in the tortuous story of this painting. A letter written by Prince Hermann von Pückler‐Muskau (1785–1871) a few years after her marriage creates a vivid portrait: Now there was nothing left to keep me in London, except Lady Garvagh is here, and alone, and so attractive! It would be wrong to avoid such a friend, especially since I have no intention of falling in love with her. … For my part, I am quite content with nothing more than a tender friendship, especially when I can see it in the gentle gaze of soulful blue eyes, hear it spoken by a mouth of crimson red, and feel it affirmed through the warm pressure of a perfectly proportioned, velvet hand. Add to this portrait the innocent expression of a dove; long, dark brown curls; a slender waist; and the most beautiful English complexion – there you have Lady Garvagh, as she lives and breathes.¹²⁶

Lady Garvagh was a little more accommodating to approaches from the National Gallery in the summer of 1856. She arranged for Sir Charles Eastlake (1793–1865) to see the picture at her London home at 26 Portman Square in her absence, and wrote to him on 10 July saying that he was welcome to visit again, and that the picture might be available for a suitably elevated price.¹²⁷ Eastlake called on her on 15 July and reported his meeting to the Trustees. The Lady Dowager referred back to historic offers from Lord Monteagle (1790–1866; Trustee 1835–66), said to be £1,500, and Lord Ellesmere (1800–1857; Trustee 1835–57), said to be £3,500, and left Eastlake with the impression that nothing less than this, and more likely £4,000, would be required in order to secure the picture. Eastlake asked to be kept informed if the family ever wanted to sell.¹²⁸ This was reported to the Trustees on 4 August, but did not result in any offer at that time.¹²⁹

Eastlake had been appointed as the National Gallery’s first Director in 1855, having previously been Keeper from 1843 to 1847. His directorship saw a transformation of the Gallery, and his ambition, which included an expressed desire to rival the collections of the capitals of Europe, emerges in the narrative of how the Garvagh Raphael was added to the collection. All of the following account points to enormous efforts to align the necessary parties behind this acquisition at a record‐breaking price. Coming just six months before Eastlake’s death, the acquisition of the Garvagh Madonna can be seen as the crowning achievement of his time as Director.¹³⁰ His interest in the picture’s pictorial qualities is made clear in letters that he wrote in pursuit of securing it for the National Gallery, further discussed below. The picture’s tone and palette, as well as the simplified ovals of Raphael’s figure style in this particular painting, coincided with his own aesthetic as a painter. This can be seen by comparing the facial type, tone and (to a lesser extent) palette of the Garvagh Madonna with Eastlake’s 1835 Portrait of Mrs Charles H. Bellenden Ker (fig. 17).

At some point in 1862 Eastlake offered £4,000 for the picture. As the 2nd Baron was abroad, a reply was delayed until after his return and on 6 November the Dowager Lady Garvagh informed Eastlake that the family would not part with the picture for less than £9,000.¹³¹ The matter was not pursued immediately, but on 18 June 1863 Eastlake had a conversation with Lord Garvagh and followed it up with a letter on 1 July, in which he summarised his understanding that Garvagh had rejected an offer of £7,000 from a European collector, and had recently received an offer ‘from a friend now in America’ for £8,000.¹³² Eastlake had understood that if £8,000 was offered by the government then this would succeed, and he asked for the negotiations with the interested party in America to be suspended while the National Gallery considered its position. Lord Garvagh replied on 4 July, saying that he would accept 9,000 guineas.¹³³ The Trustees of the National Gallery were kept apprised of these communications on 6 July,¹³⁴ and Eastlake’s long‐standing admiration of the picture (now being referred to as the ‘Garvagh Raphael’) was made abundantly clear as he informed them in a detailed letter that: It must be quite unnecessary for me to speak in praise of a work which has been extolled by eminent writers on art living & dead. The picture may be regarded as representing the characteristic excellence of the master. It is not merely a name, but the charm which the finest works by Raphael possess which gives it value.¹³⁵

Eastlake summarised the contact over the previous year, and the price now requested, and went on to opine that the price was reasonable, specifically adducing comparison with that paid to William Beckford 24 years earlier for Raphael’s Saint Catherine (£6,000).¹³⁶ As a result he was instructed to go and see William Gladstone (1809–1898), Chancellor of the Exchequer. Gladstone and the Cabinet were unimpressed by the offer that was put to them, and the Chancellor wrote to Eastlake on 11 July refusing to put a special government grant behind this purchase.¹³⁷ On 17 July Eastlake wrote to Lord Garvagh, copied to the Lady Dowager, stating he ‘was not prepared to proceed further … upon the basis put forth in his Lordship’s letter’.¹³⁸ These negotiations were at an end.

Two years later, in May 1865, the possibility was again being discussed and Eastlake wrote another long letter to the Trustees, which was read at their meeting on 5 June. He set out the most recent history of the negotiations and how they had been broken off. Once more he recommended the painting and asked ‘who would not wish to see one celebrated & first rate picture added to the National collection, giving that collection greater European importance’.¹³⁹ Reference was again made to the stipulation that the picture should be referred to as the ‘Garvagh Raphael’, and speed was encouraged to resolve the matter before the 2nd Baron departed for his regular summer sojourn in Norway. The Trustees resolved to write to the Chancellor of the Exchequer on 1 June, and a letter was immediately dispatched. The following day Eastlake received a positive reply from Gladstone, who had ‘mentioned your considerate letter to the Cabinet’.¹⁴⁰ This news was relayed to Lord Garvagh before it was even reported to the Trustees, and on Saturday 3 June the vendor instructed Eastlake to make payment of £9,000 into his mother’s bank account at Coutts.¹⁴¹ The Trustees formally agreed to purchase at this price on 5 June, and counterparty agreement was reported on 13 June.¹⁴² A receipt for the picture was issued on 27 June, and on 3 July the Director was able to report delivery of the picture to the Trustees.¹⁴³ Even before acquisition there was some discussion of whether the work needed any restoration, and where it should hang, and on 4 July Ralph Wornum could report (with evident pleasure) that the painting was now hanging with An Allegory (‘Vision of a Knight’) and Saint Catherine (NG213 and NG168).¹⁴⁴

The author is grateful to the following for their assistance in the preparation of this entry: Christopher Baker, Costanza Barbieri, Rachel Billinge, David Ekserdjian, Pier Ludovico Puddu, Nick Penny, Carol Plazzotta, Jan Sammer, Jeffrey Spier, Marika Spring and Aidan Weston‐Lewis.