1. Background and early work in Antwerp and London, to autumn 1621.

Anthony van Dyck was the seventh child of Frans van Dyck and Maria Cuypers [Cuperis], who lived at ‘Den Berendans’, a substantial house on the Grote Markt in Antwerp. His father was something of a painter as well as a prominent silk and linen merchant. His mother, who died when he was eight, was known for the beauty of her embroidery. In October 1609 the young Anthony was registered with the Antwerp Guild of St Luke as a pupil of the dean that year, Hendrik van Balen I, a competent painter of small figures who often collaborated with the landscapist Jan Breughel I. He himself was registered as a master on 11 February 1618, and four days later, by his father’s consent, he was declared of age by the Antwerp Tribunal. He had earlier, while still a minor, been involved in legal proceedings, on 3 December 1616 and 13 September 1617, on behalf of himself, his four sisters and his younger brother (his siblings all followed religious callings), to receive part payment of their grandmother’s estate. This was incumbent because his father, formerly well-to-do and respected, the President of the Confraternity of the Holy Sacrament at Antwerp Cathedral, soon after became bankrupt. In order that Anthony’s earnings should not be seized to help pay his father’s debts, in 1620 he set up on his own, renting a house large enough for a studio with assistants, the ‘Dom van Keulen’ in the Lange Minderbroederstraat. There he engaged Herman Servaes and Justus van Egmont to make copies of his several sets of Christ and the Apostles (dispersed). In this, van Dyck followed the example of Rubens, who had assistants to copy (c. 1609–10) the prototype apostelado sent to the Duque de Lerma.

Van Dyck’s series of larger compositions of the period 1616–21, both sacred and profane, were impressively of his own invention: but financial stress, and perhaps the difficulty of placing his early works except on commission, directed his precocious brilliance rather more than he might have wished to a series of three-quarter-length portraits of burghers and their wives (dated examples, 1618; Vaduz, Samml. Liechtenstein, and Dresden, Gemäldegal. Alte Meister). These were evidently inspired by contemporary work by Rubens of this kind, notably the portrait of Jan Vermoelen (1616; Vaduz, Samml. Liechtenstein).

It is not known precisely when or on what terms van Dyck entered the studio of Rubens. The older master, 22 years his senior, must have spotted van Dyck’s potential soon after he himself was established in the mansion he had built (1610–17) on the Meer. Van Dyck as a boy—probably while still in van Balen’s studio—seems to have been uniquely privileged in access to the travelling Pocket-Book of Rubens (surviving in part, and through 17th-century copies), which was later celebrated by Bellori. He plundered its contents extensively, methodically and on occasion wittily into the first of his known sketchbooks, the so-called Antwerp Sketchbook (Chatsworth, Derbys), in which he pocketed a wide range of material, from Giacomo della Porta’s physiognomic comparisons and Serlio’s 4th Book of Architecture to copies of paintings by Titian and recipes for sore eyes and for painters’ materials; the sketchbook also records figures or groups by subject categories from all manner of Italian and northern engravings such as were found in abundance in Rubens’s house. Significantly, there are neither portraits nor records of portraits. Van Dyck’s ambition throughout his working life was to be a history painter.

Writing on 24 April 1618 to Sir Dudley Carleton, English ambassador at The Hague, Rubens offered a painting of Achilles and the Daughters of Lycomedes (Madrid, Prado), valued at 600 guilders. This was ‘fatto del meglior mio discipolo’, then gone over by Rubens himself who was bargaining to part with ‘the flower of my stock’ in exchange for antiquities that Carleton had brought from his previous post, Venice. On the visual evidence of the picture itself, the ‘best pupil’ must have been van Dyck. As Bellori recorded, van Dyck had already served Rubens c. 1616–18 in the preparation of cartoons for the Decius Mus tapestry cycle, which were sent to Brussels to be woven by May 1618. In this project van Dyck’s role was that of chief assistant, transferring the highly complex designs of Rubens’s oil sketches on to canvases of heroic dimensions, painted especially to instruct the draughtsmen regularly employed on preparatory cartoons. Van Dyck also, according to Bellori and Mariette, drew modelletti for engravings intended to register and advertise afar the prowess of Rubens in design.

By 24 March 1620, in Rubens’s contract with the Jesuit Provincial Father Tiry to design 39 ceiling paintings for the aisles and galleries of the Society’s new church in Antwerp, the execution was allowed to be by van Dyck and other unnamed studio assistants. This contract also stipulated that van Dyck might make a painting for one of the four side altars at a later time. Van Dyck’s early assistance to Rubens in painting figures can already be discerned in such religious works as the Virgin and Child with Four Great Penitents (c. 1618; Karlsruhe, Staatl. Ksthalle) and Jesus in the House of Simon the Pharisee (St Petersburg, Hermitage). His blazing idiosyncracies of morphology and brushwork blended with, but were never wholly absorbed by, the more profoundly creative genius of Rubens, who kept for himself eight of van Dyck’s early masterpieces and with whom van Dyck apparently always continued to be on good terms. A letter from Antwerp of 17 July 1620 to Thomas Howard, 2nd Earl of Arundel, written by Francesco Vercellini, his Venetian gentiluomo then accompanying Lady Arundel on a continental visit, relates that:Van Dyck is still with Signor Rubens, and his works are hardly less esteemed than those of his master. He is a young man of 21 years, with his very wealthy father and mother living in this town, so that it will be difficult to get him to leave these parts, especially since he sees the good fortune enjoyed by Rubens.

In fact, van Dyck’s mother was long dead; and his parents had fallen in fortune. Despite Vercellini’s pessimism, ‘the father of virtù in England’ (i.e. Arundel) was anxious to bring this rising star to London.

On 20 October Thomas Locke wrote to William Trumbull from London: ‘Van Dyck is newly come to town…. I am told my Lo: of Purbeck sent for him hither’. John Villiers, recently created Viscount Purbeck, was half-brother to the royal favourite, the Marquess of Buckingham. On 25 November Tobie Mathew wrote to Carleton at The Hague that King James I had granted van Dyck, Rubens’s ‘famous allievo’, an annual pension of £100. On 16 February 1621 van Dyck received £100 ‘by way of reward for special service performed for his Matie’. What precise service to his Majesty this may have been remains obscure.

Of van Dyck’s paintings, Buckingham acquired the Continence of Scipio (1620–21; Oxford, Christ Church) from Arundel for York House and the Venus and Adonis (1620–21; Harari & Johns, London, 1991, see 1990–91 exh. cat., no. 17), presumably for another of his residences. The latter relates to one of van Dyck’s Antwerp mythologies: it is an allegorical portrait painted in haste to celebrate the flamboyant Buckingham’s betrothal to Lady Katherine Manners, a match regarded by her father as a mésalliance. The first of these two paintings, with full-length figures, evinces van Dyck’s youthful admiration for the scenographic taste of Paolo Veronese whose Esther and Ahasuerus (1556; Vienna, Ksthist. Mus.) was then at York House; the second his facility in adapting to his purpose both Dürer’s engraving of The Promenade (b. 94) and Cesare Ripa’s emblem into a sensuous combination of the learned and the louche, hardly suitable for the burghers of Antwerp, welcome perhaps only at the Stuart court. Of the noble Arundel, van Dyck painted an appropriately sober portrait seated half-length (1620–21; Malibu, CA, Getty Mus.). It was Arundel who on 28 February 1621 signed a travel pass and permit for him for eight months. No other paintings are known of those van Dyck might have done during this first brief stay in England, unless it were one of the self-portraits (see §IV below).

Van Dyck returned to Antwerp in March 1621 for just under eight months to order his family affairs and to prepare for Italy. Usually assigned to this phase of his activity are the portrait of Susanna Fourment and her Daughter and that of Isabella Brant (both Washington, DC, N.G.A.), wonderful presages of the fashion for full-length portraiture that he later developed to gratify the Genoese families. Also datable to this period are the superb St Sebastian Bound for Martyrdom (Edinburgh, N.G.), the Crowning with Thorns and the Betrayal of Christ (versions of both, Madrid, Prado), masterpieces that manifest his urge to be recognized, like Rubens, principally as a composer of altarpieces and other histories at least on the scale of life.

2. Italy and France, autumn 1621–autumn 1627.

Anthony van Dyck: Lucas van Uffel (died 1637), oil on…In October 1621 van Dyck followed Rubens’s example in 1600 by leaving for Italy. It is not clear how long he intended to remain there, nor whether he would have accepted, as Rubens had, princely service at this stage of his career. He arrived in Genoa, the mercantile equivalent of Antwerp, probably by late November, lodging with his countrymen Cornelis and Lucas de Wael, who were established as art dealers and painters of small-figure genre and battle scenes. Van Dyck came unprepared for the sight in the Spinola, Doria and Grimaldi palaces of the astounding advances in fashionable portraiture made by Rubens during his Mantuan service 15 years before; and there can be little doubt that he conceived his standing portrait of Marchesa Elena Grimaldi Cattaneo and his equestrian portrait of Giovanni Paolo Balbi (c. 1625; Parma, Corte Mamiano Found.) to challenge Rubens’s portraits of Brigida Spinola Doria (1606; Washington, DC, N.G.A.) and Gian Carlo Doria (c. 1606–7; Florence, Pal. Vecchio). (For van Dyck’s portrait of the so-called ‘Marchesa Balbi’, c. 1621–2, Washington, DC, N.G.A..) Between November 1621 and February 1622 he painted the portrait mentioned by Bellori of the future Doge Agostino Pallavicini (Malibu, CA, Getty Mus.), voluminously robed in crimson brocade as ambassador from Genoa to the Holy See. This was most likely the first of his grand portraits of Genoese notables. From February 1622 he seems to have been in Rome, filling his Italian Sketchbook (London, BM) with observations from life and from the works that took his eye while visiting churches and collections. He sketched (fol. 62r and v) and then he painted resplendent full-length portraits of Sir Robert Shirley and Teresia, Lady Shirley (both Petworth House, W. Sussex, NT), gorgeous in Circassian dress during their diplomatic mission to Gregory XV between 22 July and 29 August, and half-length portraits of two northern European sculptors, François Du Quesnoy (Brussels, Mus. A. Anc.) and Georg Petel (Munich, Alte Pin.). He then moved to Venice to meet Lady Arundel and to accompany her in November/December to Mantua and Milan. They reached Turin late in January 1623. He probably stopped again in Genoa and also visited Florence and Bologna. He painted in fairly quick succession, either in Venice or in Genoa, two portraits of the merchant Lucas van Uffel (Brunswick, Herzog Anton Ulrich-Mus.; New York, Met.; see fig.), the latter (New York) tense with a dazzling depiction of movement arrested on the instant.

News of the death of van Dyck’s father on 1 December 1622, leaving him worldly care for his siblings, would have reached him in Rome early in the new year. His immediate reaction to private grief was expressed in his most poignant portrait of himself, the Self-portrait with the Broken Column (early 1623; St Petersburg, Hermitage). (The public epitaph would come some six years later with the altarpiece of the Crucifixion with SS Dominic and Catherine of Siena painted soon after his return to Antwerp, in fulfilment of a promise of his father to the Dominican nuns who had nursed his last illness.)

Between March and October or November 1623 van Dyck was again in Rome. He followed the success of his portraits of van Uffel with two portraits of aristocratic persons just as vivid: the full-length of the former Nuncio in Flanders, Cardinal Guido Bentivoglio (1623; Florence, Pitti), seated alert as though giving audience, a state portrait that proved one of the most memorable Baroque achievements of its kind; and, just before that, the three-quarter-length, more intimate revelation of Principe Virginio Cesarini in Jesuit Garb (St Petersburg, Hermitage), seated in an armchair as though in private disputation, his gaze and his gesture alerting the viewer to the frailty of his existence. Radiographs have revealed that beneath the surface of the paint in the latter van Dyck had sketched the contrapposto he had wanted for the portrait of Cardinal Bentivoglio. He also portrayed another cardinal about this time, the Genoese Domenico Rivarola (best-known version in Des Moines, IA State Educ. Assoc.), either in Rome or in Genoa. Certainly it was in Rome that he met Sir Kenelm Digby (1603–65), the English Resident and a future patron. Digby’s recollections of van Dyck became the principal source of Bellori’s Vita of the artist. Significantly, van Dyck was the only Flemish artist, save Rubens and Du Quesnoy, both also active in Rome, who was esteemed enough by Bellori to be included in his publication of 1672.

From autumn 1623 to spring 1624 van Dyck seems to have been in Genoa. In emulation of Titian’s portraits of Clarissa Strozzi (1542; Berlin, Gemäldegal.) and Ranuccio Farnese (1542; Washington, DC, N.G.A.), he revivified a fashion for portraying young children without their parents. The lively charm of this Genoese series, anticipating his work at the Caroline court by eight years, has never been surpassed, from its beginnings with the portraits of Filippo Cattaneo and Clelia Cattaneo (both late 1623; Washington, DC, N.G.A.) to the Three Sons of Girolamo de’ Franchi (the ‘Balbi Children’, 1625–6; London, N.G.) to the much later portrait of Mary, Princess Royal (1637; ex-Governor Fuller priv. col., Boston, MA), which in 1647 was sent clandestinely to Holland by Charles I when he was detained at Hampton Court.

Van Dyck then moved to Palermo in spring 1624 at the invitation of the Viceroy of Sicily, Emanuel-Philibert of Savoy (1588–1624). Towards mid-May the plague struck, and the city was soon in quarantine. With the discovery of St Rosalia’s purported remains in a grotto on nearby Monte Pellegrino, her cult expanded; and van Dyck executed at least three commissions to represent her as an intercessor (New York, Met.; London, Apsley House; Ponce, Mus. A.). The plague also claimed the life of the Viceroy himself, on 3 August of that year, shortly after van Dyck painted a scintillating three-quarter-length portrait of him (1624; London, Dulwich Pict. Gal.). On 12 July the 25-year-old van Dyck visited the Italian artist Sofonisba Anguissola, then in her nineties. In his Italian Sketchbook he surrounded a drawing of this relict (fol. 110r) with notes of their conversation about portraiture. He was also commissioned to paint a large altarpiece, the Madonna of the Rosary, for the oratory of the Rosario (1624–7; in situ). This, his principal religious undertaking in Italy, he began but did not complete before he prudently withdrew to Genoa from a recrudescence of the plague. (After his return to Antwerp, his representative, Antonio della Torre, was paid by the Confraternity of the Rosary on 8 April 1628 for the altarpiece, ‘novamente fatto nella città di Genova’.) From van Dyck’s pen sketch of the composition (Hilversum, Liberna priv. col., see 1991 exh. cat., no. 45) it is evident that he had in mind Rubens’s Madonna della Vallicella Adored by Saints (1606–7; Grenoble, Mus. Grenoble), the first version of the altarpiece that Rubens himself had rejected for the Chiesa Nuova in Rome; he knew it as it hung over Rubens’s mother’s tomb in the abbey of St Michael, Antwerp, where his own younger brother Theodoor held a canonry.

In July 1625 van Dyck reputedly made an excursion to Marseille in order to visit, at Aix-en-Provence or Belgentier, the savant who had become the admiring friend and correspondent of Rubens, Nicolas-Claude Fabri de Peiresc. If so, van Dyck is likely to have drawn his portrait. A likeness was eventually engraved in Antwerp for The Iconography (m.-h. 89). On his return from Provence, he exulted in painting a dazzling series of portraits, many full-length, of men, women and children of the leading Genoese families (e.g. Genoese Noblewoman and her Son, c. 1626; Washington, DC, N.G.A.). His siblings had to declare to the Antwerp magistrate on 12 December 1625, as his brother-in-law had done on 27 November 1624, that he was still abroad.

3. Antwerp and The Hague, autumn 1627–spring 1632.

Van Dyck returned home in evident haste in autumn 1627 for family as well as professional reasons. His parting gift to his hosts in Genoa, the double portrait of Cornelis and Lucas de Wael (Rome, Mus. Capitolino), is in parts conspicuously unfinished. His sister Cornelia, a Béguine, died in Antwerp on 18 September. Payment was made on 18 December by Giovanni Francesco di Antonio Brignole to ‘Antonio Fiamengo’ for the full-length portraits (all Genoa, Gal. Pal. Rosso) of his son Anton Giulio Brignole-Sale on Horseback, of Paola Adorno, Anton Giulio’s wife, and of Geronima Brignole-Sale with her Daughter Aurelia, Giovanni Francesco’s wife and daughter. Van Dyck’s first dated portraits on return to Antwerp were three-quarter-lengths of appropriately rich sobriety but without palatial settings: that of the collector and connoisseur Peeter Stevens of 1627 and that of Anna Wake of 1628, the year Stevens took her as his wife (both The Hague, Mauritshuis). He painted also in 1628 his first major altarpiece for an Antwerp church, the St Augustine in Ecstasy, which was commissioned by Marius Jansenius for the left aisle of the church of St Augustine, flanking a vast Sacra Conversazione by Rubens and counterbalancing on the right the altarpiece of similar dimensions to his own, the Martyrdom of St Apollonia by Jacob Jordaens (all 1628; Antwerp, St Augustine; on loan to Kon. Mus. S. Kst.). Van Dyck and Jordaens were paid alike, 600 florins, establishing their status as the leading painters after Rubens in their city.

Van Dyck had made a will in Brussels on 6 March 1628. On 27 May the Earl of Carlisle wrote to Buckingham from Brussels that he had not met Rubens at his home in Antwerp but had met him the following day at ‘Monsr Van-digs’. At the Maison Professe in Antwerp, van Dyck was enrolled that same month in the Sodalitedt van de Bejaerde Jongmans, a Jesuit confraternity of bachelors. (He was to paint in 1629 for these devout bachelors the grand altarpiece of the Virgin and Child with SS Rosalia, Peter and Paul, redolent of his experience in Venice and Bologna, and in 1630 a smaller, more intimate and more distinctly personal work of devotion, the Mystic Vision of the Blessed Herman Joseph (both Vienna, Ksthist. Mus.).) In December 1628 he was presented with a gold chain, valued at 750 guilders, for a portrait of the Infanta Isabella Clara Eugenia (version, Turin, Gal. Sabauda), whose court painter, Rubens, was in Madrid on a diplomatic mission to her nephew Philip IV of Spain.

Van Dyck’s eye was already on Rubens’s eventual return to dominate painting in Antwerp and correspondingly on his own chances of a return to the Stuart court, where Charles I, a truly art-loving prince, had succeeded his father James in 1625. As his spectacular introduction to the new reign, van Dyck painted his most delectable masterpiece hitherto on a profane subject, the gorgeous Rinaldo and Armida (see fig.; for a later modello on the same theme see [not available online]), and on 5 December 1629 he wrote to Endymion Porter, whom he had met nine years earlier through Buckingham, that the painting had been passed to his agent for delivery to the King; the following March van Dyck received £72 for it. By then he had more than sufficient financial security; and on 20 March 1630, Antwerp having issued a loan of 100,000 guilders, he subscribed for 4800 guilders. In a proxy dated 27 May, written for Pieter Snayers, who was to figure among the painters whose likeness (Munich, Alte Pin.) was engraved for The Iconography (m.-h. 98), van Dyck described himself as ‘schilder van Heure Hoocheyd’ (‘painter to Her Majesty’), the Infanta Isabella. His annual salary was 250 guilders; and, like Rubens before him, he was excused residence at her court in Brussels, continuing to live in the artistic centre Antwerp. In December an Antwerp restorer, J.-B. Bruno, chanced to remark on van Dyck’s exquisite collection of paintings (see Wood; Brown and Ramsay).

On 10 May 1631 van Dyck stood godfather for Antonia, daughter of Lucas Vorsterman the elder (who, with his star pupil Paulus Pontius and with Schelte Bolswert, was to be one of the chief engravers of The Iconography; for van Dyck’s own etched portrait of Vorsterman see [not available online]). During 1631, at the acme of his powers, van Dyck painted the stupendous full-length portrait of Marie de Raet, the wife of Philippe Le Roy, whose own full-length portrait dates from the previous year (both London, Wallace). From 4 September to 16 October 1631 Maria de’ Medici, exiled from France with her younger son, Gaston, Duc d’Orléans, was in Antwerp, and van Dyck painted their portraits full-length (Bordeaux, Mus. B.-A., and Chantilly, Mus. Condé, respectively). The Queen Mother’s secretary, J. P. de La Serre, noted in his travel record his admiration for van Dyck’s ‘Cabinet de Titien’. Titian, almost from van Dyck’s first sightseeing tours south of the Alps, had largely supplanted Veronese in his utmost admiration, as is evident from his altarpiece of the Crucifixion with SS Francis and Bernardino and a Donor (c. 1620) for the parish church of S Michele di Pagana, near Rapallo in Liguria (in situ). His Italian Sketchbook is crammed with ricordi penned from Titian’s portraits and compositions. Among the nineteen works of Titian that he owned (he copied four more) were the Perseus and Andromeda (c. 1555; London, Wallace) and the Vendramin Family (c. 1543–7; London, N.G.), in which the boys seated on the altar steps with their pet dog surely prompted the purely domestic group of the Children of Agostino and Vittoria Spinola (c. 1623–5; Genoa, Pal. Durazzo-Pallavicini), itself long acclaimed as a foretaste of the Children of Charles I and Henrietta Maria (1635; Windsor Castle, Berks, Royal Col.).

Between 6 and 16 December 1631 Sir Balthazar Gerbier, the British agent in Brussels who had been Buckingham’s master of the horse, sent Charles I’s treasurer, Lord Weston, van Dyck’s Virgin and Child with St Catherine (London, Buckingham Pal., Royal Col.) as a New Year’s gift to the King. Perhaps out of pique at this high-handed transaction, van Dyck himself wrote to Geldorp that this painting was but a copy. Gerbier wrote to the King on 13 March 1632 that Rubens considered it an original and that the vendor, Salomon Noveliers, confirmed this before a notary; van Dyck himself was in Brussels, planning to travel to London.

Despite any such intention, the painter went for the winter of 1631–2 northward to The Hague, where he painted subject pictures as well as portraits for both the court of Frederick Henry and Amalia von Solms and the court of the deposed ‘Winter King and Queen’ of Bohemia, Frederick V of the Palatinate and Elizabeth Stuart. For the individual likenesses of the Palatine princelings, he adapted the pattern invented by Titian for his portrait of Benedetto Varchi (c. 1540–43; Vienna, Ksthist. Mus.). He portrayed in addition the poet and statesman Constantijn Huygens the elder, who entered in his diary for 28 January 1632, ‘Pingor a Van Dyckio …’. The appearance of the portrait (untraced) is presumably recorded in Pontius’s engraving in The Iconography (m.-h. 53), the series of engravings for which Huygens himself wrote three mottoes.

4. England, spring 1632–late 1633 or early 1634.

Not until April 1632 was van Dyck in London again. From that time Edward Norgate, an artist and writer in the service of Arundel and brother-in-law of Nicholas Lanier, the court musician who had presented his own portrait by van Dyck (1628; Vienna, Ksthist. Mus.) to the King, had 15 shillings per diem ‘for dyett and lodging’ of van Dyck. The Flemish painter moved soon to Blackfriars, London, beyond the jurisdiction of the jealous Painter-Stainers’ Company. He also had lodgings at Eltham in Kent, where the King kept his summer residence. On 5 July 1632 he was knighted and made ‘principalle Paynter in ordinary to their Majesties’, thus eclipsing Daniel Mijtens who had been portrait painter to the Stuart court since 1625. His ‘greate Peece’ displaying the royal family settled at home (1632; London, Buckingham Pal., Royal Col.), to be hung at a vantage-point in the Long Gallery at Whitehall, was decorously more staid, it must be said, than the vivid grouping of The Lomellini Family , which he had finished in Genoa. In the still greater, but now so sadly damaged, group portrait of Philip Herbert, 4th Earl of Pembroke, and his Family (c. 1633–4; Wilton House, Wilts), van Dyck, in dealing with so many individual figures, would seem to have overextended his powers of coherent composition in depth. More compact, more coherent and effective in grouping and panoply is the portrait of Johan Maurits of Nassau-Siegen and his Family (1634; Firle Place, E. Sussex), in which the count and the countess are shown seated with their four children standing beside them.

Van Dyck was kept busy on portraiture and on the restoration (in one case the replacement) of Titian’s series of Twelve Roman Emperors (1536–40; destr. 1734), purchased by Charles I from Mantua. Ten commissions from the King were paid at £280 on 8 August 1632, and on 7 May 1633 nine portraits of the King and Queen at £444; in April 1633 he had, in addition, received the gold chain and medal that he is later seen wearing in the enigmatic Self-portrait with a Sunflower (c. 1635–6; best-known, but doubtfully autograph version, London, Duke of Westminster priv. col.); and on 17 October 1633 he was promised a salary of £200 a year, besides payment for special work. A few days later he had a £40 advance for a portrait of the Queen. During 1632 he had painted the half-length of Henrietta Maria Handing Charles I a Laurel Wreath (Kroměříž Castle), an eloquent criticism of Mijtens’s effort at the same courtly composition (before 1632, overpainted 1634; London, Buckingham Pal., Royal Col.); also the portrait of Philip, 4th Lord Wharton (Washington, DC, N.G.A.). In a census of foreign residents at Blackfriars, he was described as ‘Dutch. Sir Anthony Van Dike. Portrait painter. Two years. Six servants’.

5. Antwerp and Brussels, early 1634–early 1635.

Henrietta Maria in the last week of August 1633 had tried to place Theodoor van Dyck, Anthony’s younger brother, as her chaplain. The canon can only have stayed a little while in London; by 14 March 1634 he was back in Antwerp. On private business thither the painter had perhaps even preceded him, for on 28 March he acquired an interest in Het Steen at Elewijt, the estate that Rubens was to buy in May 1635. On 14 April Anthony authorized his sister in Brussels to administer his property. At this point he had little intention of settling to enjoy country life. He was elected on 18 October dean of the Antwerp Guild of St Luke honoris causa, an honour previously vouchsafed only to Rubens. He painted a noble version of the Pietà (Munich, Alte Pin.). Then on 4 November, the Infanta Isabella having died in Brussels the previous December, Philip IV’s brother, the Cardinal-Infante Ferdinand, made his entry into Brussels as the new governor for Spain. The three-quarter-length portrait of him (Madrid, Prado) by van Dyck is noted on 16 December as ‘recently painted’, the painter being then lodged in Brussels in the house ‘In’t Paradijs’ behind the town hall. During this 12-month stay in Flanders, he painted some of the most impressive male portraits of his career: that of his friend Jacomo de Cachiopin (Vienna, Ksthist. Mus.), an inheritor of mercantile wealth who devoted a whole room in the country house he built for himself to portraits by van Dyck, including those of himself and his wife (untraced); a full-length of ineffable elegance, the Abbé Cesare Alessandro Scaglia (Hackwood Park, Hants), the worldly wise, world weary envoy of Savoy, who had returned jobless to Brussels; and two superb equestrian portraits, one of the Spanish general, the Marqués Francesco de Moncada (Paris, Louvre), the other of Prince Thomas-Francis of Savoy-Carignan (Turin, Gal. Sabauda), who had served as temporary governor for Spain between the death of the Archduchess and the arrival in the Netherlands of the Cardinal-Infante. The portrait of Cachiopin is one of the most penetrating studies of a highly civilized melancholic. It contrasts with a more public appearance calculated by van Dyck in his chalk and wash study to be etched by Vorsterman for The Iconography (m.-h. 75). In the portrait of Moncada van Dyck emulated in flattering the rider the revolutionary pattern invented by Rubens for the equestrian portrait of the Duque de Lerma in Spain (1603; Madrid, Prado), known to the painter through Rubens’s preparatory drawing (Paris, Louvre); a pattern that had guided van Dyck in Genoa for the Anton Giulio Brignole-Sale on Horseback, and which he was to elaborate in the stupendous Charles I with M. de St Antoine (1633; London, Buckingham Pal., Royal Col.). The magnificent conceit of showing Prince Thomas-Francis executing a levada on a wild and rocky eminence was to excite Bernini in his equestrian monument in marble, Constantine the Great (1654–70; Rome, Vatican, Scala Regia), as well as in his ill-fated equestrian statue of Louis XIV (1669–77; subsequently modified into a Marcus Curtius by Girardon; Versailles, Château, Gardens). Whence else came to Urban VIII’s sculptor the crucial idea of a broad cascade of drapery set aslant to show off the Roman Emperor on his rearing horse? Van Dyck in this vision of command was surely matching himself against Rubens’s equestrian portrait of George Villiers, 1st Duke of Buckingham as Lord High Admiral (1625; ex-Osterley Park House, NT, London; destr. 1949).

It is not only in comparison to these public and private triumphs in male portraiture that the portrait of Princess Henrietta of Lorraine (1634; London, Kenwood House), painted in Antwerp with her negro page, seems tame. The relaxation of van Dyck’s imaginative creativity over the preceding decade is clear when this picture is compared to the portrait painted 11 years earlier of the Marchesa Elena Grimaldi Cattaneo (see fig. above) gliding so regally on to an evocation of a noble terrace at Sampierdarena, her complexion protected by a parasol held over her head by her negro page.

6. England, spring 1635–late 1639.

Some sitters, particularly in England, evidently did not fully engage van Dyck’s interest. Yet to England he returned in spring 1635 in the fullness of his powers to paint the portrait of William Laud, Archbishop of Canterbury (c. 1635–7; Cambridge, Fitzwilliam), the only English prelate who could afford his price for a three-quarter-length. For the Stuarts he had to meet the challenge of depicting the Head of Charles I in Three Positions (1635–6; Windsor Castle, Berks, Royal Col.), as a guide for Bernini in Rome to carve a marble bust, and by July the Three Eldest Children of Charles I and Henrietta Maria for Christina of Savoy in Turin. The triple study of the King’s head, inspired in presentation by Lorenzo Lotto’s Portrait of a Jeweller in Three Positions (c. 1530–35; Vienna, Ksthist. Mus.), then believed to have been painted by Titian, caused Bernini to remark, it is reported, on ‘ce visage funeste’; and the sculptor was understandably reluctant when four years later Henrietta Maria wrote to request a bust of herself in the manner of the portrait van Dyck had essayed of her husband. She promised to send models painted by van Dyck (Memphis, TN, Brooks Mus. A.; London, Buckingham Pal., Royal Col.). The Three Eldest Children was commissioned by her for her elder sister in Turin in grateful exchange for portraits of the little Prince and Princess of Savoy. When van Dyck’s picture was ready in the autumn, the Savoyard ambassador in London had to write to his Duke that the Queen had told him that: ‘The King was angry with the painter Vendec for not having put smocks on them, as is the custom with small children, and that she should ask Madame, her sister, to have them painted in’. The King may have felt shame to have his heir depicted as an infant of five and a half not yet breeched. Nonetheless van Dyck continued as his principal painter, and on 23 February 1637 he was paid £1200 by Charles I ‘for Certaine pictures by him delivered for our use’.

It was the year after his return to London that van Dyck completed on Scaglia’s commission the last and most deeply moving of his religious works, the Lamentation (c. 1636; Antwerp, Kon. Mus. S. Kst.), for a chapel of the Antwerp Recollects. Meanwhile the long-term project to which van Dyck attached particular importance, The Iconography, was going ahead; and he wrote on 14 August 1636 to Franciscus Junius, librarian to Lord Arundel, for an appropriately learned tag to inscribe below the portrait of Kenelm Digby (m.-h. 71), engraved by Robert van Voerst (1593–1636). At the same time van Dyck, like Rubens, praised highly Junius’s new treatise De pictura veterum (Amsterdam, 1637).

Highlights of van Dyck’s painted production in the later 1630s were the portrait of two brothers-in-law seated, Thomas Killigrew and William, Lord Crofts (1638; London, Buckingham Pal., Royal Col.), a painting in which he again used the symbolism of the broken column (of fortitude in mourning) as he had done in the Self-portrait of early 1623; and, perhaps a little earlier, the haunting three-quarter-length of a man wracked by anxiety, who has usually and most likely been identified as Sir Thomas Chaloner (c. 1637; St Petersburg, Hermitage). Chaloner was to be among the parliamentary judges who condemned the King to death. By contrast to these sensitive portrayals of shared grief or inner tension, he painted in the same phase the portrait of François Langlois, the engraver, publisher and fellow art dealer, dressed as a Savoyard, fingering a musette (c. 1637–8; Cowdray Park, W. Sussex). This was not only an instance, rare since van Dyck’s Italian years, in which he felt called to convey an instantaneous impression of activity, particularly in the face and hands—indeed for that reason this portrait of ‘Ciartes’ (Langlois’s nickname, derived from his birthplace, Chartres) has sometimes been backdated to those years—but also within its Baroque form it was a startlingly dégagé image. Van Dyck also developed in England a pastoral mode of presentation for the pose, dress and setting of select sitters. His successes in this vein were the portraits of Olivia Porter (c. 1637; London, Syon House, priv. col.), wife of his friend Endymion; the earlier Philip, 4th Lord Wharton (Washington, DC, N.G.A.) aged about 19; and Lord George Stuart (c. 1638; London, N.P.G.). For those no longer in the bloom of youth, he enhanced the formal grandeur of rich clothes and hangings. A prime example is another earlier portrait, that of Robert Rich, 2nd Earl of Warwick (1634; New York, Met.) standing in court attire on the foreshore, the splendour of his stance and person set off by the rocky cliffs, while the navy that he commanded battles in the distance. This glorification consciously outclasses the more staid full-length of the same personage painted two years earlier by Mijtens (1632; London, N. Mar. Mus.). Another example is the double portrait of the Earl of Arundel and his wife, Aletheia Talbot, known as the Madagascar Portrait (1639; London, N.P.G., on dep. Arundel Castle, W. Sussex). The richly dressed Earl and Countess point to Madagascar on a large globe, alluding to a scheme for colonizing the island in which the Earl was then involved.

A memorandum (?of late 1638) by van Dyck to Charles I lists prices for twenty-five paintings not yet paid for as well as five years’ arrears in his £200 salary. He was paid £1603 on 14 December 1638 and £305 on 25 February 1639. The valuation of 14 of the paintings was reduced by the King, who, since van Dyck’s return from Flanders, had been short of revenue. Van Dyck had served his prince not only with two beguiling portrait groups of the royal infants, one of which had vexed his employer, and the study of his head in three positions but also with portraits of Charles I Armoured as a Christian Knight and Charles I in Garter Robes (1636; Windsor Castle, Berks, Royal Col.), as well as with numerous portraits of the Queen, including the enchanting earlier full-length of her in hunting costume with her dwarf Sir Jeffrey Hudson standing at her side (1633; Washington, DC, N.G.A.). But it is perhaps the portrait of the King in hunting costume, Le Roi à la chasse (c. 1635; Paris, Louvre), that represents van Dyck’s supreme tribute to his royal patron.

Anthony van Dyck: Le Roi à la chasse, oil on canvas, 2.75×2.14 m, 1635 (Paris, Musée du Louvre); photo credit: Réunion des Musées Nationaux/Art Resource, NY

7. Final years, 1640–41.

That the King seemed insufficiently grateful and a slow payer was presumably one element in van Dyck’s resolve to leave London, reported early in 1640 by the Countess of Sussex to Ralph Verney. His marriage the previous year to Mary Ruthven, one of the Queen’s ladies, offered him a position in English society that his long association with Margaret Lemon, a tempestuous enchantress, had not. Yet the dawning prospect in his native city that he might succeed Rubens, who was overburdened with commissions and mortally sick with gout, would have been a major element in his calculation. There was, however, a delay from 30 May, when Rubens died in Antwerp, until 23 September, when, once more from Arundel, a pass was obtained for Sir Anthony and Lady van Dyck to visit the Continent. Ten days later the Cardinal-Infante wrote from Ghent to Philip IV that van Dyck was expected on 18 October at the dinner of the Guild of St Luke in Antwerp. Ferdinand wrote again to his brother on 10 November that van Dyck’s pride would not allow him to undertake to complete Rubens’s unfinished paintings for the Torre de la Parada, although he would accept a fresh commission. It is not known what work he chose to do in Flanders in 1640. Probably the last commission executed in London was for the widowed Earl of Southampton, the allegorical portrait of his wife, Rachel de Ruvigny, Countess of Southampton (c. 1638–40; Cambridge, Fitzwilliam), ‘la belle et vertueuse Huguenotte’, triumphant over death. About that Bellori wrote, not quite accurately on the evidence of Kenelm Digby, that she was portrayed as Fortuna. He also executed a portrait of Inigo Jones (c. 1640; see [not available online]).

Van Dyck, having been frustrated two years earlier in his hopes of realizing through the costliest medium of tapestry a grand design to clad annually on St George’s Day the walls of the Banqueting House at Whitehall with four large tapestries of Charles I and the Garter Knights in Procession (the project advanced only so far as a swift chiaroscuro oil sketch; Belvoir Castle, Leics), turned his attention to the commission for history paintings to adorn the Grande Galerie of the Louvre. In hopes of that, he went to Paris at the beginning of December 1640. Evidently he visited the Palais du Luxembourg; and there rapidly in pen and wash, mingling with characteristic wit serious purpose, the made a copy (England, priv. col.), so as to stretch his own compositional sense to an unaccustomed scale, of Rubens’s Betrothal of Maria de’ Medici and Henry IV (c. 1621–5; Paris, Louvre). Despite this sally, Louis XIII summoned, quite unsuitably, Poussin from Italy to execute that coveted commission.

After this second disappointment, van Dyck returned to London to paint a wedding portrait, presumably his last royal portrait, of Princess Mary and William II of Orange Nassau (Amsterdam, Rijksmus.); the children married on 12 May 1641. On 13 August Lady Roxburghe wrote from Richmond to Count John Wolfert van Brederode at The Hague that van Dyck had recovered from a long illness; in ten to twelve days he would bring the promised painting to Holland on his way to Flanders. In October he was reported again in Antwerp, then on 16 November again in Paris. By then he was so ill that he had to postpone portraying ‘Monseigneur le Cardinal’ [Richelieu]; and he requested a pass to England. His wife was pregnant; and on 1 December their daughter Justiniana was born in London. Three days later he revised his will, providing not only for his legitimate family but also for his natural daughter, Maria Teresa. On 9 December he died in his house at Blackfriars. It was Jordaens who succeeded to the primacy of Rubens at Antwerp.

1. Artistic.

The enthusiasm of van Dyck for Venetian portraits, especially those by Titian, witnessed in the Italian Sketchbook not only by ricordi but by the additional list on the last leaf of those known to van Dyck, can hardly be exaggerated. Titian’s double portrait of ‘the French ambassador enditing’ (Monsignor Georges d’Armagnac and his Secretary, late 1530s; Alnwick Castle, Northumb.) was known to him after it had been appropriated by the Earl of Northumberland from the collection of the murdered Buckingham: that picture and knowledge of another then also in England, Sebastiano del Piombo’s portrait of Cardinal Ferry Carondelet and Two Companions (c. 1512; Madrid, Mus. Thyssen-Bornemizza), undoubtedly inspired van Dyck’s portrait of Thomas Wentworth, 1st Earl of Strafford Dictating to Sir Philip Mainwaring (1640; St Osyth Priory, Essex); and the close recession of columns in the picture by Sebastiano inspired the setting of the portrait of Albertus Miraeus (c. 1626–32; Woburn Abbey, Beds). The portrait of Strafford, depicted on the eve of his departure for Ireland, standing full-length in armour, a bâton in his left hand and an Irish wolfhound under his right, was a patent tribute to Titian’s portrait of Charles V with his Hunting Dog (?1532; Madrid, Prado) with an imperial hand resting on his faithful white hound. Van Dyck rarely essayed narrative in portraiture: but when he did, as in the George Gage Being Offered a Marble Statue by a Roman Dealer and his Assistant (1622–3; London, N.G.), his genius took fire from Titian’s portrait of Pope Paul III with his Grandsons (1546; Naples, Capodimonte), of which he had penned a ricordo in his Italian Sketchbook (fol. 108r).

This abounding love for Titian should not allow van Dyck’s admiration for Raphael to be overlooked: for the double portraits of Andrea Navagero and Agostino Beazzano (before April 1516; Rome, Gal. Doria-Pamphili) and Raphael and his Fencing-master (c. 1518; Paris, Louvre); for the presumed Self-portrait (c. 1513–14; ex-Czartoryski Col., Kraków), which he recorded in the Italian Sketchbook (fol. 109v); and for the portrait of Bindo Altoviti (c. 1518; Washington, DC, N.G.A.). Those double portraits were not far from the forefront of van Dyck’s mind when he painted the double portrait of Thomas Killigrew and Lord Crofts and ten years earlier that of Cornelis and Lucas de Wael. A reverberation of Raphael’s dandified Self-portrait in a furred gown is to be found in van Dyck’s Self-portrait with the Broken Column; and in the oval portrait of Sir Endymion Porter and Van Dyck (c. 1635; Madrid, Prado), the painter posed himself as Altoviti.

Of contemporary painters, Rubens apart, Guido Reni became the crucially important influence on van Dyck. This developed beyond any actual meeting in Bologna and his years south of the Alps; the pen sketch (Madrid, Real Acad. S Fernando) is not a self-portrait of Reni, but a hasty impression by van Dyck. Before abandoning Genoa, van Dyck had studied Reni’s Assumption of the Virgin in the Jesuit church of S Ambrogio (1617; in situ), and memories of that mighty altarpiece mingled with those of Rubens’s treatment of the theme in an oil sketch (Vienna, Akad. Bild. Kst.) painted within a few years of his return to Antwerp. The Mystic Marriage of St Catherine (1630; London, Buckingham Pal., Royal Col.) shows not only an addiction to Titian’s gusto in composition but also a refinement of sentiment, expressed both in silvery tone and in morphology, that evinces a lasting attraction to Reni. Whereas the Virgin and Child with SS Rosalie, Peter and Paul (1629; Vienna, Ksthist. Mus.) owes much of its grand conception to a composition by Titian of the Virgin and Child with a Donor (untraced), noted by van Dyck in his Italian Sketchbook (fol. 28v), the core appears, when reduced in scale and modified to the contemporary Vision of St Anthony (1629; Milan, Brera), to be of Reni’s mode of refinement of types and sensibility to illumination. That this affinity with Reni was deep rooted is shown nowhere better than in the three-quarter-length portrait of Nicholas Lanier (1628; Vienna, Ksthist. Mus.). The Reni-like quality of the finished portrait, the work of a week, emanates from a preparatory drawing in black and white chalk (Edinburgh, N.G.), which in its agitation of rippling light on the ridges and slower movement of shadow in the hollows of drapery so vividly recalls Reni’s drawings for the ‘gran’ Madonna dei Signori Marchesi Tanari’ in Bologna (i.e. the Virgin and Child with St John or ‘Madonna Tanari’, 1627–8; untraced). A paradigm of van Dyck’s eclectic progress as a draughtsman in chalks in less than 15 years is the comparison of the advancing Head and Front Quarters of a Grey Horse (c. 1618–21; Chatsworth, Derbys), studied for St Martin Dividing his Cloak with a Beggar (versions at Zaventem, St Martin, and Windsor Castle, Berks, Royal Col.) in the boldly plastic manner that he had learnt in Antwerp from Rubens, himself incarnate with the spirit of Annibale Carracci, and the Studies of a Greyhound (c. 1633; London, BM), seated but quivering with nervous mobility, studied as a companion to the sitter in the portrait of James Stuart, 4th Duke of Lennox (c. 1633; New York, Met.), with Reni-like vibrancy of surface.

2. Literary and Classical.

Van Dyck retained the Latin of an educated bourgeois. He was able to comment on the De pictura veterum of Francis Junius. In the ex-voto painted in memory of his father, the Crucifixion with SS Dominic and Catherine of Siena (see fig. above), his inscription and funeral device alluded in ‘ne patris svi manibus …’ to the manes, the spirit of the departed. However, as Roger de Piles perceived, he was less minded to be a scholar than Rubens. His interest in antiquity was slighter and evanescent. He drew a fragment of a Niobid frieze (Paris, Fond. Custodia, Inst. Néer.); he included one of the Arundel marbles, a bas-relief from Smyrna (London, Mus. London), in his Continence of Scipio for Buckingham; besides his conspicuous placing of a wine ewer with an ithyphallic satyr for a handle in the early Samson and Delilah (1619–20; London, Dulwich Pict. Gal.), which was intended to rival Rubens’s treatment of that subject for Nicolaas Rockox (c. 1609–10; London, N.G.), he also incorporated knowledge of the Belvedere Torso (Rome, Vatican, Mus. Pio-Clementino) and the Borghese Hermaphrodite (Paris, Louvre); and he drew in his Italian Sketchbook (fol. 33v) the statue (Paris, Louvre) believed to be of the ancient Cynic philosopher Diogenes. Moreover, he was sufficiently familiar, doubtless through contact with Rubens, with Otto van Veen’s Amorum emblemata (Antwerp, 1608), witness his Time Clipping Cupid’s Wings (1630–32; Paris, Mus. Jacquemart-André); with Virgil’s Aeneid, witness his Venus at the Forge of Vulcan (1630–32; Paris, Louvre); with Tasso’s Gerusalemme liberata (1580–81), witness his Rinaldo and Armida (1629); with Guarini’s Il pastor fido (1589), witness his Amaryllis and Myrtillo (1631–2; Pommersfelden, Schloss Weissenstein); as well as with the common literary stock of painters, Ovid, the Vulgate and the Lives of the Saints. To the end of his life he kept up with Jesuit writings, the frontispiece to a recent publication being as essential an ingredient in his conception of the painted memorial to Rachel de Ruvigny as was Rubens’s Jupiter (c. 1618; Strasbourg, Mus. B.-A.). In portraits he essayed emblem and allegory rarely. After his contact with Agostino Pallavicini in Genoa, it was not until his association with Sir Kenelm Digby in London that he was fully engaged by learned, posthumous allusions to Rachel de Ruvigny, triumphant over death, and by those contrived by Sir Kenelm for the allegorical portrait of his wife, Venetia Stanley, Lady Digby as Prudence (1633; London, N.P.G.), which were based on Juvenal. The pretty fancies of the adolescent Lady Mary Villiers with Lord Arran as Cupid (c. 1636; Raleigh, NC Mus. A.) and Mary, Duchess of Lennox as St Agnes (1637; Windsor Castle, Berks, Royal Col.) strike no deep chords.

First Florentine period, 1472–c. 1482.

The first reference to Leonardo as an artist occurs in 1472, when he was required to pay his dues to the painters’ Compagnia di S Luca in Florence. His apprenticeship in the studio of the sculptor Andrea Verrocchio is recorded by Vasari and confirmed by a reference in 1476 to his continued residence there, but the date at which this apprenticeship started is unknown. Verrocchio’s workshop, which undertook a wide range of commissions, including sculpture in bronze, stone and terracotta, decorative work in metals and various stones, paintings and at least one major feat of engineering (the orb on the top of the lantern of Florence Cathedral), provided a solid grounding for Leonardo’s subsequent versatility. Verrocchio was himself an inventive artist, particularly in figure sculpture, in which he pioneered a freedom of movement and viewpoint.

The first dated indication of Leonardo’s ability as an artist is a remarkable pen-and-ink drawing of a Tuscan Landscape dated 5 August 1473 (Florence, Uffizi), which already signals an exceptional talent and mind at work. The subsequent record of his activities before his move to Milan c. 1482 is sparse. In 1476 he was accused anonymously of sodomy, but no prosecution was sustained. His receipt of an official commission in January 1478 for an altarpiece in the chapel of S Bernardo in the Palazzo della Signoria indicates his growing reputation. The altarpiece was not executed by Leonardo and was eventually supplied by Filippino Lippi (Florence, Uffizi). The note on one of his drawings, the Studies of Heads and Machines (1478; Florence, Uffizi), saying that he ‘began two Virgin Marys’, probably refers to small panels rather than the altarpiece. A year later he made an annotated drawing of the Hanged Body of Bernardo Baroncelli (Bayonne, Mus. Bonnat), the murderer of Giuliano de’ Medici, brother of Lorenzo the Magnificent, which may have been connected with a project to depict the traitors on the outside of the Palazzo del Podestà in the customary manner. The second recorded commission from this period was in March 1481 for an altarpiece in S Donato a Scopeto. Although the subject is not recorded, the unfinished panel of the Adoration of the Magi (Florence, Uffizi) was almost certainly intended for this destination. Filippino Lippi subsequently provided a completed altarpiece of the same subject (Florence, Uffizi).

The visual record of Leonardo’s work from this first Florentine period includes a remarkable group of inventive drawings for varied artistic projects. These, together with notes and an inventory of works completed just before or after his arrival in Milan , also show the first signs of the broadening range of his interests. The surviving drawings illustrate machinery (including the precise gearing of scientific instruments), aspects of military engineering, as well as optical phenomena and geometry, and the inventory lists studies made from nature, detailed representations of surface anatomy, portrait and compositional drawings, together with ‘some machines for ships’ and ‘some machines for water’. Besides the drawings, there is a small body of paintings that can be attributed in whole or in part to him (see §II, 1 below).

First Milanese period, c. 1482–99.

At some date after the last recorded payment from S Donato for the Adoration of the Magi (Sept 1481), Leonardo left Florence for Milan. He is not firmly documented there until April 1483, but it is likely that he moved during the course of 1482. In the draft of the letter in which he outlined his talents to Ludovico Sforza (‘il Moro’), ruler of Milan and Duke of Bari, he concentrated on his capabilities as a military engineer, promising to ‘apprise you of my secrets’ (Cod. Atlantico, fol. 391r). He listed ten categories of military devices for use on land and sea, ranging from bridges and tunnels to guns and mortars ‘outside the common use’. Only at the end of the letter did he mention that he could ‘undertake sculpture of marble, bronze and clay, similarly in painting whatever can be done, to bear comparison with anyone else, whoever he is’. He also mentions that ‘work on the bronze horse may be taken on’. This refers to the long-standing scheme to erect an equestrian memorial to Francesco Sforza, Ludovico’s father and the first Sforza Duke of Milan, a project for which initially Antonio Pollaiuolo appears to have been considered. The tone of the letter suggests that Leonardo hoped his move to Milan would provide greater opportunities to develop the full scope of his work than had been possible in Florence. It remained true throughout his life that his activities flourished better within a court and in receipt of a regular income than when he needed to make a living from the completion of commissioned works of art.

The first notice of Leonardo’s activity in Milan occurs in a contract for work on an altarpiece. In company with the brothers Ambrogio and Evangelista de’ Predis, he agreed to provide the painted decoration and panels for a large sculpted altarpiece by the wood-carver Giacomo di Damiano ( fl 1469–1502) for the Confraternità dell’ Immacolata Concezione in their chapel in S Francesco Grande, Milan. In addition to polychroming and gilding the wooden architecture and sculpture, the painters were expected to provide paintings of the Virgin, prophets and angels to be set in the frame.

The subsequent history of this commission, which went through a series of protracted legal wrangles, involves some of the lengthiest and most confusing documentation for any Renaissance painting. The dispute centred on the confraternity’s claims that the painters had failed to fulfil their obligations and the painters’ assertion that the value of the panel of ‘Our Lady done in oils’ was far greater than the sum the confraternity was offering to pay. By the time a procurator was appointed in 1496 to settle the dispute, Ambrogio de’ Predis and Leonardo had appealed to a higher authority, probably the Duke. By 1503 matters were still not resolved, by which time Leonardo had left Milan. In 1506 arbitrators stipulated that Leonardo had to complete the painting of ‘the most glorious Virgin Mary’ within two years at an agreed price. The painting was finished by August 1508, when Ambrogio, on Leonardo’s behalf, was given permission to remove the painting from its frame to make a copy.

Of the two surviving versions of the painting, now known as the Virgin of the Rocks, one (London, N.G.) is known to have come from the altarpiece in S Francesco Grande, while the early history of the other version (Paris, Louvre; see fig.), stylistically the earlier of the two, is unclear. Attempts have been made to reconcile the written evidence and the two paintings, but none can be confirmed. The two most straightforward hypotheses are either that the Louvre painting was completed but withheld by the artists and sold privately elsewhere, while the London panel was a second version, made to fulfil the legal requirements; or that either the Louvre painting or some other part of the altarpiece was incomplete until 1508, and that the London version was the copy made in that year and substituted for the original. The stylistic evidence marginally favours the former hypothesis, in that the Louvre Virgin of the Rocks appears to be wholly in the style of the 1480s, while the London version exhibits features of Leonardo’s work from the mid-1490s, even if it is not wholly by him.

The nature of Leonardo’s engagements at the Sforza court is unclear in the 1480s. His Portrait of a Lady with an Ermine (Kraków, Czartoryski Col.), presumed to be the portrait of Ludovico Sforza’s mistress Cecilia Gallerani that was celebrated in a Milanese poem, should be dated, on new evidence, to c. 1490–91. In 1487 Leonardo submitted a model for the scheme to design a tiburio (crossing tower) for Milan Cathedral, although he did not undertake the commission. It is also to this period that the Codex Trivulziano (Milan, Castello Sforzesco), the first of his surviving notebooks, can be dated. For the rest of his life he kept notebooks written in mirror handwriting and filled with drawings and diagrams that record various intellectual endeavours and scientific investigations in which he was involved. The Codex Trivulziano contains, among other things, studies for the tiburio, philosophical aphorisms and Latin word lists. A sheet in pen and ink, with Two Studies of a Human Skull (Windsor Castle, Royal Lib., 19059r), dated 1489, is one of a series of anatomical investigations concerned with the brain, nervous system and senses. Although there had been earlier signs of his interest in a range of scientific and technical matters, sustained explorations of questions lying outside his immediate professional involvements are fully documented only from the late 1480s.

During the 1490s Leonardo was involved with ceremonial activities at the Sforza court, with painting and sculpture and with his own intellectual pursuits in a growing range of natural, physical and mathematical sciences. Typical of his work as a court artist were his admired stage designs for the Festa del paradiso, a spectacle by Ludovico’s leading court poet, Bernardo Bellincioni, composed in 1490 as a wedding celebration. The same year he resumed serious work on the equestrian monument to Francesco Sforza. In 1497 he is documented as nearing the completion of the Last Supper, begun c. 1495, in the refectory of S Maria delle Grazie, Milan (in situ), and in 1498 he was painting the mural decoration of the Sala delle Asse (Milan, Castello Sforzesco). His notebooks also suggest that he participated in various architectural and engineering projects, including the extensive schemes for canalization, urban planning and decoration in Vigevano, close to Ludovico’s birthplace.

In his scientific work, Leonardo began to embrace a variety of concerns. Anatomy and optics were central among these (see Anatomical studies and Science and art), but he also embarked on detailed investigations of statics and dynamics, with an almost obsessional interest in the complex patterns of motion in water. His notebooks reflect a sustained campaign of self-education in the basic theoretical concepts of Classical and medieval science, and in the elements of mathematics. In this latter ambition he was greatly aided by the arrival of the mathematician Luca Pacioli at the court in 1496, and the following year they collaborated on the illustrations of geometrical bodies in Pacioli’s De divina proportione, which was eventually published in Venice in 1509.

The visual record of Leonardo’s artistic products during the 1490s is disappointingly meagre. The project for the huge equestrian monument to Francesco Sforza progressed to the point at which the full-sized clay model could be exhibited in 1494, but the bronze was never cast, and the model seems not to have survived Ludovico’s fall in 1499. The Last Supper was his major completed achievement. Early viewers testify to its extraordinary impact; however, the partly experimental technique led to the wall painting’s rapid deterioration, and it exists today only as a fragmented ghost of its former presence. Although the Duke requested in 1497 that Leonardo start on ‘the other wall’ when he finished the Last Supper, no sign remains of other work by him in the refectory. The portraits of the Duke and Duchess and their Children added to Giovanni Donato da Montofano’s Crucifixion on the opposite end wall are too damaged to permit a definite judgement, but the underdrawings appear too routine to be attributed confidently to Leonardo.

Leonardo’s decorative painting in the Sala delle Asse, depicting trees intertwined as a great bower, also survives in an incomplete and heavily restored form. It seems likely that he was responsible for other work in the suite of rooms in the Castello Sforzesco, which Ludovico was transforming and extending, but no traces survive. Three portraits of more or less autograph quality—the Portrait of a Musician and the Portrait of a Woman in Profile (both Milan, Bib. Ambriosiana) and the Portrait of a Woman, known as ‘La Belle Ferronnière’ (Paris, Louvre)—may also be assigned to his period at the Sforza court. When Ludovico fled Milan in 1499, in the face of the invading armies of the French king Louis XII, Leonardo sent money to Florence for safekeeping. Although he apparently entered into some kind of agreement with the King, he left Milan in December. He stayed briefly in Mantua, where he made a portrait drawing of Isabella d’Este (damaged version, Paris, Louvre; for illustration see Este (i), (6)), and in Venice, where he appears to have given some advice on hydraulic engineering. His visit to Mantua is not surprising in view of the close links between the Sforza and the Este families, and of Isabella’s known interest in Leonardo’s art.

Second Florentine period, 1500–mid-1508.

In April 1500 Leonardo returned to Florence, where he was faced with the prospect of re-establishing his career. During the next six years Isabella d’Este endeavoured to obtain a painting from him. She hoped that he would make a painting based on the portrait drawing and subsequently that he would provide a subject painting—at one point she suggested an image of Christ at the age of 12. The correspondence in which Isabella pursued her frustrating quest provides the best evidence for Leonardo’s activities immediately after his return to Florence. Her plenipotentiary in Florence, the Carmelite Fra Pietro da Novellara, wrote to her on 3 and 14 April 1501, mentioning the painter’s obsession with geometry and that his pupils were making copies of his paintings to which he occasionally put his hand. The letters also describe two works by Leonardo. One was a small panel painting for Florimond Robertet, the secretary to the French king, which showed the Virgin and Child contesting the possession of a yarnwinder. Later known as the Madonna of the Yarnwinder, the best versions of this much copied painting are in the Duke of Buccleuch’s collection and a New York collection.

The other work was a large-scale cartoon (untraced) of the Virgin and Child with St Anne and a Lamb, in which the life-sized figures were cunningly compressed into a compact group. The cartoon seems to have been drawn when Leonardo was involved with a commission for an altarpiece for SS Annunziata (later finished by Pietro Perugino), which was apparently ceded to him on his return to Florence by Filippino Lippi. Leonardo was provided with accommodation in the monastery of SS Annunziata, and it was there in 1501 that he exhibited his cartoon to large crowds, though it was probably not intended as the design for the altarpiece.

That Republican Florence did not provide the most appropriate arena for Leonardo’s talents is perhaps indicated by the fact that in 1502 he accepted the appointment as Cesare Borgia’s ‘architect and general engineer’, with responsibilities that took him to Urbino and other Central Italian cities. The most spectacular product of his work for Cesare is the Map of Imola (Windsor Castle, Royal Lib., 12284). In 1503 he was again in Florence and appears to have been one of the engineers involved in Machiavelli’s ill-fated plans to divert the River Arno around Pisa, when Florence was at war with the city.

Later in the same year Leonardo received the highly prestigious commission for a wall painting of the Battle of Anghiari (destr.) in the Sala del Maggior Consiglio, the great new council hall in the Palazzo della Signoria, which the Republic had erected after the expulsion of the Medici in 1494. The subject was to commemorate a Florentine victory over the Milanese in 1440. Leonardo was provided with a room in S Maria Novella in which to make the huge cartoon (destr.), and work seems to have proceeded steadily, although it was interrupted during the autumn of 1504, when the Florentine authorities sent him to Piombino to advise on fortifications. Payments were made for materials during 1504, and one of his own notes (Madrid, Bib. N., MS. II, fol. 2r) provides evidence that he was actually painting on the wall in the summer of 1505. In 1504 Michelangelo received the commission to paint the Battle of Cascina (unexecuted) as a companion piece to the Battle of Anghiari and joined Leonardo as an apparently unsympathetic rival.

However, there were growing signs that Leonardo might eventually fail to complete the commission. His characteristically experimental technique was running into trouble, and his notebooks testify that his diverse intellectual concerns were again coming to the fore, including studies of bird flight and geometry. Finally, in 1506 a train of events marked the abandonment of the project. In May he was granted leave of absence to work in Milan for three months, perhaps in response to the settlement of the litigation surrounding the Virgin of the Rocks. Although he returned to Florence briefly in March 1507, and for a longer period from September to the following spring, his residence in Florence was effectively at an end. The winter of 1506–7 was apparently occupied with the study of anatomy, bird flight and mathematics. The last of his substantial artistic involvements in Florence seems to have been the assistance he provided, according to Vasari, to Giovanni Francesco Rustici, who was making the bronze group of St John the Baptist Preaching between a Pharisee and a Levite (1506–11) for the exterior of the Florentine Baptistery (in situ).

It is difficult to assign a single, finished, wholly autograph painting to the years 1500 to 1508. It is reasonable to assume that the Madonna of the Yarnwinder was completed, but even that might not have been entirely by Leonardo. The incomplete and partly ruined painting of the Battle of Anghiari survived until the remodelling of the council hall in the 1560s, and, although strenuous efforts have been made to discover it under the later paintings by Vasari, it has not so far reappeared. A number of projects for compositions of the Virgin and Child can be dated to these years, as can an innovative design for a painting of an Angel of the Annunciation (untraced), which developed into the later composition of St John the Baptist (Paris, Louvre). He also began work on a composition of Leda and the Swan , experimenting with a kneeling figure of Leda (reflected in versions by followers) and a standing version, the latter known to Raphael in Florence as a developed design, cartoon or unfinished painting. The final version of Leda (untraced) may not have been completed until after 1513.

The painting generally regarded as the central product of these years is the so-called ‘Mona Lisa’ (Paris, Louvre), although even it presents some problems of dating. The identity of the sitter was for many years uncertain, but Vasari’s claim that the lady in the portrait was ‘Mo[n]na Lisa’, the wife of Francesco del Giocondo (hence the alternative name of ‘La Gioconda ’ or ‘La Gioconde’), was confirmed in 1991 with the publication of the 1525 death inventory of Leonardo’s assistant of 30 years, gian giacomo Caprotti, who seems to have been in possession of a number of his master’s works, including this portrait (see Shell and Sironi). In 1495 Lisa Gherardini (b Florence, 1479) married Francesco del Giocondo, an important figure in the Republican government, whose portrait Leonardo is also thought to have painted. Earlier confusion over her identity arose, among other things, from what was previously thought to be the earliest reference to the portrait, written by Ambrogio de’ Beatis on a visit to Leonardo in France in 1517; he described a portrait of ‘a certain Florentine lady made from nature at the instigation of the late Magnificent Giuliano de’ Medici’ (the Duc de Nemours), who was Leonardo’s patron in Rome after 1513, and it had long been assumed that he was referring to the ‘Mona Lisa’. (It is now thought likely that the work de’ Beatis saw on this visit was the portrait of another Florentine woman.)

Leonard da Vinci: ‘Mona Lisa’, panel, 600×470 mm, c. 1500–07 (Paris, Musée du Louvre);

According to Vasari, Leonardo began the portrait of Lisa del Giocondo between his arrival in Florence in 1500 and the commencement of his work on the Battle of Anghiari late in 1503; after four years, however, the work was still unfinished. The appearance of the picture lends support to the idea that it was painted over an extended period, since the craquelure of the face suggests that it was executed at a different time from the hands, which exhibit the thinness of his latest manner of painting.

Second Milanese period, mid-1508–1513.

From the summer of 1508 to September 1513 Leonardo was resident in or near Milan, working initially for the French rulers of the city under the direct supervision of the governor, Charles II d’Amboise, Comte de Chaumont. He appears to have taken up a range of duties broadly equivalent to those he had performed at the Sforza court, including providing designs for ephemeral items of courtly entertainment. He also embarked on designs for another equestrian monument, for Gian Giacomo Trivulzio, an Italian general who was serving the French. The scheme progressed as far as a detailed specification and costing of the life-size horse and rider, with a substantial base and secondary sculpture (Cod. Atlantico, fol. 179va). The architectural work Leonardo is known to have undertaken for Charles d’Amboise is not clearly identifiable. Charles’s project may have been for the kind of airy, colonnaded villa with which Leonardo had been experimenting since his first Milanese period (e.g. Cod. Atlantico, fol. 158ra). He may also have been involved in the plans for the church of S Maria alla Fontana, Milan, but the executant of the work is firmly documented as Giovanni Antonio Amadeo. After the reinstatement of the Sforza regime in 1512 under Ludovico’s son Massimiliano, Leonardo remained in Lombardy for more than a year.

Among Leonardo’s scientific endeavours during this second Milanese period is a series of outstanding anatomical drawings of human musculature and the skeletal system. His exploration of certain geometrical questions, particularly problems of transformation of volume and area (e.g. the squaring of the circle), became increasingly obsessive, as did his investigation of the dynamics of fluids, whether in the guise of the motion of water or in such related forms as the turbulent flow of the blood in vessels of the human body.

The only documented painting completed on Leonardo’s return to Milan was the Virgin of the Rocks: the style of the second version (London, N.G.) is consistent with its having been begun in the late 1490s and subsequently finished by Leonardo with studio assistants on his return to Milan. By contrast, the only autograph painting that can be wholly assigned with some confidence to this period is the St John the Baptist (Paris, Louvre), which developed from his Florentine Angel of the Annunciation and is reflected in pupils’ drawings datable c. 1509. The Virgin and Child compositions on which he was working for Louis XII cannot be certainly identified with any of the surviving paintings, although it is highly likely that one of them was a variant of the theme of the Virgin and Child with St Anne. Two main types of this composition are known: the version that included the lamb, as in the lost cartoon of 1501 and a surviving painting (Paris, Louvre), and the type in which the young St John is integrated into the narrative, as in the Burlington House Cartoon (London, N.G.). The latter has sometimes been dated to 1490–1500, but the style of its draughtsmanship and of the closest preparatory drawing (pen and ink over black chalk; London, B.M.) is increasingly recognized as belonging to c. 1505–7. The Louvre Virgin and Child with St Anne and a Lamb is not clearly datable by reference to other paintings, but the related drawings (Windsor Castle, Royal Lib., 12527, 12530, 12533), handling of colour and treatment of the landscape suggest a late date, perhaps c. 1515.

5. Rome and France, after 1513.

In October 1513 Leonardo visited Florence on his way to Rome, where he was accommodated in the Belvedere under the patronage of Giuliano de’ Medici, Duc de Nemours. He appears to have been involved with the military work that Giuliano was undertaking for the Medici pope, Leo X, and worked on the design and manufacture of burning mirrors that could have military and civil uses. The continued intensity and variety of his intellectual endeavours, particularly in anatomy (cardiology and embryology), optics and geometry, coupled with his travels in Giuliano’s service, do much to explain the reported impatience of Pope Leo, who doubted whether Leonardo would ever finish anything.

It is possible that Leonardo was present at Bologna in 1515 at the meeting between Leo X and the new French king, Francis I. His elaborate red chalk drawing of the Allegory of the Wolf and Eagle (Windsor Castle, Royal Lib., 12496) may well refer to the concordat between pope and king. In any event, Francis was as enthusiastic about Leonardo’s work as his predecessor and succeeded in attracting Leonardo to France at some point between August 1516 and May 1517. For the rest of Leonardo’s life, Francis seems to have acted as an ideal patron—actively promoting new projects and keen to exploit the range of Leonardo’s talents, but also understanding of the artist’s character as a natural philosopher or seer. Leonardo was clearly regarded as an ornament of the court and, as such, was visited by Cardinal Louis of Aragon’s party on 10 October 1517, the occasion recorded by Ambrogio de’ Beatis, who was the Cardinal’s secretary. As ‘first painter and engineer’ to the King, Leonardo was provided with accommodation at the manor house of Clos-Lucé, Amboise.

Leonardo’s final years have often been seen as dominated by his geometrical obsessions and a growing sense of pessimism, expressed most vividly in the visions of cataclysmic storms in his series of drawings of A Deluge (Windsor Castle, Royal Lib.). His health was also apparently deteriorating. De’ Beatis reported that some paralysis was affecting his right side, probably as the result of a stroke. However, a study of folios and notebooks datable to the period 1516–19 reveals a remarkable range of continuing activities. There are only occasional signs of physical frailty. His assistants, most prominent among whom was the well-born francesco Melzi from Lombardy, probably played an increasing role in the physical work, but his inventiveness appears undiminished.

As in his two Milanese periods, Leonardo furnished designs for courtly entertainments, including a revised version of his design for the Festa del paradiso of 1490. His most ambitious project was for a huge royal palace at Romorantin, with associated canalization. A scheme was devised for an extensive residence, translating French château design into the language of the Renaissance. Leonardo’s concerns extended from the overall conception to such details as the design of toilet doors with counterweights. Although the project did not materialize, echoes of Leonardo’s ideas can be seen in subsequent French château design.

The evidence of Leonardo’s involvement with painting in France is equivocal, relying on secondary sources from later in the 16th century. It is virtually certain that the ‘Mona Lisa’, the St John the Baptist, the Leda and the Virgin and Child with St Anne were taken by Leonardo to France, where some may have been completed. He probably made an anamorphic painting for Francis I depicting a fight between a dragon and a lion in such a way that it made sense only when viewed from a shallow angle. A composition with a half-length female nude, known as the ‘Monna Vanna’, which exercised a notable influence on a series of erotic paintings from the school of Fontainebleau, may also have depended on a prototype by Leonardo himself, possibly a drawing or cartoon (such as that in Chantilly, Mus. Condé). However, de’ Beatis’s testimony that illness had left Leonardo unable to undertake painting needs to be taken seriously, and it is unlikely that any wholly autograph paintings were initiated and completed in France.

On 23 April 1519 Leonardo drew up his will, bequeathing most of his drawn and written legacy to Melzi. Following Leonardo’s death, Melzi wrote movingly to the painter’s brothers in Florence, one of whom’s son, Pierino da vinci, became a sculptor. Several of Leonardo’s paintings seem to have come into the hands of his assistant Caprotti, who had also travelled with his master to the French court. On 12 August 1519 Leonardo was buried in the church of St Florentin at Amboise, although his remains are thought later to have been transferred to the chapel of St Hubert at the château of Amboise.

1. Paintings.

Leonardo completed relatively few paintings, and no more than ten surviving works are generally accepted as being finished wholly by him. A further three autograph paintings remain unfinished, while a small group of works may be classified as studio products in which he played a greater or lesser role. Early sources, particularly Vasari and Giovanni Paolo Lomazzo, refer to a number of paintings now unknown, but there is often no way of telling if these works were indeed by Leonardo himself. Additionally, there is a host of Leonardesque paintings by followers, ranging from presumed copies of original works to free variations on Leonardo’s compositions. Many of the Leonardesque paintings originated from Milan, where artists may have had access to originals brought back from France by Caprotti, whose own copies, as well as those by others, transmitted the conventional image of the ‘Leonardesque’ to later ages.

Drawings.

Leonardo was one of the most innovative and fertile draughtsmen of any age (see fig.). In his hands the practice of drawing became a flexible extension of creative thought, not only expressing a series of new ideas in teeming abundance but also becoming, through a rapid confusion of scribbled alternatives superimposed on each other, a way of permitting chance configurations to aid the inventive process. Drawing became a form of visual thinking rather than a merely functional means for the design of a picture.

Leonardo da Vinci: Designs for a Nativity or Adoration of…At the beginning of his career Leonardo achieved mastery of the two most important drawing techniques of the period, metalpoint (see fig.) and pen and ink. The Study of Arms and Hands (Windsor Castle, Royal Lib., 12558) that may have served for the portrait of Ginevra de’ Benci is drawn in silverpoint with white heightening on pink prepared paper and demonstrates a meticulous control of parallel hatching to suggest graded relief. One of the last instances in which Leonardo used this traditional medium, his Studies of a Standing Horse, seen from the side and front (c. 1490; Windsor Castle, Royal Lib.), probably for the Sforza monument, may be regarded as having taken the potential of silverpoint to its limits.

Leonardo’s work in pen exhibited from the first an exceptional vitality of touch, as in the Tuscan Landscape (1473; Florence, Uffizi), which is characterized by an extraordinary suggestion of life and atmosphere. In the drawings for the Adoration of the Magi and various Virgin and Child compositions of the 1480s and early 1490s he evolved a graphic style of unprecedented rapidity and suggestiveness. Other Renaissance draughtsmen, including Verrocchio, had used pen and ink for quick sketches, but no one had approached Leonardo’s bold and dynamic method of ‘brainstorming’, in which alternative forms emerge from a tangled confusion of lines. The rapid pen studies of the Virgin and Child with a Cat (c. 1478–81; London, BM) show the complex interweavings of bodies in motion that become possible with this approach. On the verso of this sheet, the design was traced through in reverse, becoming the starting-point for a further series of variations that were clarified by the addition of an ink wash. Such paintings as the Benois Madonna reflect the way in which complex motions can be orchestrated through this manner of sketching.

Leonardo da Vinci: Head of a Man in Profile Facing…Throughout Leonardo’s career, pen and ink remained the technique he most regularly used, not only for preliminary sketches but also for scientific illustrations and representations of machinery and architecture. During the late 1490s his system of shading with pen underwent an important and influential change: the use of diagonal parallel hatching, which moved from top left to bottom right (he was left-handed; see fig.), was progressively replaced by curved pen strokes that follow the forms. The drawings for the Leda from 1506 onwards, such as the Study for the Kneeling Leda (pen and ink over black chalk; Chatsworth, Derbys) represent the extreme development of this graphic style.

Leonardo da Vinci: Head of the Virgin., black chalk, red…

For the study of the component parts of compositions, Leonardo often turned to other media. Early in his career he seems to have made studies of draperies arranged on lay figures using a fine brush and white heightening on linen (Paris, Louvre, and elsewhere) in the manner of his master, Verrocchio, and contemporaries such as Domenico Ghirlandaio, but he progressively used the softer and more flexible media of red and black chalk during the 1490s, above all in the studies for the Last Supper. The red-chalk drawings for Apostles’ heads, sometimes on reddish prepared paper, such as the Head of Judas (Windsor Castle, Royal Lib., 12547), make subtle interplay between softly luminous shading, rhythmic contours and selective areas of dense shadow. The heads of shouting warriors for the Battle of Anghiari (e.g. Budapest, Mus. F.A.; see fig.) represent the high-point of this technique. The softer and grainier black chalk was particularly suited to creating effects of sfumato modelling (see fig.). When combined with white heightening, as in a Study of a Sleeve for St Peter in the Last Supper (Windsor Castle, Royal Lib., 12546), or with heightening and wash, as in studies of drapery (Windsor Castle, Royal Lib., 12530; Paris, Louvre) for the Louvre St Anne, extraordinarily rich effects of light and shade could be attained. Comparable effects were achieved with charcoal and white heightening in the cartoon of the Virgin and Child with SS Anne and John the Baptist (London, N.G.), which retains a far greater degree of fluidity and lack of resolution than would have been normal in a full-scale drawing. Black chalk was also the favoured medium for his drawings of A Deluge (e.g. Windsor Castle, Royal Lib.), where its sombre, atmospheric qualities were ideally suited to such dark expressions of cosmic violence.

Leonardo was also an innovator in the use of colour in drawings. His geographical studies use coloured washes to distinguish forms in flat maps according to a convention or colour code, and also for more naturalistically descriptive purposes, as in the Bird’s-eye View of Arezzo, Borgo San Sepolcro, Perugia, Chiusi and Siena (1502; Windsor Castle, Royal Lib., 12680). His use of coloured pastels (see Pastel, §1), mentioned in 16th-century sources, cannot be demonstrated in fully autograph drawings, but artists in his circle, most notably Giovanni Antonio Boltraffio, certainly exploited the technique.

In his scientific and technical drawing Leonardo experimented with many of the illustrative techniques used in later textbooks, including various forms of solid section, transparency of overlying parts and exploded diagrams of components (see fig.). He pushed the descriptive potential of static drawing on a flat surface towards its ultimate limits.

Sculpture

That Leonardo practised as a sculptor is not in doubt, but attempts to attribute surviving sculpture to him have met with generally unsatisfactory results. Vasari was probably correct in saying that Leonardo made terracotta heads of women and children (perhaps of the infant Christ or the infant St John) early in his career, since such heads were part of the stock-in-trade of a sculptor’s studio in the 1470s. When he advertised his services to Ludovico ‘il Moro’, he claimed proficiency in ‘sculpture, in marble, bronze and clay’, but the subsequent records of his sculptural activity suggest that he worked as a modeller rather than as a carver.

Leonardo’s most substantial sculptural undertaking was the great equestrian monument to Francesco Sforza that was intended to be cast in bronze in Milan. The first record of his direct involvement occurs only in 1489, when the Duke expressed doubts about Leonardo’s ability to complete the work. In April 1490 Leonardo himself noted that he ‘restarted the horse’ (Paris, Inst. France, MS. C, fol. 15v). By this time it is likely that he had set aside his technically impractical scheme for a rider on a rearing horse and reverted to the more traditional walking pose, as in Donatello’s Gattamelata (c. 1447–53; Padua, Piazza del Santo) and Verrocchio’s Bartolommeo Colleoni (c. 1479–92; Venice, Campo SS Giovanni e Paolo), but on a scale of three times life size. In 1491 and 1492 he worked on the full-scale clay model, which made a great impact when it was shown as part of the marriage celebrations of Bianca Maria Sforza and Emperor Maximilian. The surviving records of Leonardo’s project in his drawings and manuscripts are far from complete, but include beautiful studies of horses from life, proportional studies, for instance the silverpoint measured drawing of a Horse in Profile to the Left (Windsor Castle, Royal Lib., 12319) and elaborate schemes for the casting of the colossus mainly in the second Madrid codex (Madrid, Bib. N., MS. 8936). The drawings show his concern to capture the nervous vitality of a highly bred horse, both in its overall motion and in the rhythmic grace of individual parts. There is little indication of the intended pose of the rider, who may not even have been included in the clay model that was exhibited. The bronze was never cast, and the model was destroyed.

Leonardo’s plans for the equestrian monument to Gian Giacomo Trivulzio are recorded in a series of pen drawings (Windsor Castle, Royal Lib., 12353, 12355, 12356), which show that an energetically striding horse and gesturing rider were to have been mounted on an elaborate architectural base. The base was to have contained a recumbent image of Trivulzio on his sarcophagus, and a series of eight ‘captives’ (bound nude male figures) were to have been attached around its margins in a manner comparable to Michelangelo’s projected scheme for the tomb of Pope Julius II (see fig.). Leonardo’s project was destined not to reach even the stage of a full-scale model, although some later drawings, such as the black-chalk study showing a fallen soldier trampled beneath his horse’s hooves (Windsor Castle, Royal Lib., 12354) which may date from 1511, show that he continued to meditate on the possibility of reviving his earlier idea for a rearing horse.

The most tangible surviving evidence of Leonardo’s qualities as a sculptor occurs in Rustici’s bronze group of St John the Baptist Preaching. Although there is no direct documentation of Leonardo’s involvement with this group, Vasari’s account of Leonardo’s participation is supported by the visual evidence. The complex yet monumentally graceful poses, the individual characterizations and contrasted expressions speak the language pioneered by Leonardo in the Last Supper, while the intricate communication between the figures and the spectator on the ground below can be seen as the realization of ideas with which Leonardo had long been experimenting. The precise roles of Rustici and Leonardo are impossible to disentangle, but it is clear that Leonardo’s presence resulted in Rustici working on a higher plane than in any of his wholly independent works.

The high probability that Leonardo worked on smaller-scale sculpture in his studio, either in terracotta or wax, has encouraged a search for surviving examples or, more realistically, bronzes dependent on his models. The surviving terracotta that deserves the most serious consideration is a bust of the Infant Christ (ex-Galludt priv. col.; see Pedretti, 1957, and Kemp, 1981), but, in the absence of any direct evidence of Leonardo’s handling of terracotta, the attribution remains provisional. Among the bronzes that bear some resemblance to Leonardo’s designs, the best contenders are a series of small-scale rearing horses with riders. Two versions (Budapest, N. Mus.; Louisville, KY, Speed A. Mus.), the latter probably representing Marcus Curtius, can be related to drawings of rearing horses of the Battle of Anghiari type, such as the sheet of studies of Horses, a Cat and St George and the Dragon (Windsor Castle, Royal Lib., 12331). However, the generalized anatomy of the Budapest horse suggests that it is at best a later variant of a Leonardo design, while the Marcus Curtius has an awkwardness in proportion and balance that points to a follower, possibly Rustici.

The topic of Leonardo and sculpture, if examined through the master’s surviving work, is not encouraging, but it would be wrong to underestimate its importance. Sculptural values, particularly those of Verrocchio, exercised a notable impact on Leonardo’s vision of form and communication in space, and some of his own ideas were absorbed into the sculptural tradition.

Architecture.

Discussion of Leonardo’s contribution to architecture is as problematic as that of his contribution to sculpture. Although he may never have built anything, documents and drawings contain tantalizing glimpses of unrealized projects and brilliant inventions, and in his letter to Ludovico Sforza he claimed to be the equal of anyone in architecture, capable of designing public and private buildings. His architecture may be described as being in the spirit of Brunelleschi, combining a reverence for the proportional principles of antique buildings (as expounded by Vitruvius) with a relatively undogmatic use of the Classical vocabulary and an inventive ingenuity in matters of engineering.

Leonardo’s architectural projects were of two kinds. The first were practical, completing or renovating extant buildings and working as a military architect; the second were theoretical and included schemes for ideal cities and plans for many types of building. It is possible that the designs in Codex B (Paris, Inst. France; see fig.) were intended to initiate a treatise on architectural building types, while the first Madrid manuscript (Madrid, Bib. N.) deals with principles of construction. The treatment of churches in Codex B, consisting of illustrations with only a minimal commentary, may have influenced later writers on architecture, particularly Serlio. Leonardo’s architectural drawings, together with those of Francesco di Giorgio Martini (one of whose architectural manuscripts Leonardo owned) and of Giuliano da Sangallo, are among the earliest known. Since no drawings of this date survive by Bramante (who worked in Milan as court architect alongside Leonardo for 19 years), they are crucial in illustrating the evolution of the High Renaissance style.

Leonardo’s series of structural studies for the tiburio over the crossing of Milan Cathedral (1487–8; e.g. Cod. Atlantico, fol. 310rb; Cod. Trivulziano) show that he attempted to devise an architectural ‘skeleton’ that had more affinities with the principles of Gothic ribs than with Roman structures, while the shape of his dome pays obvious homage to Brunelleschi’s at Florence Cathedral (see fig.).

The fundamental principle behind Leonardo’s schemes of urban planning, developed in connection with schemes for Milan and Vigevano, was to devise a functioning ‘body’ in which canals, roads and pavements would permit an efficient and healthy environment, with a highly organized stratification of social activities. The designs in Codex B (fols 16r, 37v) illustrate ideal schemes for raised pedestrian precincts and a subterranean canal system that had little hope of realization, but the planning undertaken by Ludovico Sforza at Vigevano in the 1490s appears to reflect the translation of Leonardo’s ideas into reality.

The most impressive and coherent set of Leonardo’s architectural drawings is the series of church designs in Codex B (fols 17v, 18v, 21r) and Ashburnham I (Paris, Inst. France, MS. B.N. 2037), which show his variations on centralized and Latin cross plans. Some represent free experiments with a variety of geometrical schemes, while a few (e.g. Ashburnham I, fol. 5v) depict relatively resolved structures in which complex aggregations of plastic form are erected over intricate geometrical ground-plans. The more compact of the centralized structures, relating to Brunelleschi’s unfinished S Maria degli Angeli in Florence, may have been intended as a Sforza mausoleum on a limited scale rather than as a full-sized church.

Leonardo’s inventiveness as a military architect was given full expression when he was in service with Cesare Borgia in 1502–3. His mission to Piombino on behalf of the Florentine government appears to have stimulated some remarkable schemes for fortified structures, ranging from projects for specific locations to great, ideal schemes for impregnable fortresses (e.g. Cod. Atlantico, fol. 41va, 48rb). Massive, block-like structures, with curved or slanting profiles to deflect bombardments, are disposed around circular or polygonal plans, with elaborate passages for the internal circulation of forces. Although the grander schemes inevitably remained unrealized, it is reasonable to think that Leonardo’s advice on modifications to existing structures were taken in hand by his patrons.

The grandest of Leonardo’s plans for a residential structure dates from the last years of his career, when he was in France. His scheme for the château at Romorantin, involving a large rectangular palace block, formal gardens and a rectangular network of canals, has been reconstructed by Pedretti (1972) from a group of sketches (especially Cod. Atlantico, fol. 76vb; London, BL, MS. Arundel, fol. 270r). The style appears to marry indigenous French elements, such as the round corner towers, with Italian Renaissance elements in a way that is typical of Leonardo’s undogmatic exploitation of Classical vocabulary. His sense of form and function ultimately took precedence over strict allegiance to Classical rules.

5. Ephemeral designs.

A significant factor in Leonardo’s value to his courtly patrons was his ability to organize visual entertainments, particularly those connected with celebrations. His involvement with the design of courtly ephemera is relatively well documented, but the visual record is meagre. Only a few drawings in his surviving notebooks relate directly to known schemes. The most substantial autograph record relates to a pageant on the occasion of the weddings of Ludovico and Anna Sforza to Beatrice and Alfonso d’Este in 1491. A drawing of a richly caparisoned horse is accompanied by a note that explains an astonishingly rich series of symbolic allusions involving peacock feathers, a wheel of fortune and the Cardinal Virtues (MS. Arundel, fol. 250r). His drawings for allegorical compositions illustrating the Sforzas’ reign show a comparable elaboration of arcane allusions, extreme even by Renaissance standards.

Leonardo was also involved in theatrical design, specializing in effects that required large-scale machinery. His drawings for ‘Pluto’s Paradise’ (MS. Arundel, fol. 231v) show a scheme for the opening of a mountain to reveal Pluto and his attendants who play harsh percussion instruments. As a musician of some reputation himself (he played the lira di braccio), Leonardo was well placed to design effects that would work in concert with instrumental and vocal compositions. Contemporary accounts survive of his most famed design for Bellincioni’s Festa del paradiso, which involved a great, glowing celestial hemisphere adorned with stars and planets. A slight drawing (Cod. Atlantico, fol. 385vb), apparently dating from this late period in Milan in the 1490s, suggests that he should also be credited with the invention of the shallow, perspectival stage set normally associated with Baldassare Peruzzi and illustrated in Sebastiano Serlio’s treatise on architecture (1537–51).

Evidence of his more general work as a designer of courtly diversions is fragmentary. His notebooks contain designs for pictograms (picture writing), portable pavilions, festive architecture, automata, fountains and written outlines for amusing contrivances. Such employment was both advantageous and irksome for Leonardo: advantageous in that it helped to justify his salary at court, but irksome in that it occupied considerable amounts of time with no enduring result.