Pesellino (completed by Filippo Lippi and Workshop)
NG 727, NG 3162, NG 3230, NG 4428, NG 4868.1–4, L15 
The Trinity with Saints Mamas, James, Zeno and Jerome

Dillian Gordon, 2003

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

1455–60

Egg tempera, tempera grassa and oil (all identified) on wood, c. 184.5 x 181 cm

This was the high altarpiece commissioned for their church by the Confraternity of Priests in Pistoia from Pesellino in 1455, and completed after his death in 1457 by Filippo Lippi and his workshop in 1460. In the centre is the Trinity represented by God the Father supporting the crucified Christ, with the Dove of the Holy Spirit between them, surrounded by a mandorla of cherubim and seraphim. Two angels hover at left and right. Standing below on the left are Saint Mamas, holding a martyr’s palm and accompanied by three lions (one head and two muzzles are visible), and Saint James, carrying a book and a pilgrim’s staff. On the right are the Bishop Saint Zeno and Saint Jerome, dressed as a cardinal and carrying a book. The clouds at the base of the cross separate the group of the Trinity from the earthly landscape in which the saints stand. In the foreground the grass is covered with plants, a winding river linking it to the hills beyond.

Scenes from the lives of the four saints are in the predella beneath (a fifth panel, in another collection, is discussed below). In the first panel Saint Mamas, praying behind prison bars, is about to be attacked by four lions. In the second Saint James kneels in a landscape, about to be beheaded by an executioner. In the third Saint Zeno is exorcising a devil issuing from the mouth of the daughter of the Emperor Gallienus, watched anxiously by the emperor and his wife. In the final panel Saint Jerome is removing a thorn from the lion’s paw; a river leads to the mountainous background where two friars or monks are shown reading; a path of stone steps leads from the river up to the church on the right.

Technical Notes – Main Tier

Panel construction

Total panel (overall measurements of five sawn‐up fragments reconstructed as a single main tier): height 184.5/185 cm, width 181 cm. The main tier is made up of vertical planks. A non‐original strip of 1.25 cm runs along the top and down the left (seen from the front) edge. The fragments appear to be of their original thickness, but because of the modern frame and the metal struts and battens of the back, it has not been possible to take precise measurements of thickness.

The dimensions may have been reduced somewhat by the dismembering and rejoining of the five fragments. As is discussed below, there were originally five predella panels. The minimum projected total width of these is 221.3 cm, so the main panel may have lost something from the sides, or, more probably, the predella may have projected on either side. The main panel seems also to have been cut at the bottom, but it is impossible to see the edges. The panel has not been cut at the top, since a barbe in the gesso is visible along the top edge. Before gessoing, the panel appears to have had pieces of linen glued to it: a frayed edge is visible in an X‐radiograph detail.

The altarpiece was dismembered, probably when the Confraternity from which it came was suppressed c. 1783 (see Provenance). The main panel was sawn into five (or possibly [page 263] six – see below) pieces. The five original fragments have been joined together (see figs 1 and 2); the joins are now evident in raking light. The bottom right‐hand corner is an addition, made when the panel was reassembled in 1929/30.¹ For the dimensions of the fragments, see fig. 1.

Frame

The frame is modern but reflects the tabernacle all’antica frames of fifteenth‐century altarpieces.

Underdrawing

The entire composition was drawn in considerable and precise detail. The underdrawing (see fig. 3) plays an important role in forming the figures in the painting, in that the paint in the flesh and hair was applied in a fairly thin wash and the outlines, shadows and highlights then reinforced in a linear way (see fig. 5), with very fine and dense hatching and cross‐hatching. The figure of Saint Mamas (fig. 4) was extremely highly worked in the underdrawing, with extensive and elaborate cross‐hatching. The attention paid to this figure could perhaps be the result of the particular interests of one of the priests involved in the commission (see below).

The underdrawing is visible to the naked eye in many areas – notably, Saint Mamas’ hand, feet and hair; Saint James’s face, hair, hands and feet; Saint Jerome’s face, hands, robe and book; Saint Zeno’s hands, face and white drapery; Christ’s body; the left angel’s face and hair, and the right angel’s pink drapery. Infra‐red photography, which reveals further underdrawing, shows that there were some changes in the final painting: the green drapery of the left angel was originally [page 264] drawn fluttering above both feet, and the right foot was drawn in a slightly different position; the curl in the locks of hair below Mamas’ ear has been changed and the cross‐straps of his left leggings were originally drawn lower down; Saint James’s left shoulder was originally higher up, together with a lock of hair and the neckline of his robe; the scallop‐like folds along the bottom of his robe were considerably altered in the final painting; the position of Saint Jerome’s left arm was changed slightly. Infra‐red reflectography shows that the originally single clasp of Saint James’s book was moved to the left to make way for a pearl and a second clasp. The bottom of God the Father’s robe on the left originally extended to touch Saint James’s robe.

Condition and technique

The overall condition is surprisingly good, considering the vicissitudes the painting has undergone. There are some scattered losses, and areas of thin and rubbed paint, particularly in the light colours; the dark areas are better preserved. Prominent raised cracks run horizontally across part of the picture (see below). There are several discoloured retouchings.

All the water‐gilding is well preserved, such as the tiara of God the Father, Saint Zeno’s mitre and Saint Jerome’s crozier handle. The haloes, which are mordant gilded, are fairly well preserved, but mordant gilding has broken away in several places, such as the mandorla and God the Father’s aureole, taking with it the layer of paint underneath and leaving visible the bare gesso below.

Some aspects of colour change in the paint layers were examined by the Scientific Department, confirming fading of the red lake glazes, blackening of the vermilion, darkening of the azurite, and browning of the copper green pigments.

NG 3230: Left-hand angel

The fragment is in good condition. There is some discoloration in the green of the robe, diminishing the contrast between the green of the robe and its pinkish‐yellow lining. The angel’s wings were underpainted with azurite and then with a green glaze, which has darkened.

NG 3162: Right-hand angel

This angel is more damaged than the one on the left. The pink robe has faded dramatically; a large amount of underdrawing is visible underneath. The blue ribbons seem to have been reinforced by retouching where they cross the sky. This fragment has in the past been more drastically cleaned than the other fragments.

NG 727: The Trinity

The overall condition is reasonable. The right‐hand blue cherub has lost approximately 2 cm in the wings as a result of the joining of the fragments. Some of the pigments have changed colour, especially in the cherubim, where the azurite has darkened, and in the seraphim, where details of the feathers painted with pure vermilion have blackened.

God the Father’s greyish‐blue drapery consists of ultramarine over azurite, with the ultramarine showing some signs [page 265] of blanching, further distorted by some surface dirt. Behind his head a dark glaze has been identified as consisting of a copper green over artificial malachite and black pigment; this has discoloured and was originally a richer and darker green.

Some of the mordant gilding of the mandorla has flaked off, exposing the bare gesso ground. The top of the tiara, which is water‐gilded and painted with red lake and lead white, originally had mordant‐gilded dots on the left, which have flaked away, and painted dots on the right – in other words, dots catching the light were in gold and those in shadow were painted. Similarly, the neckline of the robe has a gilded pattern on the left and a painted pattern on the right.

In the figure of Christ, the red lake of the halo seems to have faded. Shadows are cast by the nails in the hands.

L15: Saint Mamas and Saint James

The overall condition of the fragment is good, although raised cracks are particularly disfiguring across the face of Saint Mamas and on the chest and sleeve of Saint James. Saint Mamas’ legs are in poor condition, with several scratches and gouges, and it is impossible to tell whether their flat appearance, particularly that of the left leg, is due to damage or to their having been left unfinished. The mordant‐gilded haloes have been repaired along the junction with the hair. These repairs cover locks of hair, but it is impossible to determine whether the haloes originally did so.

Saint Mamas: The yellow lining of the cloak and the green folds along the hem have a few small pentimenti; the folds have been very densely hatched and cross‐hatched in the highlights and shadows. The green is slightly abraded and at the left‐hand side the cloak is brown rather than green, since the glaze here is relatively thin and over lead‐tin yellow, while the glaze on the right passes over a darker and bluer underlayer. With time the contrast between the two has become more pronounced.

As in God the Father, the neckline has mordant‐gilded spots in the light and painted spots in the shadows on the right. Similarly, the embroidered band along the blue hem is part painted, part gilded. The belt is painted in ochre with red spots. It seems possible that it was originally intended to be mordant gilded, but was left unfinished (see below for the interruption in the execution of the painting).

Saint James: The head of the saint is in excellent condition. The red lake of the cloak may have black undermodelling in the shadows (evident in the infra‐red photograph).

NG 4428: Saint Zeno and Saint Jerome

This fragment has suffered from abrasion.

Saint Zeno: His mitre, which is in excellent condition, has been water‐gilded, painted with a white layer and then incised with a feather pattern; finally, the gold thus exposed has been punched. The border of the saint’s cope has been water‐gilded and then decorated with ultramarine and with a red lake glaze, which has faded; it has then been incised and punched. Each white pearl casts a shadow. (For the saint’s intended halo, see below.)

Saint Jerome: The figure has suffered similar abrasion to that of Saint Zeno. There is a pentimento in the position of the left arm (visible in the infra‐red photograph), in the folds of the red robe and in the position of the book. Saint Jerome’s halo has been punched, unlike all the others. The mordant gilding of the halo is applied in such a way as to appear transparent, allowing the crozier behind to be seen through it. The bottom right‐hand corner with Saint Jerome’s robe and feet is a modern reconstruction and is now somewhat darkened and discoloured.

The Landscape: In view of the fact that the landscape background runs across two different fragments, each with a different provenance, it has survived remarkably unscathed. It is in good condition, although evidently much discoloured and darkened in the greens.

Technical Notes – Predella

• NG 4868.1, Saint Mamas in Prison thrown to the Lions: 30/30.2 x 42.5 cm, thickness 1.2/1.8 cm.
• NG 4868.2, The Beheading of Saint James the Great: 30 x 42 cm, thickness 1.5/1.8 cm.
• NG 4868.3, Saint Zeno exorcising the Daughter of the Emperor Gallienus: 30 x 42 cm, thickness 1.5/1.8 cm.
• NG 4868.4, Saint Jerome and the Lion: 29.4/29.8 x 43 cm, thickness 1.5/1.8 cm.

Condition and technique

All the panels have been cut on all four edges. All have borders of modern grey paint extending up to the edges. At the sides the borders have gilding underneath, which extends diagonally across the corners. These may have been regilded. The individual scenes were probably divided by flat gold bands, although it is not impossible that they were divided by raised gilded mouldings. The damage to edges of the gesso and paint at the top and bottom of the panels was probably caused by the removal of a gilded moulding. All the panels have scattered losses and small areas of repaint.

The lines of the architecture in all the panels have been incised. All the panels have two thick lines of underdrawing [page 267] running horizontally, one across the top and one along the bottom, approximately one centimetre from the edge, visible to the naked eye at the base of the James panel and at the top of the Jerome panel.

Infra‐red photography shows that the compositions were laid out in some detail (figs 6 and 7).

NG 4868.1: Saint Mamas panel

A large knot in the wood has caused extensive and visible damage to the painted surface in the figure of Saint Mamas.

NG 4868.2: Saint James panel

X‐radiographs show where nails were driven in for the vertical battens or struts of the predella,² resulting in damage visible on the paint surface. Infra‐red photographs show pentimenti, also visible to the naked eye, in the skirt and right leg of the executioner, and in the red robe of Saint James. The face of the saint has suffered some damage. Saint James has the only halo which has rays; all the others have a punched decoration.

NG 4868.3: Saint Zeno panel

This panel, like the Saint James panel, has damages to the picture surface caused by nails driven in from the front to supporting struts. There is a pentimento in the position of the saint’s arm. The hands of the fainting girl have been repainted.

NG 4868.4: Saint Jerome panel

This panel has been extremely thinly painted and the colours may have become more translucent with time.

Reconstruction of the predella (fig. 9)

Recently the fifth predella panel has been identified.³ All the predella scenes were painted on a single plank of wood, and the break in the pattern of the wood grain visible on the back of the London panels shows that the central scene is missing. First Pèleo Bacci,⁴ then Giuseppe Marchini⁵ and most recently Jeffrey Ruda⁶ had all drawn attention to a missing panel, but without speculating on its identity or subject matter. The missing predella panel is the Vision of Saint Augustine (fig. 8), now in the Hermitage Museum, St Petersburg (inv. no. 5511).⁷ Several scholars have in the past noted the stylistic affinities of this painting with the Trinity altarpiece, but without realising that the panel was in fact part of the predella.⁸ However, not only is the panel of the same height as the predella panels in the National Gallery, but it has the same technical features, such as the horizontal lines of underdrawing across the top and bottom of each panel, described above. The design of the painting and the structure and treatment of the landscape are clearly related harmoniously to the scenes with Saint James and Saint Jerome, which would have been on either side. The subject matter relates thematically to the Trinity in the main tier above. Saint Augustine, while at work on a treatise about the Trinity, was walking on the seashore and met a child (the Christ Child), who was spooning water from the sea into a hole in the ground. Augustine pointed out the [page 268] [page 269] [page 270] absurdity of such a task, whereupon the child told him that it was even more futile for Augustine to attempt to understand the Trinity.⁹ Here the sea is shown as a river and the three blue heads at the top right‐hand corner not only represent the Trinity, but also link with the blue cherubim in the main panel, visually making the connection between the main tier and the predella.

The addition of the fifth predella panel means that the altarpiece had a minimum projected width of 221 cm, compared with the present width of the main tier, which is 181 cm. Either the predella projected approximately 20 cm on either side to support the frame, or the main panel has been cut considerably at the sides:¹⁰ it may be that Saint Jerome had a lion with him,¹¹ or possibly his cardinal’s hat, which is sometimes shown on the ground beside him; in this case it would have been behind him.¹² The width of the predella corresponds with that of the antependium (228 cm) showing the Madonna della Misericordia (see below and fig. 15), which went below,¹³ if one takes into account that some framing elements and dividing borders must be missing.

Drawings

The drawings that have been associated with the commission are: Seated Monk (fig. 13) and Saints Zeno and Mamas (fig. 10) (both Florence, Uffizi, Gabinetto dei Disegni, inv. nos 674E and 164F respectively), and the Trinity (St Petersburg, Hermitage, inv. no. 4793; fig. 11). They are discussed individually below.

History of the Commission

The altarpiece was painted for the Company/Confraternity of Priests dedicated to the Trinity (Compagnia della Trinità) [page 271]for their church in present‐day Piazza Gavinana (formerly Cino), Pistoia. The commissioning and payments are exceptionally well documented. The documents consist mainly of records kept by the treasurer in charge of the commission, the accounts of the Company, and the legal representations which followed the death of the painter.¹⁴ The sequence of events is as follows:

The decision of the decision of the compagnia

On 10 September 1455 the Company, on the initiative of the archpriest Jacopo di Bartolomeo Bellucci, decided to commission an altarpiece for their church;¹⁵ it was to cost between 150 and 200 florins, and the priests agreed to sell grain in order to pay for it (350 omine of grain were sold).¹⁶ All those present at the meeting agreed that the altarpiece should show the Trinity in the centre and two saints on each side; the three initially selected were Saint James, as patron saint of Pistoia,¹⁷ Saint Zeno, as patron saint of the clergy of Pistoia,¹⁸ and Saint Jerome. Bacci suggests that Saint Jerome was chosen to be beside Saint Zeno in homage to the canon of the cathedral, Girolamo d’Andrea Zenoni.¹⁹ It is also possible that Jerome was chosen because he was said to have had a vision of the Trinity;²⁰ he is the only one of the four main saints shown looking inward and upward at the Trinity. Saint Jerome may originally have been shown with the lion or his cardinal’s hat.²¹

Because no decision had been made on the fourth saint, the Company’s treasurer, Pero ser Landi – who, with messer Filippo di ser Giovanni, priest of Sant’Andrea, was in charge of the commission – begged to be allowed to have Saint Mamas, to whom he was devoted. This was agreed upon; Saint Mamas’ feast day was to be celebrated by the Company and Pero ser Landi promised to give 6 omine of grain every year.²² Saint Mamas may have proved acceptable because his attribute of lions balanced that of Saint Jerome. Saint Mamas was very rarely shown in Tuscan painting, but his cult was strong in the Veneto, and it is also possible that Pero ser Landi’s choice of him as the fourth saint was in some way connected with the inclusion of Saint Zeno, who, as well as being patron saint of the clergy of Pistoia, was patron saint of Verona:²³ Mamas is shown next to Saint Zeno in an altarpiece from San Pietro Martire, Verona (now in Verona, Museo di Castelvecchio, no. 373, inv. no. 152/1B 373), attributed to Giovanni Badile (1373–1451),²⁴ and Mamas and Zeno are shown together in what is presumed to be a preparatory drawing for the Trinity altarpiece (fig. 10). Mamas was an early Christian martyr, who owed his name to the fact that he called his foster mother Ammia mama. He was imprisoned and thrown to lions, although not concurrently as shown in the predella.²⁵

On 6 September, four days before the meeting of the Company, Pero ser Landi had been to Florence to fetch a drawing to show to his fellow members (‘il disegno a vedere alla Compagnia’).²⁶ O’Malley has plausibly argued that this first drawing probably showed the overall structure.²⁷

The commissioning of the altarpiece

On 17 September, one week after the meeting, Pero ser Landi went to Florence to draw up the contract and for drawings he had commissioned (‘per far scritture e per li disegni feci fare’); he claimed for the paper (‘fogli’) needed for this and gave the commission to Pesellino (‘quando alloghai la taula a Francesco Pesello maestro dipintore’).²⁸ It is reasonable to argue that one of these ‘disegni’ could have been the black chalk drawing of Saints Zeno and Mamas in Florence (fig. 10). This drawing could only have been made after the meeting of 10 September, since only then was it decided which saints were to be shown. It was Berenson who first saw this drawing [page 272] as preparatory for the altarpiece.²⁹ A number of changes were made between the drawing and the painting: Mamas’ gesture in the drawing is used for Zeno in the painting; just as Mamas carries a book in the drawing, so Jerome does in the painting; Mamas was moved to the other side in the painting and substituted by Jerome; Zeno’s face in the drawing was given to Jerome in the painting. Another preparatory drawing, the Trinity now in the Hermitage (fig. 11), was first published as being connected with the altarpiece by John Shearman.³⁰

A week later, on 24 September, Pero ser Landi paid 1 lira 2 soldi to a certain ‘magistro Lorenzo dipintore in Firenze… per sue fatiche di quello primo disegno ch’io fe’ fare’ (‘to Master Lorenzo, painter in Florence, for his trouble over that first drawing which I had made’).³¹ This first drawing may also be the one referred to as having been fetched from Florence on 6 September.

On 26 September and 8 October Pesellino was paid 20 fiorini larghi (105 lire) in two instalments,³² and on 4 December he was reimbursed one fiorino largho (5 lire 5 soldi) for the linen he had bought for the altarpiece.³³ On 26 September and 24 December, three payments amounting to 58 lire 2 soldi (= between 11 and 12 florins) were made to the carpenter Antonio Manetti da Firenze.³⁴

On 10 September 1456, exactly a year after the decision had been taken to commission the altarpiece, Pero ser Landi went to Florence ‘per la taula’;³⁵ thereafter he went to Florence almost monthly, presumably because progress was slow (or as a pretext because he enjoyed going to Florence).

The death of Pesellino

On 10 July 1457 Pesellino was too ill to continue and was paid for his work,³⁶ and on the same day there was a payment to a notary for an estimate of the state of the work made by Filippo Lippi and Domenico Veneziano and for the transporting of a panel from the carpenter (‘dello lodo fu dato per frate Filippo e Domenico da Vinegia e per portatura della taula dello legniame’).³⁷

On 29/30 July 1457 Pesellino died.³⁸ On 2 August his partner, Piero di Lorenzo, asked Jacopo Bracciolini, the Vicar of the Bishop of Pistoia, Donato de’ Medici, that payment for the altarpiece be suspended or put on deposit until he had sorted out a disagreement with Pesellino’s heirs, from whom he was claiming half the payment. He said that the arbitrators, whom he did not name, had judged the painting to be half‐finished (‘detti albitri g[i]udichorono fussi fatta la metà’); he also said that Piero di Cosimo de’ Medici had originally been going to judge how much of the sum of between 115 and 200 florins allocated for the payment of the painting should be paid.³⁹

On 20 and 27 September 1457 Piero di Lorenzo appeared before the tribunal of the Mercanzia, arguing that he had been a partner of Pesellino since 1453 and was entitled to half of the 20 florins that Pesellino had so far been paid (that is, 60 lire).⁴⁰

On 10 October 1457 Monna Tarsia, Pesellino’s widow, appeared before the Mercanzia, disputing Piero di Lorenzo’s claim. The altarpiece was described as almost complete – ‘quasi fornita et tucto quello che insu decta tavola fosse et appartenessesi del maestro, colori et ori et qualunche altra cosa’ (‘almost finished and with everything regarding the panel and appertaining to the master, colours and gold and each other thing’).⁴¹

On 14 June 1458 part of the rest of the payment due on the altarpiece (20 florins) was paid to Monna Tarsia (‘per parte di resto della taula’); this brought the total paid to date to Pesellino and his widow to 40 florins.⁴²

It is clear from the documents concerning Piero di Lorenzo’s claim that he had not participated in the painting of the panel and that his interest was purely a business one: whenever the painting is mentioned, it is said to have been painted by Pesellino.⁴³

Thr commission given to Filippo Lippi

More than a year after Pesellino’s death, between 27 September and 3 October 1458, the panel was brought to Pistoia.⁴⁴

On 31 October the commission to complete the altarpiece was given to Filippo Lippi (‘quando s’allogò la taula a frate Filippo’).⁴⁵ Not only had the latter been involved in the estimation of the state of completion of the altarpiece, but he and Pesellino had collaborated on at least one other altarpiece⁴⁶ and Lippi would therefore have been well capable of adapting his style to make it congruent with what had already been painted,⁴⁷ although ultimately the completion of the altarpiece seems to have been left largely to his workshop.

On 5 November 1458 the panel was transported to Prato,⁴⁸ where Lippi and his workshop were working (1458–65) on the frescoes for the choir chapel of the Duomo.⁴⁹ His named assistants, who collected some of the Trinity altarpiece payments, were Don Diamante and Domenico, Lippi’s discepolo.⁵⁰

On 5 June 1460 the altarpiece was brought from Prato to Pistoia and on 8 June the Bishop Donato de’ Medici judged the painting to be perfect and worth 200 florins (the maximum payment originally agreed by the Company), so Lippi was paid 115 florins, 85 having already been paid to Pesellino and his heirs.⁵¹

The contributions of Pesellino and filippo Lippi

Because of the interruption in the execution of the painting, much modern scholarship has been devoted to trying to establish how much of it had been completed by Pesellino before he died, and how much can be attributed to Filippo Lippi and his workshop. The arbitrators, Filippo Lippi and Domenico Veneziano, had judged the painting to be half‐finished, and it was described at the tribunal as nearly complete. Since 85 florins were paid to Pesellino and his heirs, representing just over half of the minimum sum originally agreed (150 florins), it seems likely that the altarpiece was indeed half‐finished. As noted above, the maximum sum agreed (200 florins) was eventually paid, with Lippi receiving 115 florins, which might have included payment for the predella, attributed to Lippi and his workshop and not mentioned in the initial documents (see below), although it is surely probable that the predella was planned from the start, and there is some evidence that this was the case.⁵²

The problem is what exactly half‐finished (‘fussi fatta la metà’) or nearly finished (‘quasi fornita’) meant. Scholars have been divided in their views, while accepting the overall design as by Pesellino. Bacci, Hans Gronau and Martin Davies saw the two pairs of saints as by different hands.⁵³ Marchini saw the two saints on the right, much of the Trinity and the whole of the Saint Mamas predella panel as being by a single hand, and the women in the Saint Zeno predella panel as by yet a third hand.⁵⁴ Ruda felt that the differences in the appearance of the two pairs of saints could be due to the distinction between the two age groups and their different positions in relation to the light source, and saw the colour choices in the costumes as Lippi’s.⁵⁵ Davies wrote that ‘Indeed, the only at all prominent parts of this main panel which might pass as purely Lippi’s are the hands of God the Father, and the Dove’.⁵⁶ However, this minimalist assessment does not seem entirely accurate. It is possible to make some observations concerning the sequence of execution, and therefore the attribution, based on a minute examination of the painting.⁵⁷ This divides roughly into three: those areas which one might argue are securely attributable to Pesellino, those probably attributable to Lippi and his workshop, and a confusing middle ground where it is impossible to be certain.

Parts of the main tier can, with a degree of confidence, be attributed to Pesellino himself. Evidently the preparatory stages – that is, the underdrawing and those areas which are water‐gilded – were carried out by Pesellino. All the figures attributable to Pesellino are characterised by a very strong grasp of form and structure and follow the technique described above, using a thin wash of paint over a strong drawing and then reinforcing that drawing by the hatched application of paint. Drapery is modelled with a final layer of hatching and cross‐hatching to create highlights and shadows. The gauge for attribution of the actual painting is the figure of Saint Mamas. Bacci⁵⁸ points out that had the figure of Saint Mamas not already been painted when Pesellino died, then the Company of Priests would certainly have had it deleted from the painting, since the person who had particularly requested [page 274] this figure, Pero ser Landi, was in disgrace by October 1457 (before the commission was given to Lippi), for financial mismanagement and for making excessive trips to Florence when two would have sufficed.⁵⁹ The figure of Saint Mamas was, however, not entirely completed: the belt was reserved and the leggings may not have been finished. The lions’ manes and muzzles show partly over and partly under Saint Mamas’ yellow cloak and were therefore also part of the first stage. The figure of Saint James is by the same hand as Saint Mamas. The head and feet of God the Father, which are firmly modelled, seem also to be by the same painter as Saint Mamas, and the ultramarine robe has the same type of hatching in lead white to create highlights as is found on Saint Mamas’ sleeve. The angels seem also to have been completed by Pesellino;⁶⁰ the drapery of the left‐hand angel (fig. 12) is similar in technique to the robe of Saint Mamas, with an underlayer of yellow beneath the green.

It may also be that the light brown landscape, thinly painted, was laid in by Pesellino, as well as the foreground: the strongly drawn foliage of the foreground is similar to that found in paintings attributed to Pesellino, for example the cassone panels with David and Goliath (NG 6579 and 6580; see pp. 288–9), and very different from that found in paintings by Filippo Lippi, such as NG 666 (see p. 142). The thick paint layers have been identified as consisting of artificial malachite over the layer of solid black pigment, and then glazed with a ‘copper resinate’‐type green, which is now rather discoloured.⁶¹

The large trees on the left behind Saint Mamas, which have been reserved, and the trees on the right, thickly painted in a manner very similar to the foreground, also belong to the first stage. An incised circle shows that Saint Zeno was originally intended to have a halo, and this area is now covered by the leaves of the tree behind him (see detail, p. 277). It is possible that it was decided not to give Saint Zeno a halo because of the potential confusion in the overlap with his crozier and Saint Jerome’s halo. The trees on the left seem to have had their final details of dots for foliage added by Lippi, since they have Lippi’s characteristic stippled application and their tips go over the angel above.

Probably by Fra Filippo Lippi and his workshop is the final detailing which disguises the juncture in the middle ground between the two types of landscape. The very fine downward strokes delineating the bushes and shrubs and the reeds and grasses, particularly along the river bank, are of the kind found in the Saint Jerome predella panel, and the dark rough trees and hedges with foliage executed in light spots (also found in the Saint Jerome predella panel) are characteristic of paintings generally accepted as attributable to Lippi, such as NG 667 (see p. 143).

Mordant gilding was among the final stages in the execution of a panel painting, and it seems likely that this, too, was carried out by Lippi and his workshop. One of the characteristics of the mordant‐gilded areas in this painting is that those parts of the decoration which are intended to catch the light (which comes from the left) are executed in mordant gilding, while those intended to lie in shadow are painted; for example, both the border and the spots along the neckline of Saint Mamas’ blue tunic, as well as God the Father’s tiara, the spots on his outer robe, and the left‐hand lining and brown border of his pink robe. This technique is also found in works accepted as by Lippi; for example, in the trees in the background of NG 667. It may be that the original intention was to mordant‐gild Saint Mamas’ belt, and that it was left blank by Pesellino and eventually painted by Lippi with dots, which he scattered liberally as the decorative elements of many of his paintings.

One of the more problematic areas concerns the painting of the figures of Saint Zeno and Saint Jerome (see details on pp. 276 and 277), which are in a somewhat abraded condition. Both appear to be by the same hand, but probably not the one that painted Saint Mamas. In the flesh painting there is no evidence that the underlying design is any different from the rest of the composition. However, it does appear that the paint application is not quite the same as in the other figures and the modelling shows less grasp of form and disegno than is found in the painting of Saints Mamas and James. Examination of samples from the flesh painting of the two pairs of figures, Saints Mamas and James and Saints Zeno and Jerome, reveals two varying methods of constructing the flesh tones: in the first pair the flesh is constructed with light pinkish layers of vermilion and white over pale underlayers containing either a little green earth or azurite, while in the other pair there is only a low proportion of lead white, and the earth pigments (principally red and yellow‐brown) and some black are painted directly onto the gesso ground, with no preparatory layer of green earth.⁶² Caution is needed in making any final assessment, as one might expect different treatment for the representation of older saints. Nonetheless, the difference in technique may suggest a different hand, although not necessarily that of Filippo Lippi, but possibly someone in his workshop to whom he may have allocated the completion.

Another problematic area is the centre of the composition (see detail on p. 279). It seems possible that it was this area which was left only partially painted. There are some unresolved and illogical aspects of the mandorla: It seems likely that this central area was started by Pesellino and completed by Lippi’s workshop, hence the anomalies and inconsistencies.

• (i) the concentric rings above the cross do not follow through below the cross – there are patches of sky instead;
• (ii) the blue cherub at the upper right (see detail on p. 283) is painted with highlights of ultramarine and white (similar to God the Father’s robe), while the other blue cherubim are painted with azurite only and are not as light in colour (evident in the infra‐red photograph, fig. 3);
• (iii) the head and feet of God the Father are probably by the same painter as Saint Mamas, but the painting of his hands is exceptionally weak: they are very pink and feebly modelled, and the sleeves appear to have been painted late in the process. Moreover, in the St Petersburg drawing (fig. 11), which is a fairly detailed study for this part of the painting, two areas were left unfinished: the body of Christ (this is explicable by the fact that this was simply a conventional pose the painter could take from a previous crucifix or Crucifixion) and the hands of God the Father. It seems likely that Pesellino had deliberately left the painting of God the Father’s hands until he could complete an individual study, mindful of the difficulties of showing hands bearing the full weight of the cross;
• (iv) the head of Christ is similar in style to the left‐hand angel; the body of Christ is difficult to assess, since it is painted with only a single layer of paint and may not be completely finished. His beard seems to be by a different hand from that of whoever painted the hair, and may have been painted in the later stages, to integrate the head with the thinly painted body;
• (v) the dove has been painted on top of the cross and the red drapery below and would seem to belong to a later stage;
• (vi) below the arms of the cross the tips of the wings of the cherubim and seraphim go over the landscape and over God the Father; the clouds supporting God the Father’s feet at the base of the cross were among the last things to be painted since they go over the blood dripping down the cross.

It seems likely that this central area was started by Pesellino and completed by Lippi’s workshop, hence the anomalies and inconsistencies.

Lippi and his workshop are generally thought to have carried out the predella, with opinions varying as to the extent of the involvement of Filippo Lippi himself.⁶³ Martin Davies pointed out that the predella is not referred to when the Company decided on the programme for the altarpiece, nor in any of the payments to Pesellino, or indeed to Lippi. It is first surely referred to in 1465, although it is more than likely that it was planned from the beginning.⁶⁴

It has been suggested that the first two scenes of the predella may have been designed by Pesellino and the predella then completed by Don Diamante, who was working with Filippo Lippi on the frescoes in Prato Duomo.⁶⁵ Recently the predella scene with Saint Mamas has been attributed to the so‐called Master of the Johnson Nativity (Domenico di Zanobi di Piero?) working in Lippi’s workshop.⁶⁶ Others have given the predella entirely to Lippi’s workshop.⁶⁷ Examination with infra‐red photography helps to clarify the problem of attribution. The drawing of the main tier is done with meticulous cross‐hatching (figs 3–5), while the underdrawing of the predella is much more free (figs 6 and 7). It seems likely, from the evidence of the underdrawing of the predella scenes, that the composition of the whole predella was designed by Filippo Lippi – which would, of course, be the case if the piece of wood brought from the carpenter’s shop in 1457 was indeed the predella panel. The underdrawing differs considerably from that of the main tier, but compares well with that of NG 667 (see figs 6 and 7 on pp. 146–7).⁶⁸

Two drawings are connected with the scene of Saint Jerome in the predella: a Seated Monk (fig. 13), sometimes attributed to Don Diamante,⁶⁹ and on the verso a rough sketch showing Saint Jerome in Penitence (fig. 14).⁷⁰ The existence of two sketches, one with the saint before a desk (that is, Saint Jerome in his study) and one of the saint in penitence,⁷¹ suggests that these may have been two preliminary ideas for the final predella scene. Eventually, the scene with the lion was chosen, perhaps for the same reasons of balance as may have dictated the iconography of the main tier, since Saint Jerome’s lion is balanced at the other end of the predella by Saint Mamas’ lions. There was almost certainly some input from Filippo Lippi himself in the actual execution of the predella. The quality of the Saint Augustine and Saint Jerome scenes is noticeably higher than that of the others. As Ruda points out, ‘the figure and landscape types are Fra Filippo’s’,⁷² and the wash‐like thin painting of the background behind the Saint Augustine and Saint Jerome is particularly fine.

The Other Church Furnishings ordered by the Compagnia

Between July 1456 and July 1457 the Company had been preparing their church to receive the altarpiece, buying 18 braccia of linen for a curtain and yarn for sewing it, ordering a wooden predella (altar step) in intarsia, and preparing the masonry of the altar and stone altar steps.⁷³

In 1460 Lippi was asked to paint the curtain and also to paint a wooden antependium, for 18 florins.⁷⁴ He was finally paid for the antependium between 11 September 1466 and 8 September 1468.⁷⁵ This antependium, which showed the Madonna of Misericordia (‘una Nostra Donna cho’l mantello grande tenuto da due angeli e sotto detto mantello sono molte figure di diversi frati’), has been identified as that once in Berlin, presumed destroyed (fig. 15).⁷⁶

In 1462 the Company commissioned a stained‐glass oculus showing the Trinity, to protect the altarpiece from the elements: ‘conciossa cosa che l’aria ci guastava la soprascritta tavola.’⁷⁷

In 1465 they added shutters and a frame and canopy to the altarpiece.⁷⁸ The shutters were to close across the whole altarpiece, including the predella and canopy, which was presumably made of wood: a payment was made to Jacopo di Giovanni Cristiani, alias di Meuccio, ‘per resto degli sportelli ch’egli à fatti, de l’altare della chiesa della Trinità e della predella sta sotto la tavola de l’altare e del sopracelo’. The shutters were attached to the altarpiece by (iron?) bands nailed in place [page [279]][page 280](‘sei bendelle e sei chancri [hinges?] per chonficare agli sportelli che serrano la tavola d’altare’). The painter Meo di Bocchi painted the shutters, canopy and frame with azurite bought specially from the Gesuati in Florence.⁷⁹ Also in 1465 Meo di Bocchi painted a chest for candles and four shields bearing the insignia of the Company, namely the Trinity.

In June 1465, nearly ten years after the altarpiece had first been commissioned, the whole ensemble was finally hoisted into place.⁸⁰

Subject and Iconography

The Trinity as the main subject of an altarpiece in Tuscan panel painting is first found at the end of the fourteenth century and into the beginning of the fifteenth.⁸¹ However, Pesellino has adapted the old‐fashioned polyptych scheme to suit a Renaissance‐type pala by converting the Trinity into a contemplative Crucifixion set in a landscape with attendant standing saints instead of the Virgin and John the Evangelist. The disposition of the figures in a panoramic landscape was to prove extremely influential.

The four types of tree found in the landscape – cypress and cedar(?) on the left, olive and palm on the right – may be intended to represent the four woods of which Christ’s cross was supposed to have been made.⁸²

It may be that Pesellino derived the pose of Saint Mamas (fig. 17) from Donatello’s David (fig. 16).⁸³ It has been noted that a painting of a lion in a cage was also in the Palazzo Medici (see p. 395, note 31).⁸⁴ In the Trinity altarpiece are found motifs similar to those in other altarpieces attributed to Pesellino. The organisation of the two pairs of figures is very similar to that in the small panel of the Virgin and Child enthroned with Saints (New York, Metropolitan Museum) attributed to him.⁸⁵

Influence

The altarpiece was extremely influential. Probably while the painting was in Filippo’s workshop in Prato, the two flying angels were copied in the Nativity (fig. 20) attributed to Don Diamante, painted for Santa Margherita, Prato, now in the Louvre, Paris.⁸⁶ The figures of Saint Mamas and Saint Jerome were copied in two drawings (Florence, Uffizi, inv. no. 379E, and Frankfurt, Städelsches Kunstinstitut, inv. no. 5585, respectively; figs 18 and 19) attributable to someone working in the workshop in Prato.⁸⁷ The pose of Saint Mamas was adapted for the ‘Majordomo’ in the Feast of Herod in the frescoes for Prato Duomo: a drawing (Florence, Uffizi, inv. no. 673E v; fig. 21) for the figure, derived from Saint Mamas, has been attributed to Don Diamante.⁸⁸

Apart from its immediate impact on Filippo Lippi’s workshop, the altarpiece was, as Ruda rightly notes, ‘a milestone in Florentine painting’.⁸⁹ It was widely copied: the Trinity group was adapted in an engraving of the Triumph of Eternity⁹⁰ and copied by a local, somewhat mediocre, Pistoiese painter,⁹¹ while the angels, also much copied, frequently reappear, for example in Benozzo Gozzoli’s frescoes of 1459 in the Medici Chapel in the Palazzo Medici, Florence.⁹² The figure style was to influence Verrocchio, and the landscape the Pollaiuoli.

Provenance

Provenance of the whole altarpiece

The altarpiece was commissioned in 1455 by the Company or Confraternity of Priests dedicated to the Trinity, the Compagnia della Santissima Trinità, for their church variously called Santissima Trinità or San Jacopo, in present‐day Piazza Gavinana (formerly Cino), Pistoia, as described above. The altarpiece is mentioned in an inventory of 1492.⁹³ It was [page 281] [page 282] noted in Pistoia by Vasari in 1550, and with slightly more detail in 1568.⁹⁴ The confraternity was suppressed c. 1783 and the church, except for the façade, subsequently destroyed by fire.⁹⁵ In 1821 the altarpiece was described as having been sold to a foreigner,⁹⁶ presumably William Young Ottley (see below).⁹⁷

Provenance of the dismembered parts

The central piece – NG 727 – entered the collection of William Young Ottley, who was living in Italy c. 1791/9; it was later seen in the collection by Waagen.⁹⁸ Warner Ottley sale, 30 June 1847 (lot 19), bought by the Revd Walter Davenport Bromley (1787–1862). Seen in the Davenport Bromley Collection by Waagen.⁹⁹ Purchased at the Davenport Bromley sale, 13 June 1863 (lot 172).

NG 3162 (the angel on the right): This angel and its companion appear to have been owned by the Lombardi in Florence.¹⁰⁰ Bought by the Countess Brownlow from Raffaelle Pinti c. 1867.¹⁰¹ Lent by the Countess Brownlow to the New Gallery, 1893/4 (as Masaccio, no. 109), and by Lord Brownlow to the Burlington Fine Arts Club, 1906 (no. 19).¹⁰² Bequeathed by the Countess Brownlow, 1917.

NG 3230 (the angel on the left): Apparently, like NG 3162, with the Lombardi in Florence. Lent by Lord Somers to the British Institution, 1866 (as Masaccio, no. 37), and to the RA , 1873 (no. 182); lent to the RA , 1896 (no. 159), by Lady Henry Somerset (daughter of the 3rd Earl Somers). Purchased from Lord Somers, Temple‐West Fund, 1917.

NG 4428 (Saints Zeno and Jerome): Waagen saw, in the Ottley Collection in 1837, not only NG 727 but also ‘auf zwei anderen Tafeln die Heiligen Jacob und Zeno’.¹⁰³ He probably meant NG 4428 and L15 (see below). NG 4428 is first definitely recorded in 1906 in the collection of the German Emperor William (II).¹⁰⁴ After the First World War it was shown for a time at the Schloss Museum in Berlin. Presented by the NACF in association with and by the generosity of Sir Joseph Duveen in 1929, the year Duveen became a Trustee of the National Gallery.¹⁰⁵

L15 (Saints Mamas and James): Like NG 4428, this fragment is clearly one of the pieces seen by Waagen in 1837 in the William Young Ottley (1771–1836) Collection. Bought in 1846 from his brother, Warner Ottley,¹⁰⁶ through Ludwig (Lewis) Grüner, and presented by Queen Victoria to the Prince Consort on 26 August 1846. It hung for many years at Osborne House (see p. xxiv, fig. 1),¹⁰⁷ but was transferred in 1902 to Buckingham Palace.¹⁰⁸ On loan from the Royal Collection since 1919.

NG 4868 (four predella scenes): Found in some buildings attached to San Desiderio in Via Laudesi, Pistoia, which had belonged to a nunnery suppressed in 1786; after they had been in the possession of the Amati family, the Gelli family acquired these buildings for use as a factory.¹⁰⁹ Writing on the back of the scene with Saint Zeno suggests that this panel was in the hands of Antonio Gherardini between 7 April and 1 May 1897, perhaps for repair. The four panels were lent by Cav. Antonio Gelli to an exhibition in Pistoia in 1899 (Sala XVI, no. 2). After his death they passed to his daughter, Signora Michelozzi‐Roti, in Florence. Later in the Felix Warburg Collection, New York; presented by Mr and Mrs Felix M. Warburg, through the NACF , 1937.

Exhibited

London 1945/6, NG , National Art Collections Fund Exhibition (31). For details of loans of the various panels, see under Provenance for NG 3162, NG 3230 and NG 4868.

Select Bibliography

• P. Bacci, ‘Il Pesellino, Fra Filippo Lippi, Domenico Veneziano, Piero di Lorenzo, Fra Diamante, Domenico discepolo di Fra Filippo Lippi etc. e la tavola pistoiese della “Trinità” nella Galleria Nazionale di Londra’, in Documenti e Commenti per la Storia dell’Arte, Florence 1944, pp. 113–51.
• G. Marchini, Filippo Lippi, Milan 1975, pp. 168, and cat. 51, pp. 212–13.
• J. Ruda, Fra Filippo Lippi. Life and Work with a Complete Catalogue, London 1993, cat. 53, pp. 449–52.
• D. Gordon, ‘The ‘missing’ predella panel from Pesellino’s Trinity altar‐piece, BM , 138, no. 1115, 1996, pp. 87–8.

Notes