When bought in 1937, this painting had an interesting and varied critical history. It first appeared in the collection of the Duke of Marlborough in the early eighteenth century, described as “The Family of Rubens done by Vandike of Ten figures.” However, by mid century, the picture was instead attributed to Rubens or to painters in his circle, and the family became anonymous. While well known when it hung at Blenheim palace, its distinctive style meant that the attribution remained in dispute: Van Dyck’s early work – strongly influenced by Rubens – was little known. In 1886, when the Duke of Marlborough was forced to sell much of his collection, the painting was returned to Van Dyck, perhaps at the prompting of Wilhelm Bode, an influential scholar and museum director.1 Although the attribution was largely accepted, doubts persisted well into the twentieth century.2
Van Dyck, born in 1599, was a precocious artist – in part thanks to his great natural talent, but also due to Rubens’ influence on him as a young man.3 By his late teens, he was already working with Rubens, entrusted with great responsibility in the workshop, and he would soon set up as master in his own right. This painting seems to have been part of his independent production, made before he left Antwerp in 1621: he cannot have been older than twenty-two when he painted it.4 His early work is difficult to date, a result of the tension between his precocious facility and the requirement, as an assistant in the Rubens’ workshop, to work within defined parameters. His proximity to Rubens, however, enabled the young painter to develop very rapidly, clearly with a desire to match his master’s extraordinary talent and with the need to distinguish himself. Van Dyck grasped the older painter’s critical reflections on the different artistic traditions, Northern and Italian, as seen in the synthesis of the two in Rubens’ mature work. In the Rubens’ workshop, he would have had access to an important collection of Italian art, originals and copies. Let the Children Come shows the influence of these varied strains; technically extravagant and formally bold, it is an ambitious attempt to impress. So overt is the bravura application of paint and the freedom with which Van Dyck approached the composition, that we see the painter claiming mastery. Yet these are also the elements that reveal a young artist at the limits of his ability.
This canvas is one of a number of biblical scenes painted by Van Dyck at this time. While his subject matter was not common,5 there were precedents for depicting this scene, as there were for his choice to integrate biblical figures with others who appear to be from the life. Yet, while Van Dyck intended to make the family look distinct from the biblical characters, he elected not to dress them in contemporary clothing. Rather, they wear fantastical costume, derived from older paintings. This avoids jarring viewers’ sensibilities through the sharp contrast of ancient and modern, and instead makes the picture “timeless.”6
One precedent with which Van Dyck may have been familiar is a painting by Adam van Noort (illus. 1) a prominent Antwerp artist and Rubens' and Jordaens’ teacher.7 It shows a group of women presenting children for Christ’s blessing. Many of the elements seen in Van Dyck’s painting are here: the bi-partite composition, the use of pseudo-historic clothing, modelled upon examples seen in Italian paintings, and the inclusion of heads clearly drawn from the life amongst others more generically treated.
Van Dyck chose to show the moment at which Christ blesses the children, after having admonished his disciples for not permitting them to approach him. “Let the children come to me, and do not hinder them, for to such belongs the kingdom of heaven.” (Matthew 19.14.) The three apostles show a range of reactions: surprise, consternation, acceptance.
A drawing (illus. 2) shows Van Dyck first exploring the subject. While the composition is reversed and the drawing is schematic and difficult to read in parts, it is worth comparing with the painting.8 Both are composed with a clear divide between the divine and the mundane, Christ and his apostles on one side, distinct from the other figures. Both Christ’s pose and that of the apostle directly behind him are close to those in the painting. By contrast, the gathering opposite is less resolved, and – following iconographic tradition – is not a single family, but a group of children and adults. Among them, a child blessed by Christ, holds his hands together as if in prayer, while his mother supports him; another, later hidden as the artist reworked the drawing, looked out at us. Another mother cradles her child and directly behind her, an amendment adds a bearded man who reaches forward to offer a comforting touch on a child’s shoulder, so creating a coherent family group. In this respect, it is worth noting that the original group comprised only women and children, rather like the Van Noort painting mentioned previously. Although Van Noort’s work is not dated precisely, it may well have inspired Van Dyck as there are many similarities between it and the drawing; for instance, both show Christ to the right and the children to the left, which would be reversed in Van Dyck’s canvas.
The elements described here find an echo in the painting; clearly Van Dyck was intent on creating stronger formal and psychological relationships among the figures. In the finished work, this results in a focus on one family, but we would note that at the underpainting stage, a third adult was present. It seems clear that Van Dyck was already exploring the subject – and possible compositions – before the commission for a family portrait arrived. He must have proposed it to his client, who was willing to accept this novel idea.
Like Rubens, Van Dyck employed painted “tronies” or study heads, which could be inserted into finished compositions. We see their use in the figures of Christ and the apostles. Two of the study heads survive, used for the two apostles to the left of Christ;9 those for Christ and the third apostle can be observed in other paintings from the 1610s.10 Such figure types met viewers’ expectations for the apostles – expectations formed by artists – and, in the case of Christ, may have recalled an “authentic” likeness – in this instance, a relic thought to be an accurate portrait.
The tronie employed for the apostle at Christ’s shoulder (illus. 3) demonstrates one distinctive feature of Van Dyck’s use of them – perhaps stemming from the fact he would delegate less to assistants, so that the tronie need not be so fully finished. The canvas shows two views of the same man, both painted brusquely and, in places, very thinly.11 Overall, it is rather schematic – very different from the tronie by Rubens included in this exhibition, which is more polished and detailed, and with a more descriptive palette and touch. Van Dyck’s abbreviated and stylized approach here imposes limits on the final painting: the result lacks a sense of veracity. In other cases, tronies are elaborated and adapted for the needs of the final composition in ways that transcend the limitations of the studies.12
For the family group, Van Dyck also relied upon painted studies for at least the figures of the father, the boy being blessed and the child in the foreground who looks at us. Others have not survived, possibly due to the fragility of the support: the studies of the children are painted on paper, while the father is on a very fine canvas, perhaps a scrap of domestic linen.
The Study of A Boy’s Head and Hands is the only one broadly accepted as Van Dyck; the other two have not yet been publically exhibited. All these studies are painted directly, share a characteristic touch, and a shared approach to modelling – for instance, compare the treatment of the eyes and brows. They are done mostly wet-in-wet with narrow brushes, and the paint is often dabbed or flicked across the surface to avoid it becoming muddy through blending. Recent examination in preparation for the exhibition allows confirmation of their authorship.13
The Study of a Child’s Head was painted rapidly and economically – likely necessary to quickly capture a child’s likeness. The consequent lack of information necessitated some improvisation as it was worked up in the Ottawa painting. To some degree, the study has been used simply to map volume and highlight placement with the result that the child in the painting appears fuller faced and a perhaps couple of years older.
Conversely, the Portrait of a Man, a study for the father, contains a great deal of information. Fully worked up, it has the appearance of a portrait, including a painted background and convincing clothing. The head has been transferred to the painting with a great degree of accuracy, but with slight embellishment of his hair, moustache and beard.
The family group has every appearance of veracity, both in terms of likeness and character – and this is reinforced here by comparison with Christ and his apostles. Both groups – the divine and the human – use the painted study to different ends, the results guided by the conventions for representing the two.
Close examination of the Ottawa painting has clarified an unusual aspect of the work’s evolution. In the initial conception, the painted area was only to have taken up approximately eighty percent of the available stretched canvas.
The fabric support for the painting is a single piece of herringbone weave linen; Flanders was an important centre of textile production. Herringbone, designated a “fancy” weave, was frequently available in widths wider than the standard. The present size of the original fabric is 131.5 × 198 cm, although the edges are somewhat uneven, since all sides have been trimmed. There is strong cusping at the top and bottom of the painting, and clearly the current height is close to the maximum width of available fabrics.
It was previously thought that the canvas had been stretched onto different wooden supports during painting, after it was removed from a priming loom. Rather, tacking holes within the composition seem to be the result of later reformatting, and the making of this painting more straightforward – the canvas stayed on its priming loom until it was completed.
The canvas was stretched onto a flat-faced four-member wooden support, and the nature of the stretching can also be deduced – from the ragged uneven edge of the priming application, and the evidence that it was applied up to a ruck in the fabric, we know there was no clean tacking edge to prime over, and it seems that the fabric was attached at its very edges to the face of the priming loom with iron tacks – this from iron staining at two surviving original points of attachment.
The priming was applied in two layers: the first comprised calcium carbonate in oil, presumably to muffle the prominent weave with an inexpensive aggregate; the second of lead white and a finely ground charcoal black to produce a blue-hued mid grey.14 This second layer reads well in X-radiography, and we can see that it was applied, as was often done, with something like a plasterer’s flat or trowel: there are large application sweeps where the tip or fault at the edge of the tool removed the denser priming or scratched the layer below; at the edges there are areas of texture made from the tiny peaks produced by lifting the flat away from the tacky priming. The X-radiograph also gives us the information stated above about the nature of the priming loom – the trowel ran against the inner edge of the loom members and dis-levels in the joinery as the priming was applied, and these features caused less priming to be applied there. Overall, from the pattern of the application arcs, it would appear the canvas was primed upright in bold, practiced sweeps, and the edges left rather ragged and uneven – this unevenness, and the loom-edge feature, can be seen at all edges, indicating the painting has not been cut down.
Infrared reflectography (IRR) does not detect any dry underdrawing; the mid-tone priming would have permitted the use of pale chalk to set down initial locations of form, and this would be undetectable. There are reserves at many locations, however, that reveal the roughing in of form with a translucent brown paint, and it may be that the majority of the “planning” or drawing was applied freehand with this paint. In some areas, this paint reads more clearly in IRR, where it seems to contain more black for added force in shadows. The use of tronies may have freed the artist to work in this bold way. There are no significant pentiments, aside from the suppression of a woman’s head between the mother and father.
In the initial conception, the plan was to use only a portion of the entire primed surface. However, the remaining space would not be large enough to allow for re-use, so logically the decision was for pictorial reasons rather than simple economy.
In the original plan, the height of the composition was from the top of the canvas to a point just below the hands of the boy being blessed – the darker roughing in of the drapery of the foreground apostle can be seen showing through the upper paint layers, and indicates the artist set a lower limit to the composition with a straight line. The leftmost limit of the composition is unclear at that stage, but would have been about 20 cm in from the present edge. As the figures took shape and volume however, it appears the paint surface was expanded in two stages to ultimately fill the entire primed surface of the canvas. In the first expansion it seems the artist roughed the painting out to the extreme edges, then made use of the slight disruption in the priming at the edge of the loom as a convenient vertical and horizontal limit to crop the painting slightly – even drawing this limit in dark paint where it had become less clear in the thicker paint of the golden draperies. Eventually he transgressed these edges as the painting was finished, and it appears the ultimate, intended composition is that which we see today.
In respect of finish, the stages of expansion are visible today as a result of changes in the paint and the effects of overcleaning; they would have been much less obvious. The interesting exception can be seen on the column at left however, where opaque finishing paint and highlighting end at the loom-feature crop edge. Clearly the artist did not think refinishing was worthwhile, since the composition worked visually. A further point to note is the residual female head placed between the mother and father – presumably that of a maid or nanny – is now visible because of change within and damage to a smalt-containing upper layer, and the presence of overpaint and staining. This head would have been fully painted out, the area reading as sky. Each of the three adult heads have reserves in the underpaint of the sky, indicating that Van Dyck initially intended to paint all three – as the roughing in progressed he simplified the grouping.
The traditional reading of this painting as a family portrait set into a biblical scene is well established. The circumstances of its commission will remain unclear until documentary evidence is found to identify the family. The repeated changes in size of the painting indicate there were no stipulated measurements – often considered essential when commissioning a work – but the ultimate increase is only an additional twenty percent.
When the painting was purchased in 1937 there was no record of any recent restoration or conservation. The painting came with a single glue-paste lining on a later 19th-century stretcher. The Gallery conducted no treatment until 1975, when it was cleaned of water-removable surface dirt then given a new coat of varnish over the existing varnish; some minor inpainting was carried out. In 1997 through 1998, the painting was comprehensively cleaned and restored.
Thomas Tollemache (c. 1651–1694); perhaps his sale, London, 27 March 1713; bought jointly £255 by John Churchill, 1st Duke of Marlborough (1650–1722) and Sarah, Duchess of Marlborough (1661–1744), “The Family of Rubens done by Vandike of Ten figures, bought at Mr. Talmashes Auction” according to a ms. inventory dated 1718 (British Library, Add. Ms. 9125, Blenheim Jewels and Pictures, fol. 6v.); by descent; Blenheim Palace sale, Christie’s, London, 24 July 1886, lot no. 64 as by Rubens, although announced at the sale as likely by Van Dyck (The Times, London, 26 July, p. 12); bought £840 by Charles Fairfax Murray for Bertram Woodhouse Currie (1827–1896); by descent to Bertram Francis George Currie, his grandson (1899–1959); sold Christie’s, London, 16 April 1937, lot no. 125 as “Sir A. VanDyck”; bought £1,050 by Heather for Bottenwieser, London; with Asscher & Welker, London; bought 1937, £3,000.
Tollemache’s sale was announced in the Post Boy, the week of 21–24 March and again on 24–26 March 1713 as “On Friday next the 27th instant, will be sold by Auction at the Pall-Mall Coffee-house in Pall-Mall, the Pictures of Col. Robert Wilkins, Executor to the late Gen. Talmarsh, taken by a Commission of Sequestration out of the High Court of Chancery for the Benefit of the Creditors of the said General. … Catalogues may be had at the Place of Sale.” (No catalogue found.) A variant advertisement, implying earlier sales, is given in the Daily Courant, 25 March 1713, “On Friday next, the 27th Instant, will be sold by Auction, at the Pall-mall Coffee-house in Pall-mall, the remaining Part of the Pictures of the late General Talmarsh, for the Payment of his Creditors …”
A later sale links the names of Tollemache and Roger Sizer, suggesting their collections were purchased abroad in the late 1680s, “To be Sold by AUCTION, At Young's Long Room next Door to the Chymist's at the Foot of the Hay-market, on Wednesday the 17th of this Instant April, Original Pictures by eminent Masters, collected abroad by General Talmarsh and Roger Sizer, Esq; about 40 Years ago, and never before exposed to Sale at Home. …” Daily Post, 13 April 1728. This is perhaps the Roger Sizer or Syzer who was clerk to the PayMaster General of the English army, active in Flanders at this time. See J.C.R. Childs, The Nine Years’ War and the British Army 1688–1697: the Operations in the Low Countries. (Manchester, 1991), p. 62.
For the references see the online database “The art world in Britain 1660 to 1735” at www.artworld.york.ac.uk [accessed 27 March 2013].
It is interesting that the earliest reference to this work, in the 1718 inventory of the Marlborough collection noted above, correctly identified the artist. Van Dyck’s early work would likely have been no more easily identified in the early eighteenth century than later, when the picture was attributed to Rubens and others.
See Waterhouse 1978, pp. 8–9, 25, which clarifies its provenance. At the 1886 Blenheim sale, Murray acted as agent for both Charles Butler and Currie; Butler was the better known and some, but not all, contemporary sources assume Murray bought solely on his behalf. Currie’s memoirs make it clear that he bought the Van Dyck – which he continued to attribute to Rubens – at the Blenheim sale; see Bertram Wodehouse Currie, 1827–1896. Recollections, Letters and Journals. (Roehampton, 1901), vol. 2, p. 179, and the statement in the anonymous Catalogue of the Collection of Works of Art at Minley Manor (privately printed, 1908), i.e. Currie’s collection, p. 7.
Wentworth Hubert Charles Beaumont, 3rd Viscount Allendale (1922–2002); sold Christie’s, London, 30 June 1961, lot no. 75 as “Van Dyck,” “Portrait of a Man”; bought £1,500 by Leonard Koetser, London. With Martin Asscher, London by 1970. With Newhouse Galleries, NY; bought Mr. & Mrs. F. Howard Walsh, Fort Worth, Texas; Walsh Family Art Trust; sold Heritage Auction Galleries, Dallas, 10 November 2006, lot no. 25009 as Van Dyck, “Portrait of a Man (Frans Snyders).” Private collection.
The man is sometimes identified as Frans Snyders, in fact childless. Bode 1887, p. 61, is the first to publish this; the claim may have circulated earlier among the bidders at the 1886 Blenheim sale of Let the Children Come (note the anonymous annotations to that catalogue in Lugt mf. 0-2503, no. 1165).
James, Earl of Arran, later 4th Duke of Hamilton (1658–1712) by 1695, when listed in an inventory as “A Boyes Head with two Hands by Vandyke” (Bodleian Library, Mss. Beckford, Hamilton inventories, A list of the right honourable the earle of Arans pictures at London, 12 August, 1695, no. 15); by descent; Hamilton Palace sale, Christie’s, London, 8 & ff. July 1882, pt. 4 (this lot 8 July), lot no. 1033, “Van Dyck,” “A Youth Praying”; bought £47.5 by Edouard Warneck (1834 – c. 1923/24); his sale, Galerie Georges Petit, Paris, 27–28 May 1926, lot. no. 37, Van Dyck; bought 37,000 French frs by Leo Blumenreich, Berlin. Fritz Hess, Berlin; his sale Cassirer & Fischer, Lucerne, 1 September 1931, lot no. 13, Van Dyck; sold 10,500 Swiss frs. Phillips, London, 20 April 1993, lot no. 47 as workshop. With Philip Mould Ltd., London. Henry Weldon, NY, by 1997. Private collection.
The connection to Let the Children Come is first made in print in Exposition de tableaux de Maîtres anciens au profit des inondés du Midi. (Paris, 1887), no. 35, referring to a copy of this study then also owned by Warneck (now Louvre); see below.
I – Louvre, Paris, inv. R.F. 1961–83.
After Anthony van Dyck
Head of a Young Man
Oil on wood, 41 × 28.5 cm
Provenance : Edouard Warneck (1834 – c. 1923/4); his sale, Galerie Georges Petit, Paris, 27 & ff. May 1926, lot no. 36; bought 41,000 French francs by the Président Charles d’Heucqueville, Paris; his sale, Paris 24–25 March 1936, lot no. 88; bought Schoeller for Victor Lyon (1878–1963); gift of Hélène and Victor Lyon, 1961, as attributed to Van Dyck.
Selected References : Burchard 1938, as one of two autograph studies, their exact priority left unexplored; Waterhouse 1978, figs. 8–9 as a first study, subsequently revisited in the picture exhibited here; De Poorter in Barnes et al. 2004, under no. 1.15 as an inferior version of the work exhibited here; Foucart 2009, p. 135 as a copy; Pérez Preciado in Madrid 2012–13, p. 194, note 5, as a copy, summarizing earlier attempts to relate this and the original.
II – Unknown Location.
Oil on wood. Measurements unknown.
Known from a photograph in the Witt Library, box no. 742 (Sir A. van Dyck).
Selected References : McNairn in Laskin & Pantazzi 1987, p. 100 as a third study; Pérez Preciado in Madrid 2012–13, p. 194, note 1, as a copy.
1 Waterhouse 1978, p. 8. Bode was among the most influential scholars of his generation and his opinions, although often controversial, carried weight. Note that his attribution of the Christ exhibited here to Rubens himself, and not to his workshop, was an important factor in the Gallery’s purchase of that work.
2 Notably, Rooses, the author of an important catalogue of Rubens’ works, rejected it. Waterhouse 1978, pp. 9, 25 notes that at the 1937 sale there was some doubt among the art trade, and that staff at the National Gallery, London – worried by the apparent “split” between the left and right halves, so different in style – suggested it was collaborative, perhaps painted by van Dyck and Jordaens. Burchard 1938 resolved the issues and has remained the foundation for research.
3 The best introductions are Nora de Poorter’s catalogue of the artist’s early work in Barnes et al., 2004 and Madrid 2012–13.
4 It is unclear when he set up an independent studio and began to sell works on his own; technically he could only do this after he registered with the Antwerp painters’ guild, which he did in 1618. The evidence – much of it circumstantial and contradictory – suggests he ran his own studio only around 1618–21, see Lammertse 2002. Yet this was also the period of his closest collaboration with Rubens.
5 Judging from a survey of the material published in Duverger 1984–2004.
6 Princeton 1979, no. 21; Gordenker 2001, pp. 41–42.
7 See Pérez Preciado in Madrid 2012–13, no. 38 for a succinct analysis of Van Dyck’s sources.
8 Vey 1962, no. 32. While this is the only drawing related to the painting to survive, judging from his practice at this time, Van Dyck would likely have explored the composition in others, ranging from quick, abbreviated sketches to more finished drawings, perhaps supplemented by studies of individual figures; in this spectrum, the drawing discussed is relatively well-developed. Note that it includes colour notations, quickly applied red and blue: Christ wears, as in the painting, a red garment and a blue cloak (admittedly, the boundaries of the two are not clear and the choice of colours is not uncommon); the child to the lower left a red garment.
9 De Poorter in Barnes et al. 2004, nos. I.90 and I.91 and Madrid 2012–13 nos. 35 and 36.
10 The figure types appear in a set of Christ and the Apostles by Van Dyck; see the chart in de Poorter (in Barnes et al. 2004), pp. 68–69. The head of Christ appears in other works from this period: De Poorter in Barnes et al. 2004, no. I.14, argues that the same profile view of Christ appears in a doubtful piece, Christ and the Penitent Sinners, (Bayerische Staatsgemäldesammlungen, Munich, her no. I.A6). And compare the physiognomy of Christ in The Healing of the Paralytic (Munich, Alte Pinakothek), itself perhaps based on Rubens prototypes, see de Poorter in Barnes et al. 2004, no. I.10. Tronies do not need to be after the life: Pérez Preciado in Madrid 2012–13, p. 198, notes that this type of Christ derives from an older painting claimed to be an authentic likeness; Van Dyck likely knew it through a copy by Rubens. It is unclear whether Van Dyck sought a model who resembled the image, or whether he relied upon the prototype, and adapted it.
11 See Madrid 2012–13, no. 35.
12 See for instance Madrid 2012–13, no. 36: a double-headed tronie on paper, the left side head is used in Let the Children Come; the right as St. Philip in Van Dyck’s Apostles series c. 1618–20. When employed to paint the apostle, his features are softened, the modelling less rugged, and subtle changes made to turn a man into a saint. This was undoubtebly driven by him being the sole focus of the work, and not a detail.
13 The attribution of studies made after the life by Van Dyck – including both tronies destined to be incorporated into other works, and portrait studies made as aids in painting larger portraits – can be controversial. For the latter type see Barnes et al. 2004, no. II.47, a study of Elena Grimaldi; a pentiment clearly links it to the final portrait, yet its attribution has been doubted due to its restrained handling and perhaps its condition. As in the Ottawa painting, it seems clear that the study was part of Van Dyck’s carefully planned preparatory work. Relying upon studies might not have been ideal, but they could be used if the subject was unwilling or unable to sit, or due to some other constraint; for an example of Rubens taking studies as a solution to dealing with a busy sitter and the lack of a proper canvas, see Ruelens and Rooses 1887–1909, vol. 2, pp. 250–251.
14 Most information about paint given here was determined analytically by Marie-Claude Corbeil and Kate Helwig of the Canadian Conservation Institute, Ottawa, in 1998.
Adam van Noort, Suffer Little Children to Come Unto Me, oil on canvas, 111 × 135 cm. Koninklijk Instituut voor het Kunstpatrimonium, Brussels (Inv. 4703). Photo: KIK-IRPA, Brussels. www.kikirpa.be
Anthony van Dyck, Suffer Little Children to Come unto Me, c. 1618, pen and ink with wash and red chalk on paper, 17.7 × 19.5 cm. Musée d’Angers (inv. MTC 4993). Photo : Pierre David
Anthony van Dyck, Double Head Study of a Bearded Man, c. 1618–20 oil on canvas, 45 × 67 cm. Private Collection. Photo: John David Allison
Anthony van Dyck, Study of a Boy’s Head and Hands, c. 1618–20, oil on paper, later mounted to canvas, 43 × 28.9 cm. Private collection
Anthony van Dyck, Study of a Child’s Head, c. 1618–20, oil on paper, later mounted to wood, 31.4 × 22.9 cm. Courtesy of Philip Mould and Company, London