Masaccio 1401–1428/9(?)
NG 3046 
The Virgin and Child

Dillian Gordon, 2003

Extracted from:
Dillian Gordon, The Fifteenth Century Italian Paintings, Volume I (London: National Gallery Company and Yale University Press, 2003).

1426

Egg tempera (identified)¹ on wood, 139.0 x 76.0 cm

The Virgin is seated on a throne of grey stone (pietra serena) ornamented with classical columns with capitals of the Ionic, Corinthian and Composite orders. She is clothed in a voluminous blue cloak, with a transparent veil over her head and hair, and wearing a dress of red glazed over silver. On her knee is the naked Child, sucking his fingers and reaching for the bunch of grapes which his mother holds in her hand. At the base of the throne two seated angels are playing lutes, and at the sides kneel two more angels with their hands joined in prayer. This was the central panel of an altarpiece painted by Masaccio in 1426 for the family chapel of Ser Giuliano degli Scarsi in Santa Maria del Carmine, Pisa.

Technical Notes²

Panel construction

Height, including false strips, 137.5 cm on left edge, 139.2 cm on right edge; width, including false strips, 76.0 cm. Painted surface: height on left edge 134.8 cm, on right edge 136.0 cm; width 73.5 cm. Thickness of panel 3.0/3.5 cm.

The grain of the wood runs vertically. The panel, constructed of three planks, has two vertical joins: the left plank (seen from the back) is 15.1/15.4 cm wide, the centre plank 46.1/46.4 cm, and the right‐hand plank 11.4/11.7 cm. All the butterfly keys are modern and were inserted in 1920 (see conservation dossier). The present frame is modern.³ The panel has been cut on all sides. The tip of the exposed arch left by the removal of the original frame is missing, but the gilded background finishes in a point. The arcs of the pointed arch show that the panel has not been greatly reduced at the sides. On the back, a band 18.5 cm high, situated 14 cm from the top, has been planed level to accommodate a horizontal batten. A second recess 10 cm high runs along the base, which indicates that the panel has lost probably at least about 8 cm of planed area, plus possibly several more centimetres of panel from the bottom. This would allow space to complete the feet of the seated angels. When the bottom edge of the panel was cut off, a great deal of paint was lost. This has been made up with a band of later grey and brown paint.

Four nails fixing the battens along the back were hammered in at regular intervals from the front. The two nail‐heads under the picture area are covered with diamond‐shaped patches, probably parchment⁴ (visible in the X‐radiograph, fig. 1).

A space was left for a capital at each side, as indicated by renewed gold at capital level and the edge of the canvas at the left side (visible in the X‐radiograph), which was not taken up over the frame.

Condition and technique

The history of the panel is complicated, as several campaigns of restoration have evidently been carried out. All the picture sight surfaces of the surviving fragments of the altarpiece, including NG 3046, were at one stage made up into rectangles and this would seem to account for the strip of gilded gesso on the left‐hand side, which runs under a later gilding of the spandrels. This was probably done before 1640, since by then the altarpiece had been dismembered and probably dispersed. Subsequently the pictures were restored to their present shapes. In NG 3046 the border around the spandrels was regilded and toned. The innermost punched part of the gilded spandrels is original: the gold leaf is the same as that of the main panel. The punching in the left and right spandrels has been done with different tools and follows a different approach: the foliage in the right‐hand spandrel has been outlined with a small star punch and the intervening gaps then densely punched, whereas the stippling in the left‐hand spandrel merely fills the gaps. If the two spandrels had been a later restoration, they would probably have been identical.

Numerous incised lines have been drawn for the architecture of the throne. In the throne step, indented dots mark off the spacing between the strigillations; these dots do not appear in the sections to the left and right sides between the angels and the side of the panel, where the strigillations become wider and change direction.

A small amount of underdrawing has been detected with infra‐red reflectography (fig. 2). The Virgin’s right eye was originally drawn lower down, and smaller. The painted surface is much damaged, as is patently visible in the infra‐red photograph and the X‐radiograph (fig. 1).

The Virgin and Child: The flesh in both figures (see detail opposite) is considerably abraded, and the paint is particularly thin at the junction between highlight and mid‐tone, resulting in false highlights of exposed gesso. The green earth shows through more than was originally intended – for example, in the centre of the Virgin’s right cheek. The Virgin’s hair is abraded, especially in the strands superimposed on the robe to the right of her left eye. So, too, is the lead white of her veil. The blue of her cloak is disfigured by discoloured retouchings which seem to have been made with Prussian blue (a pigment introduced in the early eighteenth century). The reconstruction of the missing parts of the Child’s feet and the Virgin’s left hand has darkened.

The mordant gilding around the Virgin’s neckline is well preserved and has retained the red lake glazes indicating shadow at the left (fig. 3). Only small fragments survive of the original line of mordant gilding on the blue cloak over her head; that around her cuffs is completely modern and that of the blue cloak has been reinforced in parts. The gold hem along the bottom is largely original, with small passages of restoration.

Much of her dress, which is of silver leaf glazed with red lake and incised with small vertical strokes, probably intended to suggest a textured fabric,⁵ has survived, but most of the red of her left sleeve is restored. This sleeve was never covered with silver leaf – perhaps forgotten.

Her halo, which is in good condition, contains lettering in pseudo‐kufic approximating very vaguely to the words AVE GRATIA PLENA. The Child’s halo is also well preserved, with only some losses of the red paint in the lozenges; its front edge is more densely punched in order to create the seeming rim of a disc. The reflection of the Child’s hair in the golden disc is indicated with yellow‐brown paint.

The kneeling angels: The left‐hand angel is in reasonable condition: the face is relatively well preserved, but the hands are damaged. The robe has only a few retouchings, and the mordant gilding around the neck is original. The face of the right‐hand kneeling angel is probably the best preserved part of all the angels, and the hands, mordant gilding and robe are all in good condition except for a damage down the edge of the panel, which has also affected the back of the head.

In both kneeling angels, the wings and haloes are entirely false: they go over new gold and the haloes are punched into new gold.

The seated angels: The seated angel on the left has suffered considerable damage: most of the left side of the face, the whole of the left shoulder and most of the upper arm have been repainted. The halo is regilded. Both hands and the lute are relatively well preserved, particularly the left hand with the outlining of the fingernails, and the plectrum is still visible.

The face of the right‐hand angel (see detail opposite) is fairly well preserved with the exception of the profile, which has been affected by the regilding of the halo, as has the outline of the hair. The right hand is extensively damaged, the left hand better preserved. The robe below the slit has been repainted, falsifying the form of the leg and the fall of the drapery. The neckline has been water‐gilded, painted with ochre and incised freehand. The lute is comparatively well preserved. In both angels the wings retain their original painted feathers, although there are some retouchings.

The throne: The throne is fairly abraded and has suffered large losses. Much of the detail of the architecture has been lost, particularly the bottom right‐hand capital. The base of the throne has suffered few losses but is somewhat rubbed.

Attribution and Identification of the surviving parts of the altarpiece

It is accepted on stylistic grounds that NG 3046 is by Masaccio, and that it formed the centre of an altarpiece seen and described by Vasari in the church of Santa Maria del Carmine, Pisa, before it was dismembered.

Vasari referred to the altarpiece in general terms in the 1550 edition of the Vite.⁶ However, the 1568 edition contains a more detailed account, and it is this which provides the basis for identifying the surviving panels and reconstructing the original appearance of the altarpiece:⁷ Nella chiesa del Carmine di Pisa, in una tavola che è dentro a una cappella del tramezzo, è una Nostra Donna col Figliuolo, ed a’ piedi sono alcuni Angioletti che suonano: uno de’ quali, sonando un liuto, porge con attenzione l’orecchio all’armonia di quel suono. Mettono in mezzo la Nostra Donna, San Pietro, San Giovanni Battista, San Giuliano e San Niccolò; figure tutte molto pronte e vivaci. Sotto, nella predella, sono di figure piccole storie della vita di quei Santi, e nel Mezzo i tre Magi che offeriscono a Cristo; ed in questa parte sono alcuni cavalli ritratti dal vivo tanto belli, che non si può meglio desiderare; e gli uomini della corte di que’ tre re sono vestiti di vari abiti che si usavano in que’tempi. E sopra, per finimento di detta tavola, sono in più quadri molti Santi intorno a un Crucifisso.(In the church of the Carmine in Pisa, in a panel which is in a chapel of the choir screen, is a Virgin and Child, and at her feet are some little angels playing music; one of them, playing a lute, is listening intently to the harmony of the sound. In the centre are the Virgin, Saint Peter, Saint John the Baptist, Saint Julian and Saint Nicholas; the figures are very alert and lively. Below, in the predella, are small figures with stories from the lives of those saints, and in the centre the three Magi making their offerings to Christ; and in this part are some depictions of horses done from life and very beautiful, so that one could not desire better; and the courtiers of the three kings are dressed in the fashions of those times. And above, completing the altarpiece, are many saints in several panels around a Crucifixion.) It is also accepted that the altarpiece is identifiable with that commissioned from ‘Maestro Mazo dipintore di Iohanni da Firenze del popolo di San Michele Bisdomini’ (that is, Masaccio) for the family funerary chapel in Santa Maria del Carmine of the Pisan notary Giuliano di Colino degli Scarsi da San Giusto (1369–1456), who kept detailed records of the payments for the commission.⁸ No contract survives, but according to Ser Giuliano’s records, Masaccio began the altarpiece on 19 February 1426 and the first payment for it was made on 20 February 1426,⁹ the final one on 26 December 1427 (1426 modern).¹⁰

NG 3046 was first published by Berenson in 1907¹¹ as being the centre of the altarpiece seen by Vasari in the Carmine, and this has since been accepted by all scholars, bar two.¹² The four full‐length standing saints on either side of the Virgin and Child, identified by Vasari as Peter, John the Baptist, Julian and Nicholas, are all now lost.

The predella has been identified as three panels: the Adoration of the Magi (fig. 6), singled out for admiration by Vasari, and two panels showing paired scenes corresponding, as Vasari said, to the saints in the main tier – The Crucifixion of Saint Peter and The Beheading of Saint John the Baptist (fig. 8) and Saint Julian killing his Parents and Saint Nicholas dowering the Three Daughters (fig. 7).¹³ (All three panels are now in the Gemäldegalerie, Berlin, inv. nos 58A, 58B and 58E.)

Vasari describes the upper tier as the Crucifixion surrounded by several saints: the Crucifixion (fig. 9) is now in the Museo di Capodimonte, Naples (inv. no. 36),¹⁴ the extreme foreshortening of the body of Christ confirming that it was an upper tier panel. The structure of the panel is identical to that of NG 3046, in that it consists of three planks of which the width of the central one is identical to the width of the central plank of NG 3046 (46 cm). X‐radiographs confirm that the vertical pattern of the grain of the wood is continuous through both panels and that the Crucifixion and the Virgin and Child were painted on the same planks of wood.¹⁵

Although Vasari does not identify the saints surrounding the Crucifixion, it is accepted by most scholars that two saints of similar format to that panel came from the upper tier:¹⁶ Saint Paul (fig. 10), Pisa, Museo Nazionale di San Matteo (inv. no. 27)¹⁷ and Saint Andrew (fig. 11), Los Angeles, J. Paul Getty Museum (inv. no. 79.PB.61).¹⁸ None of the other upper tier saints, who probably numbered four, is known to survive. It has been proposed by Christa Gardner von Teuffel that the missing figures could be Saint James, for Giuliano’s wife, Iacopa, and Saint Mark, included for Giuliano’s nephew, who, together with Giuliano, is thought to be one of the two lay figures in the predella (see below).¹⁹ Four small panels with Saint Augustine (fig. 15) and Saint Jerome (fig. 12) and two [page 209] unidentified Carmelite saints (figs 13 and 14), now in Berlin, Gemäldegalerie (inv. no. 58D), are also thought to have come from the altarpiece, although opinions differ as to their position.²⁰ Missing are probably the other two Doctors of the Church, Saints Gregory and Ambrose.

The total height of the predella, Virgin and Child and Crucifixion panels combined gives a minimum projected height of about 262.6 cm. The total height of the altar and altarpiece together is given in the documents as 8⅔ braccia, that is, about 506 cm; Gardner von Teuffel estimates the average Tuscan altar as about 120 cm high, plus altar‐step 17 cm. This allows approximately 106 cm for framing elements, and an inscription socle.²¹ The predella gives a minimum width of 195 cm.²²

NG 3627 has also been said to have come from the Pisa altarpiece, although this is unlikely.²³

The Original Appearance of the Altarpiece

The exact date when the altarpiece was dismembered is unknown. It must have been after 1568, when Vasari described it in such detail – possibly before 1574, when the church was being remodelled.²⁴ The altarpiece is not mentioned in the description of the Carmine in Pisa by Tronci, c. 1635,²⁵ and the Saint Paul panel, which was itself first documented in the Tronci Collection in 1640, was said to have been previously in another private collection, implying the passage of years since the dismemberment.²⁶

The component panels of the altarpiece are not generally doubted. However, the form the altarpiece took – whether it was a multi‐tiered polyptych of several compartments or an altarpiece with a unified main picture surface – and the related question of how the upper tier was completed have been the subject of debate.²⁷ It is curious that no altarpiece survives, [page 210][page 211] either preceding or subsequent, which appears precisely to correspond with any of the proposals put forward for the original design. Relevant to this issue is whether Masaccio was or was not involved in the design ab initio.

Most reconstructions of the altarpiece have proposed that it was a polyptych with the main tier divided into three distinct parts.²⁸ For example, Mario Salmi in 1948 and Ugo Procacci in 1951 showed the main tier side saints as paired under a double arch, each pair surmounted by two half‐length saints, including the surviving Saint Paul and Saint Andrew on either side of the Crucifixion, the small‐scale saints in the lateral pilasters, and Scenes from the Lives of Saints Peter and John the Baptist and Scenes from the Lives of Julian and Nicholas on the left and right sides respectively of the Adoration of the Magi (fig. 16).

The arguments for the altarpiece having had a unified pictorial surface were put forward by John Shearman in 1966 (fig. 17).²⁹ He argued that the Virgin and Child panel was originally part of a single unified surface because NG 3046 has been cut on all sides, that the horizontal batten helped to hold firm the full width of the altarpiece, and that the centre was marked off by pendant capitals at the sides. He considered that the unified space and light with cast shadows drew the centre figures and side saints into a form of sacra conversazione, which was later to become common. He showed the throne steps extending across the entire picture surface, and the main tier saints spreading down the steps on different levels. Shearman conceded that reconstruction of the upper tier is ‘wholly mysterious’ and that the ‘upper profile of the whole altar [piece] was substantially gothic in form, with a broken cresting of [page 212] semi‐isolated panels punctuated by finials’.³⁰ Everett Fahy has suggested that a small panel (81 x 43 cm) showing the Virgin and Child with Saints Paul, Nicholas of Bari, Catherine of Alexandria(?), Francis, John the Baptist, and Peter with the Crucifixion above (Athens, Byzantine Museum), attributed to Borghese di Piero, was influenced by Masaccio’s altarpiece, particularly in its design, and that it confirms Shearman’s hypothesis.³¹

However, Gardner von Teuffel argued, against Shearman, that the altarpiece was a traditional multi‐tiered polyptych, while pointing out that one cannot be sure of its original appearance. Her reconstruction represents a refinement of that made by Salmi, based on a detailed analysis of the measurements, examination of the carpentry of the panels, the documents, an analysis of Masaccio’s compositional methods, and placing the altarpiece in the context of contemporary Tuscan altarpiece design. She outlined the various possibilities for a reconstruction of the altarpiece as a unified surface beneath a corbelled arcade, or with saints paired under arcades, or as separate figures in a pentaptych.³² She argued persuasively for the central panel having been separated from the side compartments by some sort of framing elements, noting the self‐enclosed design and the niche‐like effect of the throne (although her argument that the central composition is self‐contained because of the upward‐folding angels’ wings is now negated by the fact that the wings have been found to be false).³³ She put forward the possibility of separate figures in a pentaptych, but also wrote that ‘we cannot be sure if the saints were themselves in separate picture fields or divided by consoles’,³⁴ and ‘that discrete vertical members were set over a unified panel after the completion of the painting… cannot be excluded’. She pointed out that the incidence of the incised parallel and diagonal lines of the throne must have occurred beyond the present picture surface, and therefore considered it likely (although not certain) that the throne step appeared completed in the side panels.³⁵ She also noted that ‘it is very rare to find a simple [read ‘single’] half‐length figure in the upper zone surmounting paired full‐length saints in the main register. This convention suggests that there were originally four half‐length figures above four distinctive side divisions in the main storey.’³⁶ Gardner von Teuffel suggested that five upper pinnacle panels complemented the five surviving predella scenes, and were either separate from each other and narrower than the main tier panels below, or of the same width and contiguous.³⁷

Fresh examination of the surviving panels has made it possible to suggest that the form of the altarpiece is more likely to have been a triptych and to propose some refinements to the reconstruction (fig. 18).³⁸

The predella

The predella panels were not painted on a single plank of wood, and the wood grain is not continuous. The top frame moulding of the panels was formed of rectangular dentellations, visible in the X‐radiographs.³⁹ X‐radiographs of the panels (fig. 5) reveal where nails originally secured the planks to vertical struts or to one horizontal batten, fixing them within a containing frame. Because it was not painted on a single plank, the predella cannot be reconstructed with absolute certainty. However, the distribution of the nails⁴⁰ suggests that the panels were positioned as follows: left, Scenes from the Lives of Saints Julian and Nicholas; centre, the Adoration; right, Scenes from the Lives of Saints Peter and John the Baptist[page 215](fig. 4). This sequence places the titular saints of the altar on the Virgin’s proper right.⁴¹ Although John the Baptist is usually found on the immediate left or right of the Virgin, three contemporary altarpieces show him positioned at the outside, in all three cases paired on a single panel with another saint, and therefore in a position which has never altered.⁴² In previous reconstructions the predella has been reconstructed with the Scenes from the Lives of Saints Peter and John the Baptist on the left (the Virgin’s proper right), and the Scenes from the Lives of Saints Julian and Nicholas on the right (the Virgin’s proper left), presumably on the assumption that Vasari was describing the saints in a sequence from left to right. However, placing Saints Julian and Nicholas on the Virgin’s proper right would fit with Vasari’s sequence if he was describing the saints in pairs from the centre outwards to the right, and then from the centre outwards to the left – that is, Saints Peter and John the Baptist (to the right) and Saints Julian and Nicholas (to the left). Vasari names the saints from the centre outwards when describing the Trinity altarpiece by Pesellino (see p. 287, note 94), the altarpiece then in the Certosa which he attributes to Fra Angelico (it is in fact by Starnina), and Fra Angelico’s San Pier Martire altarpiece. All these altarpieces show the saints in positions which cannot have been changed.⁴³

The main tier and the pinnacles

The fact that the panels with the Virgin and Child and the Crucifixion were painted on the same planks, together with the position of the joins of the three planks that make up the central panel, with the widest plank centrally placed, suggests that this was an independent section,⁴⁴ with columns applied to the surface after final assembly of the altarpiece. The corners of the throne steps and probably the angels’ wings (assuming that they had wings like those of the seated angels) would have encroached into the painted area of the side panels.

The two main tier side panels, each containing two saints, would have been the same width as the centre panel (about 76 cm): this is suggested by the fact that all three predella panels are approximately the same width (about 62 cm).⁴⁵ The fact that the paired scenes from the lives of each saint are separated only by a thin gold band, rather than by a projecting frame moulding, indicates that the full‐length pair of saints above were not materially separated from each other – Vasari described them as ‘pronte e vivaci’ (‘alert and lively’); see extract above. Moreover, even allowing for the same kind of discrepancy in scale between the side saints and the Virgin as is found in the Cascia di Reggello altarpiece (fig. 19), the saints would have been implausibly thin if they had each had to be squeezed into a panel that was half the width of the central panel, as would be the case if the altarpiece had been divided into a pentaptych.

Doubts have been raised as to whether the Saint Paul and Saint Andrew panels originally belonged with the altarpiece, principally on the basis of the thinness of their planks ( c. 1.8 cm) compared with the 3.0/3.5 cm of the central planks. However, in many other respects they match the other panels of the altarpiece and seem to have been part of it. The Saint Paul and Saint Andrew pinnacle panels both have an off‐centre join. Assuming that the side panels of the main tier below were painted on the same planks as the pinnacles and that above each was a pair of pinnacle saints, the pinnacle with Saint Paul can be placed immediately to the left of the Crucifixion, and that with Saint Andrew to the right. The position of the joins in the planks indicates that the structure of the panels was probably tripartite, with a wide central plank and two narrower lateral planks, similar to the panel with the Virgin and Child and the Crucifixion. This further suggests that the pinnacle saints were contiguous, and this is confirmed by the dimensions of the two surviving pinnacle panels (each c. 35 cm wide), which roughly correspond to half the width of the predella panels (each c. 62 cm wide).

The batten marks on the back of the panel with Saint Paul cannot be matched up with the one on the back of the Crucifixion, which is toward the top.⁴⁶ Moreover, aligning the three panels showing Saint Paul, Saint Andrew and the Crucifixion, or placing them immediately side by side on the same level, would present a bizarre and incongruous juxtaposition of arches, and Saint Andrew seems to be looking up at the Crucifixion. The panels with the main tier saints were therefore lower than the central panel. It would have been most unusual for the arches of the side panels to be level with that of the central Virgin and Child.

How the altarpiece was terminated at the top seems impossible to resolve. Salmi’s second reconstruction of 1967 showed a fourth tier consisting of five crowning gables,⁴⁷ also proposed by Gardner von Teuffel. She argued that the measurements given for altar and altarpiece together could accommodate such a tier, and that it is unlikely that the altarpiece would have finished with flat pinnacles, as in Procacci’s reconstruction of 1951⁴⁸ and in Salmi’s first reconstruction.⁴⁹ Whatever the case, Berti and Foggi’s reconstruction of 1989,⁵⁰ which shows at the top of the altarpiece the four small‐scale saints (a reconstruction first tentatively suggested by Shearman),⁵¹ is probably wrong. These small‐scale saints are more likely to have come from lateral pilasters, as in numerous contemporary Tuscan altarpieces, than to have perched on the very top of the altarpiece. Not only are there no precedents for such an arrangement in Tuscan altarpiece design, but the fact that three of the saints face to the right and only one to the left results in an extremely awkward and unbalanced design.⁵²

The altarpiece was probably a hybrid, halfway between a Gothic polyptych and a Renaissance pala with a unified surface. It was probably constructed in three vertical units and one horizontal unit (the predella panels united in a single frame), easy to transport, and not impossibly painted in Florence and shipped to Pisa along the Arno (although see also below).

The decision about the form of the wooden structure of the altarpiece was probably arrived at by members of the Order, together with the patron, painter and carpenter.⁵³ Indeed, the involvement of the prior, Fra Antonio, is documented twice: the completed altarpiece had to meet with his approval, as did the marble vault.⁵⁴ The altarpiece was carpentered by a Sienese carpenter resident in Pisa, a certain Maestro Antonio di Biagio da Siena della Cappella di Santa Cristina, at a cost of 18 florins;⁵⁵ in addition, he made a wooden step (‘predella’) [page 216]for the altar,⁵⁶ and also the altar frontal and altar cover. A coperchio was to go on the altar together with a painted altar frontal: ‘lo pectorale dell’altare et lo coperchio di sopra al altare, incollati et con spranghe et paracintule et chiavati in sul’altare colli orli da lato infine alle colonne dell taula’ (‘the altar frontal and the cover above, [both] glued and [strengthened] with bars and crossbars and [both] keyed to the altar with the side edges reaching to the columns of the painting’).⁵⁷ The altar frontal and cover are mentioned in two different places in Ser Giuliano’s records as being made by woodworkers.⁵⁸ Every time the coperchio is mentioned in the documents, it is in conjunction with the wooden altar frontal that was eventually carpentered by Antonio di Biagio, who made the altarpiece, and painted by Cola d’Antonio: the two items were commissioned together and delivered together, then glued and nailed to the altar, reaching to the side columns.⁵⁹ The coperchio must therefore have been made of wood.⁶⁰ Unfortunately, no altarpiece documented as having been carpentered by Antonio di Biagio survives.⁶¹ Joannides points out that the payments to the carpenter are not listed in the account books, but only in the subsequent résumé, where Ser Giuliano appears to group items by function rather than chronologically,⁶² so the fact that the payment to the carpenter for the altarpiece is listed before the payment to Masaccio does not necessarily mean that Masaccio had no hand in the design of the wooden structure. In fact, the hybrid nature of the structure rather suggests that he did.⁶³

Subject and Iconography

The choice of saints in the altarpiece was presumably determined by Ser Giuliano degli Scarsi in collaboration with the prior of the Carmine. Giuliano’s name saint was Julian, and Nicholas was the name saint of both his parents, whose remains were buried in the chapel together with those of his sister Bacciamea, his nephew Marco, his daughter Nanna, and his nieces Teccia and Lucretia. It is worth noting that the scenes chosen from the lives of Saints Julian and Nicholas both involve ‘family’ incidents, connected with parents and with daughters. The altar itself was dedicated to Saints Julian and Nicholas.⁶⁴ It has been suggested that John the Baptist, patron saint of Florence, was included because Florence at that time controlled Pisa, and that Peter was included as a reference either to Giuliano’s grandfather or to the papacy.⁶⁵ Eliot Rowlands has suggested that Saint Andrew was chosen for Andreuccia, Giuliano’s sister‐in‐law, and that John the Baptist held a special relevance for the Carmelites.⁶⁶ It has also been proposed that Giuliano and his nephew Marco, who was also a notary (d. 1421), are the two figures in contemporary dress in the Adoration of the Magi.⁶⁷ This would have obviated the need to include Giuliano as a miniature and therefore unrealistic donor figure in the main panel, which was still a current device, as can be seen, for example, in an altarpiece of 1426 by Bicci di Lorenzo.⁶⁸ The small‐scale Carmelite saints thought to be from the pilasters localise the altarpiece within the convent of Santa Maria del Carmine.⁶⁹

Although Gardner von Teuffel fully explores the significance of the eucharistic symbol of the grapes in the context of the theme of the Virgin as a vine in Carmelite litany,⁷⁰ she also points out that the same motif is found earlier in the Cascia di Reggello altarpiece of 1422 (fig. 19).⁷¹ This has no Carmelite connections, and indeed the several surviving examples of the motif suggest that it evolved independently of a Carmelite context, even if it was chosen in this instance as peculiarly apposite to the Carmelites’ liturgy. The motif of the Child sucking his index finger first occurs in Giotto’s Stefaneschi altarpiece (Rome, Pinacoteca Vaticana) and it may have been a Florentine motif adopted by Ambrogio Lorenzetti in his altarpiece of the Purification of the Virgin (Florence, Uffizi) painted for Siena Duomo in 1342.⁷² In the Cascia di Reggello altarpiece it was extended by Masaccio to involve the eating of grapes. A roughly contemporary example of the Child putting his fingers in his mouth while eating grapes occurs in a painting of the Virgin and Child attributed to Arcangelo di Cola da Camerino (Bibbiena, Propositura dei Santi Ippolito e Donato), where the Child is standing on his mother’s lap with one hand in his mouth, the other reaching down for the bunch of grapes she holds in her left hand.⁷³ Maetzke⁷⁴ has related this to the Cascia di Reggello altarpiece rather than to the Pisa polyptych. It is a difficult painting to place chronologically: Berti assumes that it comes after the Pisa polyptych.⁷⁵ The motif of the Child with grapes is also found in a painting (private collection) seemingly heavily influenced by the Pisa Virgin and Child. This is sometimes considered to be an early work by Fra Angelico, although not all scholars agree with the attribution;⁷⁶ in the painting the Child is reaching out for a bunch of grapes held by the Virgin, but leaning back at the same time as if reluctant to take them.

The eucharistic symbolism of the grapes is part of the strong symbolism that runs vertically through the altarpiece, from the Crucifixion with the Tree of Life in the upper tier, through the main tier with the Child eating grapes (symbols of his blood shed on the cross), to the throne step, which alludes to Roman and Early Christian sarcophagi (see below). This sort of symbolism is found also in Masaccio’s fresco of the Trinity in Santa Maria Novella.⁷⁷

Location, Context and Commission

The altarpiece by Masaccio was part of a comprehensive programme undertaken in the Carmine in Pisa between 1414 (1415 Pisan style) and 1429 by Giuliano degli Scarsi, which involved the construction of a family burial chapel by Pippo di Giovanni di Ghante, an altar with a consecrated stone, a curtain painted by Mariano di Maestro Piero della Valensana, a wooden step, iron torch‐holders painted red with yellow stripes, and seats to the sides and in front of the altar. A wooden cover or coperchio (see above), as well as a wooden altar frontal painted red with a surrounding fringe, containing the half‐length figure of Saint Julian within a quatrefoil, painted by Cola d’Antonio, also decorated the altar.⁷⁸ Because the Carmine – a single‐aisled, extremely tall church with main chapel and two side chapels at the east end – was remodelled in the late sixteenth century,⁷⁹ it is impossible to be certain of the exact location and form of Ser Giuliano’s chapel.

Gardner von Teuffel has argued (probably wrongly) that the chapel was one still extant to the right of the choir. She writes that the ‘most probable interpretation of “cappella” in the Pisan documents would appear to be a marble framework around the wooden altarpiece set within a pre‐existing architectural structure’. The context of the documents she sees as indicating an identification of this architectural setting as the chapel to the right of the main choir acquired by Ser Giuliano after 1414.⁸⁰ However, it seems more likely that the ‘cappella’ made by Pippo di Giovanni di Ghante was not a marble framework, but a small marble vaulted structure projecting into the nave, placed on the tramezzo (choir screen) next to the choir,⁸¹ as argued by Hall on the basis of Vasari’s statement that the altarpiece was in ‘una cappella del tramezzo’.⁸² According to Giuliano’s account books, the chapel, which was constructed by Pippo at an initial cost of 100 florins, rising to a final 140 florins, was begun on 29 November 1425. The vault was made by the mason Bartholomeo da Montemagno. Pippo di Ghante provided a drawing of the proposed structure, which had been approved by the prior, Fra Antonio.⁸³ The chapel was to be ‘nella nave d’essa chieza dirimpetto al choro, cio[è] colle reni al choro, che duri dalla porta, cio[è] dalla [e]ntrata del coro, insine al muro laterale della chieza di verso mezo dì’ (‘in the nave of this church in front of the choir, that is, with its back to the choir, extending from the door, that is, from the entrance to the choir, up to the side wall of the church on the south side’).⁸⁴ It consisted of a white marble groin‐vaulted structure resting on two columns and a half‐column above an altar: ‘una volta cio[è] due crociere in su due colonne e mezo di marmo bianche’ (‘a vault, that is to say, two crossings above two columns and a half [column] in white marble’), described elsewhere as ‘una crociera di marmo in colonne di marmo biancho, alta che vi chappia sotto una taula sopra l’altare che è alta la taula coll’altare braccia octo e due terzi, o circa’ (‘a marble vault [supported] on white marble columns, high enough to surmount the altarpiece above the altar, and the altar with the painting are together eight and two thirds braccia [high] or thereabouts’).⁸⁵ The payments made on 17 March 1427 to Pippo for moving the sepulchral monument to the family chapel also suggest that it was in the nave set against the tramezzo, since the original monument had to be moved from the centre of the church to the side, a little further forward, but still nearby: ‘Lo quale monumento era posto nel mezo proprio d’essa chieza, donde io il feci levare et fare dinanzi al mio altare quine presso, un pocho più da lato et più inanzi a quello’ (‘The which monument was in the exact centre of the church, whence I had it taken up and replaced in front of my altar which is nearby, a little more to the side and a little further forward’).⁸⁶ Joannides makes the point that the light in the altarpiece, which emphatically comes from the left, would have been inappropriate in a chapel to the right of the choir, but appropriate on a tramezzo lit from the clerestory on the north side of the nave.⁸⁷

The first payment to Masaccio (10 florins) was made in Pisa in the presence of the prior of the Carmine, with Masaccio himself present, on 20 February 1426,⁸⁸ and a month later, on 23 March, also in Pisa, he was paid 15 florins. On 24 July 1426 (1427 Pisan style) he received 10 florins; on that occasion he was with Donatello, who was working in Pisa (see below). On 15 October, when he was paid 25 florins, Masaccio promised not to undertake any other work until he had completed the commission, which suggests that he may have [page 218]been engaged on another project; this payment was received by his brother Victorio, called Giovanni (Lo Scheggia),⁸⁹ and the mason Maestro Leonardo. On 9 November 1426 Masaccio received three lire, which he owed to a tailor. On 18 December 1426 he was present in Pisa when he received a payment of four lire. On 24 December 1426 his garzone, Andrea di Giusto, received the sum of eight lire and five soldi, and on 26 December Masaccio received the final payment of 16 florins and 15 soldi to complete the payment of 80 florins initially agreed. He received this in the Chapter House of the Carmine in Pisa in the presence of the prior, on what one may assume to have been a fairly formal occasion when he agreed to hand over the altarpiece after its approval by the prior.⁹⁰

Masaccio and his Workshop

The Pisa altarpiece is the only surviving one by Masaccio, apart from the Cascia di Reggello altarpiece, which he painted without the collaboration of Masolino (see also p. 244): Masolino was in Hungary throughout the duration of the commission, from 1 September 1425 to around the middle of 1427.⁹¹

No one has ever questioned that Masaccio himself was responsible for the central panel. The outstanding quality is apparent, notwithstanding its damaged condition, and the central panel alone is revolutionary in many ways. It brings to fruition the nascent qualities of the Cascia altarpiece: the strong unifying directional light flooding from the left, found even in the predella;⁹² the cast shadows noted by Shearman, which are to be found not only on the throne steps, but also in the glazes over the shadowed part of the mordant‐gilt neckline of the Virgin’s dress and on the back of the throne, and in the reflection of the Christ Child’s hair in his halo; the sense of three‐dimensional space, most obvious in the (unprecedented) elliptical halo of the Child, and the classical architecture of the throne. The quality of painting in the Crucifixion and in the Adoration of the Magi is also particularly high, with dramatic lighting from the left, and again both works are attributable to Masaccio himself. Masaccio must have designed the whole ensemble, but some parts of the side predella panels show a drop in quality of execution. As stated above, the documents reveal that Masaccio’s garzone was Andrea di Giusto da Firenze, who collected a payment ‘per sua commissione’ on 24 December 1426 (1427 Pisan style), that is, when the altarpiece was nearing completion.⁹³ The painting of the Scenes from the Lives of Saints Julian and Nicholas has been attributed to Andrea by Douglas and de Nicola.⁹⁴ Some of the small‐scale figures in the Scenes from the Lives of Saints Peter and John the Baptist are comparable to those in the predella of an altarpiece attributed to him (Prato, Galleria Communale, inv. no. 8) dated 1435.⁹⁵ Andrea had been in Bicci di Lorenzo’s shop from 1420 to 1424, together with Masaccio’s brother, Victorio Lo Scheggia, documented there 1420–1, who collected a payment in Pisa in 1426⁹⁶ and may therefore also have been involved in the commission. The pilaster panel with the younger Carmelite saint has been attributed to Filippo Lippi.⁹⁷

Florence or Pisa?

One of the unknown factors in the commission is whether Masaccio painted the altarpiece in Florence or in Pisa. Sassetta painted his double‐sided altarpiece for San Francesco in Borgo San Sepolcro in Siena (see p. 347), and it is not impossible that Masaccio painted the Pisa altarpiece in Florence. Ser Giuliano’s insistence on 15 October 1426 (1427 Pisan style) that Masaccio undertake no other work until this commission was complete⁹⁸ might suggest that he was not in Pisa, possibly because he was engaged on the Carmine frescoes.⁹⁹ The nature of the wooden structure of the altarpiece proposed here would have been easily transportable. However, it seems clear from the documents that all the other wooden furnishings for the chapel were made in Pisa, and the carpenter of the altarpiece was resident there.¹⁰⁰ Is it possible that Masaccio rented the shop of the local painter Turino di Vanni in which to paint the altarpiece? The altar frontal and coperchio were originally to be made by the carpenter Bartolomeo Angeli, who had collaborated with Turino di Vanni on an altarpiece in 1403, as mentioned in a document drawn up by Ser Giuliano on 18 June 1429 (1430 Pisan style).¹⁰¹ It may be that Turino, who was aged 77, hired out the use of his shop for all the painted items for Ser Giuliano’s chapel that were to be made by Antonio di Biagio and Bartolomeo Angeli, who seem to have been compagni, and then painted by Cola d’Antonio and Masaccio respectively.

The Influence of Sculpture and Architecture

The profound influence of sculpture on Masaccio has been remarked upon by numerous writers, Beck even suggesting that Masaccio was himself a sculptor in his early years.¹⁰² It is evident in NG 3046 that Masaccio had looked at local [page 219]sculpture in Pisa. The pattern of the throne step derives from Roman and Early Christian sarcophagi,¹⁰³ of which there were numerous examples in the Pisan Camposanto (fig. 21).¹⁰⁴ Eve Borsook has pointed out the influence of the local sculptures of Giovanni Pisano, particularly in the figures of the Virgin and Child in both the main panel and the central predella,¹⁰⁵ while Joannides shows that some motifs have been taken from the Pisa pulpit of Nicola Pisano.¹⁰⁶ Burns also makes the observation that rosettes like those on the throne are to be found in the bronze doors of the Pisa Duomo.¹⁰⁷ Boskovits made the point that much of the classicising influence was transmitted through a Tuscan ‘proto Renaissance’, particularly in the sculpture of the Pisani; he makes an apt comparison with a statue of a Virgin and Child enthroned (Detroit, Ford Collection) attributed to the school of Nicola Pisano, and also notes the derivation of the throne from the Santa Cecilia altarpiece in the Uffizi.¹⁰⁸ The Corinthian and Ionic columns, and the use of pietra serena for the throne, also reflect the interest in classical architecture of the contemporary architect Brunelleschi.¹⁰⁹ Beck takes this further, noting that the classical orders are used in a way corresponding to Brunelleschi’s practice in San Lorenzo and the Ospedale degli Innocenti, and pointing out that, like Donatello, Brunelleschi was present in Pisa in 1426.¹¹⁰ In fact, the throne resembles a building more than a throne.

The Child in NG 3046 is developed from that in the Sant’Anna Metterza altarpiece in the Uffizi,¹¹¹ which Offner convincingly suggested was modelled on an Etruscan bronze, either the figure in the Museo Gregoriano in the Vatican or one similar to it.¹¹²

However, the influence of sculpture is not confined to a superficial adoption of individual motifs. The figures themselves are extraordinarily three‐dimensional. Comparisons with Donatello’s sculpture are significantly close, particularly in the upper tier figures, where Saint Paul has been compared to Donatello’s Jeremiah. Some similarities have been noted with reliefs, one in terracotta, possibly by Donatello himself, and two from his circle in marble and pigmented stucco respectively, showing the Virgin and Child with angels, and with angels and saints (fig. 20).¹¹³ Kreytenberg has suggested that, among other things, the predella of Donatello’s Saint George was an influence on the central predella panel.¹¹⁴ Donatello was working in Pisa at the same time as Masaccio,¹¹⁵ preparing the tomb of Cardinal Rinaldo Brancacci for shipment to Naples.¹¹⁶ On 24 July 1426 he was with Masaccio when he collected a payment for the Pisa altarpiece, and on 18 December 1426 he witnessed one of the payments due to Masaccio.¹¹⁷ Berti has suggested that the influence was two‐way and that Donatello was influenced by Masaccio.¹¹⁸ Sir John Pope‐Hennessy suggests that Masaccio’s Pisa Madonna was based on sculpted models made by Donatello and that this explains the payment, which he interpreted as being due to Donatello for his direct involvement.¹¹⁹

The monumentality of the figures and the coherence of the composition in NG 3046 are, however, not merely attributable to the influence of sculpture. They are also the result of Masaccio’s experience as a fresco painter and as Masolino’s collaborator in the probably more or less contemporaneous Brancacci Chapel – painting on a monumental scale with all the discipline of design, lighting and perspective that such a complex scheme entailed.¹²⁰

Provenance

Santa Maria del Carmine , Pisa, until at least 1568 (when recorded by Vasari; see above). Belonged to Miss E. Woodburn in 1855;¹²¹ Samuel Woodburn Sale, 9 June 1860 (lot 21), bought by Clarke;¹²² stated to have belonged to the Revd F.H. Sutton by 1864;¹²³ in the collection of the Revd A.F. Sutton, Brant Broughton, 1907;¹²⁴ purchased Temple‐West Fund, from Canon Sutton, with the aid of a contribution from the NACF , 1916.

Exhibited

London 1911, Grafton Gallery (7);¹²⁵ London 1945/6, NG , Exhibition in Honour of Sir Robert Witt… of the Principal Acquisitions made for the Nation through the National Art‐Collections Fund (28); London 1975, NG , The Rival of Nature – Renaissance Painting in its Context (145); London 1982, NG , Watch this Space – an Exhibition about Perspective in Painting (no catalogue); London 1985, NG , The Artist’s Eye – Francis Bacon ; London 1998, NG , Henry Moore and the National Gallery (no catalogue); London 2001, NG , Masaccio: the Pisa Altarpiece (no catalogue).

Select Bibliography

• M. Salmi, Masaccio, 2nd edn, Milan 1948, pp. 33–52, reconstruction pl. 35.
• U. Procacci, Tutta la pittura di Masaccio, Milan 1951, pp. 26–9.
• L. Berti, Masaccio, Milan 1964, pp. 44ff. and 87–9 (English trans., University Park and London 1967, pp. 85–9 and 144–5).
• M. Boskovits, ‘Giotto born again’, Zeitschrift für Kunstgeschichte, XXIX, 1966, pp. 58–66.
• J. Shearman, ‘Masaccio’s Pisa altar‐piece: an alternative reconstruction’, BM , 108, no. 762, 1966, pp. 449–55.
• L. Berti, L’Opera completa di Masaccio (presented by P. Volponi), Milan 1968, cat. no. 9, pp. 89–91.
• C. Gardner von Teuffel, ‘Masaccio and the Pisa Altarpiece: A New Approach’, JBM, 1977, pp. 23–68.
• J.H. Beck, Masaccio: The Documents, New York 1978, pp. 17–24, and appendix, pp. 31–50.
• L. Berti and R. Foggi, Masaccio. Catalogo completo dei dipinti, Florence 1989, cat. no. 6, pp. 45–74.

  

[page 220]
• J. Beck, ‘Masaccio’s Madonnas’, in Masaccio 1422/1989. Dal Trittico di San Giovenale al Restauro della Cappella Brancacci. Atti del Convegno del 22 Aprile 1989, Pieve di San Pietro a Cascia‐Reggello, Figline‐Valdarno 1990, pp. 27–35, esp. pp. 31–4; reprinted in Masaccio 1422. Il Trittico di San Giovenale e il suo tempo, ed. C. Caneva, Milan 2001, pp. 279–85.
• P. Joannides, Masaccio and Masolino. A Complete Catalogue, London 1993, pp. 153–70, and cat. no. 21, pp. 382–98.
• J.T. Spike, Masaccio, New York, London and Paris 1995, pp. 142–63, cat. no. 6, pp. 198–202.
• R. Fremantle, Masaccio: catalogo completo, Florence 1998, cat. no. 4, pp. 114–16.
• C. Strehlke, ‘The Case for Studying Masolino’s and Masaccio’s Panel Paintings in the Laboratory’, in The Panel Paintings of Masolino and Masaccio: The Role of Technique, ed. C. Strehlke with C. Frosinini, Milan 2002, pp. 12–27.
• R. Bellucci and C. Frosinini, ‘Working Together: Technique and Innovation in Masolino’s and Masaccio’s Panel Paintings’, in The Panel Paintings of Masolino and Masaccio, op. cit. , Milan 2002, pp. 28–67.
• J. Dunkerton and D. Gordon, ‘The Pisa Altarpiece’, in The Panel Paintings of Masolino and Masaccio, op. cit. , Milan 2002, pp. 88–109.
• R. Bellucci, C. Frosinini and M. Parri, ‘Technical Catalogue’, in The Panel Paintings of Masolino and Masaccio, op. cit. , Milan 2002, pp. 165–201.
• E.W. Rowlands, Masaccio’s Saint Andrew and the Pisa Altarpiece (forthcoming).

Notes