P.2A,B (COPIES) Madonna and Child with St. John the Evangelist

P.2A Madonna and Child with St. John the Evangelist

c. 1513

A. Tours, Musée des Beaux-Arts, no. 947-3-1.

Panel, 95 x 62.

Fig.P.2Ab bw, restored
Fig.P.2.Ac bw, top straight
Fig.P.2.Ad bw, before restoration
Fig.P.2.Ae bw, Child and Saint, before restoration

A horizontal crack running the full breadth of the picture just above the Virgin’s head indicates the separateness of the upper arched area, 12 cm. high. Although both parts of the panel seem to be poplar, the wood of the upper section is lighter in color. However, the painted surface appears to be consistent. It is possible that the panel was first constructed for a rectangular picture and then altered before the actual execution of the painting itself was begun for otherwise one would expect the panel to be continuous and not joined. M. Jean-Marie Girard, of the museum, believed in 1971 that the architecture is a later addition, but that this need not be wholly the case, see below.  The bottom of the panel is cut, removing the toes of Christ’s left foot and the saint’s left hand of which only one finger remains.  Christ’s whole left foot appears in the other copy.

The tonality of the painting is dark. The Virgin’s drapery is reddish purple; the drapery over her head and the cover of her book are dark green. The saint’s robe is dark olive-green; his mantle is yellow-orange.1

PROVENANCE: Acquired before 1947 from the Château du Plessis.


Valentino Pace, “Carlo Portelli,” Bollettino d’Arte, series V, LVIII, 1973, I, 27, 29, ns. 9–12, 30, cat. no. 1, attributed the picture to Portelli as his earliest known work, datable before 1555. He sees it as a copy of the other version (see below), by Rosso or Machuca, with the saint and the niche added by Portelli, but possibly not the top of the niche that he believes could be the result of a later restoration.

Carroll, 1987, 14, 33, n. 16, as a copy of Rosso’s lost early painting mentioned by Vasari.

Costamagna, 1989 (1984–1985), 18–19, 23, n. 14, P1. 9, as Portelli, with some reservation, after Rosso, the head of the saint Portelli’s only contribution, and done c. 1543.

Ciardi and Mugnaini, 1991, 13, 17, 48–49, no.4, and Fig., as Rosso around 1514.

Franklin, 1994, 14, 15, 16, P1. 13 (color), 17–18, 68, 83, 118–119, 212, as a copy of Rosso’s lost picture, the round top not original; he states that a convincing identification of the copyist may never be made and that it impossible to know if the architecture, not found in the other copy (P.2B), is Rosso’s although it does effectively define the pictorial area; he also relates the head of the saint to a terracotta bust of Christ (17, Pl. 15) in Rome attributed to Baccio da Montelupo.

Ciardi, 1994, 61, Fig., 63, 66–67, 90 n. 28, 96, ns. 137 and 139, as a copy of the other version of this composition (P.2B), where the figure of the saint has been painted out, the head of St. John related to heads by Puligo; he also thinks that Rosso’s composition may have been painted after his Assumption.

The picture was formerly attributed to the School of Andrea del Sarto although, as reported by M. Girard, Pontormo’s authorship has been suggested. M. Girard recognized the relationship of this painting to the other version of it and in response to his letter of 1969 Longhi, who thought the latter was by Machuca, acknowledged this relationship and added that he believed the Tours picture was a Florentine work copied from the other with the copyist altering the character of the head of the Virgin and adding the figure of the saint. Pace’s attribution to Portelli is based primarily upon the appearance of the saint whom he does not attempt to identify. But the style of other parts of the painting also seem suggestive of Portelli’s authorship. Both Longhi and Pace were wrong in seeing the figure of the saint as an addition to the picture made up partly of the pile of drapery in the other version (P.2B).

P.2B Madonna and Child, Milan?

B. Formerly Rome, Marandotti, dealer, who in 1959 reported that the picture was in a private collection in Milan; it may subsequently have been on the art market in New York.

Canvas, 86 x 63.5.

Fig.P.2B bw, Milan

Briganti (see below) states that the color of the picture is dark and intense with strong contrasts, the cushion a marine green, the drapery covering it a bright coral red. These colors are related to those of the saint’s garment in the Tours picture. Ciardi, 1994, 96, n. 137, states that in photographs once owned by Briganti it could clearly be seen from the thickness of the paint that the saint, seen in (P.2A), was painted over. In a letter of 17 July 1992 he also told me that in the photographs a raking light shows that the picture has been cut and what was left of the saint’s head and the niche has been painted over.


Briganti, 1953, 53, recognized the painting as an autograph work by Rosso and dated it around 1520 as contemporary with the Madonna in Glory in St. Petersburg, which he accepted as Rosso’s (wrongly in my opinion, see RP.23), following the Assumption, that he dated in 1517 (wrongly, see P.3), and immediately preceding the Volterra Deposition of 1521.

Barocchi, 1958, 236, accepted the attribution to Rosso and commented that the picture is not finished; she also saw it in relation to the St. Petersburg picture and connected it as well to Rosso’s painting in Los Angeles which she dated in Rosso’s Florentine years (wrongly I believe, see P.24).

Griseri, 1964, 17, reported Longhi’s attribution to the Spaniard Machuca.

Borea, 1965, maintained the attribution to Rosso and dated the painting around 1518 or slightly thereafter.

Longhi, 1969, 36, published his attribution to Machuca, dating the picture in 1518–1520 during a period he supposed this artist was in Florence.

Costamagna, 1984, 94–95, as by Machuca.

Angelini, 1986 (1987), 91, no.5, as an early work by Rosso.

Carroll, 1987, 14, 33, no. 16, as a copy of Rosso’s early painting mentioned by Vasari.

Costamagna 1989 (1984–1985), 19, 23, n. 15, as probably after Rosso.

Costamagna and Fabre, 1991, 25, n. 4.

Franklin, 1994, 14, 15, 17, P1.14, 18, as a copy of Rosso’s lost painting, but a convincing identification of the copyist may never be made.

Ciardi, 1994, 20, 61, Fig., 66–67, 94. n. 28, 96, ns. 137 and 139, as not by Machuca, and difficult to exclude or affirm as an original by Rosso made for Fra Jacopo perhaps after the Assumption in thanks to him for the commission and to obtain another.

The distinctive personal style of the painting that differs so markedly from that of the other and more inept but probably more faithful copy does not suggest Rosso but does bear a resemblance to Machuca’s as has several times been pointed out. This could date the copy between 1517 and 1524 only a few years after the time that Rosso painted the original. Its color suggests that it was painted with the lost original at hand. It is not known if the painting was executed on canvas or transferred to this support from a panel. Rosso’s painting was almost certainly on panel, as of all his easel pictures only the Moses (P.14) is painted on canvas and this probably for very particular circumstances that would not be related to those of the earlier picture.

Vasari in 1550, 796 (Vasari-Ricci, IV, 243) and again in 1568, II, 105 (Vasari-Milanesi, V, 156) wrote that Rosso painted “una Nosta Donna con la testa di San Giovanni Evangelista, mezza figura” for “maestro Giacopo frate de’Servi”. This picture is mentioned immediately after the lost coat-of-arms of Cardinal Lorenzo Pucci at SS. Annunziata, for which Rosso received payments in October and November 1513 (L.8), and just before the fresco of the Assumption in the atrium of the same church that was paid for between 20 November 1513 and 18 June 1514 (see P.3). The same friar at the Annunziata was, again according to Vasari but also as supported by the documents, responsible for Rosso’s receiving the commission for this fresco apparently following upon his having done the Madonna and Child with St. John the Evangelist.

The subject of the lost picture mentioned by Vasari is so unusual as to have caused three scholars to assume that it is a mistaken reference by Vasari for a picture of the Madonna and Child with St. John the Baptist,2 a subject that is certainly more probable, especially in Florence. The inclusion of St. John the Evangelist in such a scene appears to be unprecedented in that city, as well as elsewhere. Nevertheless, the painting in Tours seems to show the rare subject that Vasari describes and in a style that indicates a painting that could have been designed in Florence around 1513. Hence that picture deserves to be considered to be Rosso’s, although as a copy of his picture rather than as an autograph work. The same image, but without the saint, is known from the painting formerly in Rome that has actually been attributed to Rosso. As Ciardi reported it too once showed the saint and the niche.

Before looking further at the styles of these pictures it is necessary to see how seriously the identity of the saint can be established as St. John the Evangelist. His halo, assuming it is original, indicates that he is, in fact, a saint, and his traditional costume – a simple undergarment with a mantle thrown over it – would seem to describe an old-established saint and not one of recent date who might be depicted in something approaching contemporary costume. He is young with only a slight indication of moustache and beard, and he carries an open book from which it might be supposed he has been reading. While it is not possible to insist that these attributes designate the young Evangelist and no other saint it is also true that they allow for the possibility of recognizing him as this St. John. One should also point out that his appearance here half-length, as mentioned by Vasari, in front of the Virgin and Child is, whatever saint he might be, highly unusual. Supposing that the Tours picture is a copy of a painting of 1513 the placement of a saint in this position may actually be unique for this particular moment. It, too, would seem to have no precedent. Still, the fact that he holds an open book and that the Virgin also holds one would seem to suggest some kind of reciprocal significance. In other words, the book held by the saint may not be only his symbol but also an element in the drama of the painting to indicate a close relationship to the Virgin and her Child. Hence, it is very probable that the saint here is not only original to the image but is also the youngest of the evangelists. His unusual appearance with the Virgin and Child, and half-length as mentioned by Vasari, in a picture the style of which is very much what one would expect in Florence around 1513 makes it very likely that the Tours picture is a copy of Rosso’s lost early painting.

The unusualness of the saint’s appearance in the Tours painting may be the reason for his having been painted out in the other version. In that painting part of this figure, his drapery, and the book have been converted into a tasseled pillow with two passages of drapery laid over it that very closely correspond to the saint’s drapery. This transformation has produced a quite unnecessary pile of stuffs having no function in relation to the Virgin and Child.

The version formerly in Rome is more expert in its execution and is therefore more interesting in certain respects, but it may also exhibit more effect of the stronger personality of its own author, as, for instance, in the elongation of the Virgin’s head, in the greater slickness of the surfaces of all forms, and in the spreading out and swelling of the Child’s features. As already noted, this version also does not have the niche that appears in the other picture and this may be another omission. But the height of this picture, extending to just above the Virgin’s head, and the straight edge of the top of the picture seem more appropriate to the image than the flat, rather than fully rounded, arch at the top of the Tours picture, which is, in any case, an addition to this panel. The elimination of the top of the arch would not mean that the lower part of the niche was not described at all in the lost painting for a niche without its arched top appears in Rosso’s Dei Altarpiece. And as Ciardi indicated signs of the niche can be seen under the repainted surface of the picture formerly in Rome. Although the niche and the ledge before it in the Tours painting enclose a space that is somewhat incomprehensible in regard to the placement of the Virgin in it this awkwardness does not necessarily indicate that such an arrangement was not part of a picture by the very young and inexperienced Rosso.3 The unusual inclusion of the half length saint in the picture could account for its unusual architecture for without it it might be difficult to show such a half-length figure related as it is to the Virgin and Child. Certainly the ledge is required for the placement and pose of the Child. It should be noticed that the entire left foot of the Child is visible in the Rome picture indicating that while the rather awkward painting at Tours may preserve more of the lost original, the other version may be more accurate in a few details. The height of the Tours picture without its top section is 83 cm., almost identical to the height of the other painting, 86 cm. But if one adds a few centimeters to the Tours painting for the missing part of the Child’s foot the difference would be made up. This again suggests that the arch at the top of the Tours painting did not belong to the lost original image.

Stylistically – and reading somewhat more from the less facile but more complete version at Tours – the Child is very similar to the angels in Rosso’s Assumption of 1513–1514, his head most specifically to those at the lower left of the Virgin in the fresco, his twisting pose to the angels tumbling about in the clouds (Fig.P.3e). His upraised arm is a gesture that Rosso employed in his later works, in the Standing Nude Woman, in the Uffizi (D.5), in the man leading a camel in the copy of the Rebecca and Eliezer in Pisa (Fig.P.15a), and in his Bacchus in a Niche in Besançon (D.18), also known from Caraglio’s print (E.42). The slight moustache and beard and the flat-topped ear of the saint also characterize the figure of St. John the Baptist in Rosso’s S. Maria Nuova altarpiece of 1518 (Fig.P.5a). A comparison can also be made between the open mouth of the saint, baring his teeth, with that of St. James in Rosso’s Assumption (Fig.P.3c). The rippled edges of the saint’s drapery can be found in the drapery falling over the shoulder of the fourth apostle from the right in that fresco (Fig.P.3b) and in the drapery falling over the knee of the Virgin in the altarpiece of 1518 (Fig.P.5a). In the Tours painting the blunt features of Virgin are related to those of the same figure in the Assumption (Fig.P.3e). These similarities to autograph works by Rosso are sufficiently compelling to indicate a high degree of probability that the painting in Tours and the one formerly in Rome represent the lost early work by Rosso that Vasari describes.

As to date of the lost painting in 1513, as suggested by Vasari’s placement of it in his Life of Rosso, one might point out, as also noted by others, that the complex pose and physical exertion of the Child may well reflect the study of Michelangelo’s Battle of Cascina from which Rosso would seem to have drawn about that time (L.3). The Child could also be partially inspired by the children and angels in Sarto’s Marriage of St. Catherine in Dresden,4 of around 1512–13, and by the children in the Scalzo Charity, datable in 1513.5 But Rosso’s picture is not altogether, or even essentially, Sartesque. The physiognomy and sobriety of the Virgin in the Tours picture, her upright pose seen in contrast to the very active Child, and the use of a niche can be associated with such a work as Bugiardini’s Madonna del Latte in the Museo del Cenacolo di San Salvi, Florence.6 Bugiardini’s Madonna and Child in Munich exhibits some of the same characteristics.7 It is very possible that following a moment in 1511 or 1512 when Rosso worked in Andrea’s shop that he moved to Bugiardini’s for a while. Here he would have been introduced to that artist’s Bartolommesque manner. The Madonna and Child with St. John the Evangelist indicates such an experience combined with Sartesque elements that Rosso had already acquired. A similar combination, more fully integrated, appears in Rosso’s immediately subsequent and grander Assumption at SS. Annunziata. That Rosso’s lost original was painted after the fresco of 1513–1514 as thought possible by Ciardi seems unlikely to me. It might be relevant to note here that in 1518 Bugiardini was called upon to make an evaluation of Rosso’s S. Maria Nuova altarpiece (P.5).

One detail of the version of Rosso’s painting formerly in Rome that does not appear in the other painting is the strands of hair that fall around the Virgin’s neck and onto her chest. They are not characteristic of Florentine representations of the Virgin whose hair is always modestly pulled back behind her head with only rarely a few wisps visible at the sides of her face. But comparable loose strands do appear in other figures by Rosso, in some of the apostles of the Annunziata Assumption, in the woman at the far right of the Moses (Fig.P.14a), and in the drawing of a Standing Nude Woman in the Uffizi (D.5). It is, therefore, possible that the author of the Rome picture, who may not have been either a Florentine or an Italian, adapted these strands from other works by Rosso and used them here in a way that the Florentine Rosso would never have and never did in depicting the Virgin.8

1I should like to thank M. Girard for kindly sending me information on this painting in 1971 to supplement the notes I took on it a decade earlier. M. Girard’s description of the color is: La Vierge: robe violet éteint, à base d’ocre rouge, très décoloré dans les clairs, bordure du col un peu plus ocre rouge + sienne naturelle. Voile et dos du livre vert foncé. Visage moins soutenu que ceux de l’Enfant Jésus et du Saint, modelé Bans les ombres avec de la Terre d’Ombre. Lèvres violacées. Cheveux Terre de Sienne et Terre d’Ombre. Modelé des mains dans ces ocres jaune et rouge plus soutenues.

Enfant Jésus: Modelé à base d’ocre Jaune, de Sienne naturelle et de Terre d’Ombre. Cheveux Terre de Sienne naturelle et Jaune de Mars. Un peu plus ocre rouge pour le modelé du visage et de la plante des pieds.

Le Saint: Robe vert olive foncé. Draperie jaune moyen et Jaune de Mars, tirant à l’orangé. Modelé du visage de même que celui de l’Enfant Jésus, plus soutenu dans les ocres. Blanc des yeux verdâtre. Cheveux Terre de Sienne naturelle avec un peu de Terre de Sienne brûlée et de Terre d’Ombre.

Rebord horizontal: Terre de Sienne naturelle.

Fond et niche: vent olive et Terre d’Ombre.

Franklin, 1994, 15, relates that St. John has red hair and wears a grey tunic covered by bright yellow-orange drapery.

2See Kusenberg, 1931, 182, n. 7, no. 2, and 1935, 62; also Shearman, 1965, I, 166, who suggested that a painting in Frankfurt showing the infant St. John the Baptist may be Rosso’s picture which he dates about 1515 (see RP.10). Berti, 1993, 57–59, 60, n. 25, and Color P1. III, thought that a tondo in Liverpool also showing the young Baptist is Rosso’s lost painting (see RP.13).

3 See, for example, Franciabigio’s pictures of the Madonna and Child with a ledge or parapet (McKillop, 1974, Figs. 18, 105, 106). However, see also his Marriage of St. Catherine of around 1519 in the Borghese (McKillop, 1974, 163–4, no. 30, Fig. 80) which is comparable to the Tours painting in showing the saint half-length. Here there is no landscape but instead of a plain background as in the Rome version of the picture under consideration there is a pair of divided curtains that serve a spatial function similar to that of the niche in the Tours painting. In fact, a wholly plain background in Florentine panel pictures of the early sixteenth century showing the Madonna and Child is very rare if it exists at all.

4Freedberg, 1963, II, 34–35; Shearman, 1965, II, 210–211.

5 Freedberg, 1963, II, 42–43; Shearman, 1965, II, 298–299.

6 Dated around 1512–1515 by Freedberg, 1961, I, 208, and 1971, 61, and in the middle of the second decade or c. 1515–1518 by Pagnotta, 1987, 48–49, 204–205, no. 30, Color Pl., II, Fig. 30. If influential on the painting for Fra Jacopo it would have to have been done c. 1512 or 1513, although other pictures by Burgiardini, now lost, could have been seen by Rosso.

7 On the attribution of this picture to Bugiardini, see Federico Zeri, “La riapertura della Alte Pinakothek di Monaco,” Paragone, 95, 1857, 68; Fiorella Sricchia Santoro, “Per il Franciabigio,” Paragone, 163, 4; and McKillop, 1974, 205. The painting must date from about the same time as Bugiardini’s Madonna del Latte in Florence, as indicated by Pagnotta, 1987, 49–50, 245, no. 31, Fig. 31, as middle of the second decade or c. 1515–1517 (see preceding note).

8 In an inventory of Fra Jacopo’s possessions left to the Servites made on 10 February 1538, the year after his death (ASF, Corp. RS, 119, vol. 52, fols. 135v–136v, as published in Franklin, 1994, 272, n. 59), there appears on fol. 135v: “Una Nostra Dona con San Giovanni di mano di Raffaello.” This is immediately followed by: “un Piattà di mano di predetto.” If, as Franklin believes, the first picture is the original of the copies catalogued here, then the attribution to Raphael is obviously wrong. Consequently wrong, too, might be the attribution of the Pietà to Raphael, Franklin surmised. He thought that picture may also have been by Rosso. Although it is unlikely that either picture was by Raphael, it is also not certain that the first one showed the Evangelist rather than the Baptist. There is, of course, some probability that the first picture was Rosso’s, based on the supporting evidence supplied by Vasari, but it does not follow that the Pietà, which is not mentioned in any other source, was also his, even if it was sufficiently important to receive, like the other work, an attribution to Raphael.