Lippo d’Andrea, Two Scenes from the Legend of Saint Michael

James Jackson Jarves (1818–1888), Florence, by 1859

The panel support, of a horizontal wood grain, has been thinned to 1.0 centimeters, cradled, and impregnated with wax. A barb along all four edges of the picture surface indicates that it has not been reduced significantly in any dimension. Pre-1960 photographs (fig. 1) show the panel in the heavily overpainted state in which it has been known to most scholars. Cleaning in 1960 reduced the paint film to a network of lacunae, particularly extensive through areas of dark color and earth tones. Very little of the narrative is more than imperfectly legible. It remains unclear how much overpaint was left on the panel, despite the radical damage from solvents and scraping: there is some evidence that the child in the left scene may once have had a halo, and many of the rocky fissures into which the demons are cast in the right scene appear to be built up with layers of later paint.

Initially thought to be by Spinello Aretino¹—or to have been executed by the school of Spinello² or in the manner of Spinello³—the Jarves Legend of Saint Michael was more accurately described as by a follower of Agnolo Gaddi by Hans Gronau and George Kaftal.⁴ In 1927 Richard Offner annotated a photograph of the painting at the Frick Art Reference Library as “Florentine, ca. 1460,” while Federico Zeri annotated his own photograph in 1967 with the correct attribution to the Pseudo-Ambrogio di Baldese. This attribution was repeated by Boskovits⁵ and endorsed in written communications by Everett Fahy (1978), Luciano Bellosi (1987), and Carl Strehlke (1998). The evident justification for this attribution was demonstrated by Katherine Smith Abbott in comparing the scene at the right of the Jarves panel, showing Saint Michael and his host defeating the Rebel Angels, to that of the same subject painted by the Pseudo-Ambrogio di Baldese in a predella panel in the Museo Diocesano d’Arte Sacra at San Miniato al Tedesco (Pisa) (fig. 2).⁶ Smith Abbott also argued for accepting the identification of the Pseudo-Ambrogio di Baldese with Lippo d’Andrea as first proposed by Serena Padovani rather than with Ventura di Moro as suggested earlier by Enzo Carli.⁷ Both identifications still appear in the scattered literature concerning the artist, but the identification with Lippo d’Andrea seems far more likely to be correct for the majority of paintings included in this large and somewhat heterogeneous group.

While the scene on the right of the Jarves panel can unequivocally be recognized as the Fall of the Rebel Angels, the scene on the left has so far eluded precise identification. It shows a bearded saint standing before the door of a chapel or hermitage, addressing a child standing before him. Approaching from the left is a cohort of mounted knights holding spears and an imperial banner. These details could relate to the story of the army sent by Nero to arrest Saints Nazarius and Celsus in the wilderness, but the apparition of a host of angelic warriors led by Saint Michael in the background above the chapel does not occur in the narrative of Saint Nazarius as related in the Golden Legend. James Jackson Jarves identified the scene as a “Vision of Constantine,” although Osvald Sirén confessed to be unable to see any reason for such an identification.⁸ Kaftal described the scene as illustrating the appearance of Saint Michael to the bishop of Siponto, promising him victory on the eve of battle, but qualified this in a footnote as “tentative identification: very doubtful.”⁹ Possibly it represents a local legend of Saint Michael not included in the Golden Legend or in other written sources.

Only one attempt has so far been made to reconstruct the original context of the Jarves Legend of Saint Michael. In 1950 Gronau proposed reuniting three dispersed predella panels all showing different episodes from the legend of Saint Michael with a fragmentary altarpiece by Agnolo Gaddi in which the Archangel appears alongside Saints Julian and James in one of the lateral panels. Three of the five panels included by Gronau in this reconstruction are, by coincidence, in the collection of the Yale University Art Gallery: the just-mentioned lateral panel showing Saints Julian, James, and Michael (Agnolo Gaddi or Lorenzo Monaco, Saints Julian, James, and Michael), a predella panel showing the Apparition of Saint Michael at Mont-Saint-Michel and the Miracle of the Bull at Monte Gargano (Fra Angelico, Predella: Two Scenes from the Legend of Saint Michael), and the present panel. This reconstruction has been correctly rejected by most authors: the two predella panels at Yale are by different artists and from different dates, while the third predella panel—showing the Apparition of Saint Michael above the Castel Sant’Angelo and now in the Pinacoteca Vaticana¹⁰—is by yet another artist. None of these three is by the artist responsible for the Saints Julian, James, and Michael lateral panel at Yale, which was painted either by Agnolo Gaddi or by Lorenzo Monaco in, or recently emerged from, Gaddi’s workshop, and all of them date twenty or more years later than it does. The link tenuously uniting the works in Gronau’s reconstruction was the fact that the second Yale predella panel, showing the Apparition of Saint Michael at Mont-Saint-Michel and the Miracle of the Bull, was discovered in the mid-nineteenth century framed together with the final panel in his proposed altarpiece, a Virgin and Child by Agnolo Gaddi, now in the Contini Bonacossi Collection at the Gallerie degli Uffizi, Florence.¹¹ This assemblage, however, was a pastiche, and all five works in this group are, in fact, entirely unrelated.

It may be possible to suggest one and perhaps two predella panels by the Pseudo-Ambrogio di Baldese/Lippo d’Andrea that could plausibly have stood alongside the Jarves panel in a single altarpiece. In 1932 Bernard Berenson published an Adoration of the Magi (fig. 3), then “homeless,” that corresponds closely to the Yale panel in style and almost exactly in size, reportedly measuring 30.1 by 71.7 centimeters.¹² That such a panel might have stood in the center of the predella of which the Yale panel formed the left or right member is suggested by analogy with the predella to Lippo d’Andrea’s altarpiece from Santa Maria Nuova, now in the Galleria dell’Accademia, Florence.¹³ In that altarpiece, the center panel of the predella is approximately the same width as either of the side panels, and each of the latter is divided into two scenes drawn from the legends of the saints portrayed in the main register above them. In the case of the Yale panel, it is difficult to know whether the two scenes refer to different saints or whether both are intended to commemorate miracles of Saint Michael. If they relate the stories of different saints, it is reasonable to suppose that the other lateral panel from the predella mirrored it in format and also contained narratives from two different saintly legends. If they both celebrate Saint Michael, then the other lateral predella panel is likely to have shown either additional scenes from the legend of Saint Michael, as Gronau supposed, or two scenes (possibly one long scene) from the legend of another saint. Such a panel could have resembled the Martyrdom of Saint Acacius and the Theban Legion in the Musée des Beaux-Arts, Dijon (fig. 4). Close in style to the Adoration of the Magi and the Yale Scenes from the Legend of Saint Michael, this panel also corresponds to them in height (30 centimeters); it is only 43.1 centimeters long but has clearly been cropped at both sides. No altarpieces or fragments of altarpieces by Lippo d’Andrea are known that portray either Saint Michael or Saint Acacius.

Dating the Yale panel and its possibly related companion panels is, given their compromised state and the relative paucity of comparative material, largely intuitive. The compression of the narrative of the Fall of the Rebel Angels into a nearly square format relative to the more expansive composition in the San Miniato predella, as well as the looser handling of both scenes in the Yale panel, suggests that it follows rather than precedes the San Miniato example. The latter has been dated shortly after 1413 on the basis of a donation of land to the Dominicans in San Miniato to endow a chapel of Saint Michael in the church of Santi Jacopo e Lucia, the first mention of such a dedication in the historical record.¹⁴ The Yale panel is even closer in style to the scenes in the predella of the Accademia altarpiece, which is dated 1430 by inscription, although whether it might have preceded or followed that work is unclear. The dating proposed here, ca. 1430, must therefore be understood as both approximate and tentative, pending verification of other fragments of the same structure and a better understanding of the development of Lippo d’Andrea’s style over the second half of his career. —LK

1. Brown 1871, 16, no. 34. ↩︎

2. Jarves 1860, 45, no. 32. ↩︎

3. Sturgis 1868, 40, no. 34; Sirén 1916a, 63–64, no. 23; and Seymour 1970, 36–37, no. 21. ↩︎

4. Gronau 1950a, 41–47, pl. 25/fig. 4, pl. 27/fig. 7; and Kaftal 1952, cols. 740–41, fig. 834. Kaftal incorrectly listed the painting as in the collection of Maitland Griggs, New York. ↩︎

5. Boskovits 1975b, 298. ↩︎

6. Katherine Smith Abbot, in Smith Abbott et al. 2009, 38–39, 82–83, fig. 10. Abbott thought this predella panel was one of a series from an altarpiece with another showing the Martyrdom of Saint Catherine of Alexandria, also in the Museo Diocesano in San Miniato al Tedesco (Pisa). Both of these predella panels stood beneath large votive panels intended to be hung on pillars, not beneath lateral panels of an altarpiece. Both votives survive: the Saint Michael the Archangel is by Lippo d’Andrea and the Saint Catherine of Alexandria, together with its predella, is by Rossello di Jacopo Franchi. ↩︎

7. Serena Padovani, in Tesori d’arte antica 1979, 55–56; and Carli 1971, 109–12. ↩︎

8. Jarves 1860, 45, no. 32; and Sirén 1916a, 63–64, no. 23. ↩︎

9. Kaftal 1952, cols. 740–41. ↩︎

10. Inv. no. 528. ↩︎

11. Inv. no. Contini Bonacossi 29. ↩︎

12. Berenson 1932c, 177. The painting was last recorded at sale at Sotheby’s, New York, January 26, 2006, lot 269. ↩︎

13. Inv. no. Dep. n. 18. See Pisani 2001, 1–36; and Daniela Parenti, in Hollberg, Tartuferi, and Parenti 2020, 89–95. ↩︎

14. See Ciardi et al. 2013, 87. ↩︎