Beccafumi’s authorship of the three predella panels at Mells – this Presentation of the Virgin, The Visitation and The Adoration of the Kings (MM26, MM27 and MM28) – has never been questioned and they are universally agreed to have been painted late in his career, in the 1540s. But questions about their original state and purpose remain open. Firstly, is the series complete? The principle of selection of these three episodes is not obvious and other episodes in the Virgin’s life could have been chosen, such as the Birth of the Virgin, the Marriage of the Virgin, the Annunciation, the Adoration of the Shepherds, the Presentation of the Infant Christ, among others. Secondly, how were these three panels arranged? Did they abut one another directly? Or were they separated either by intervening episodes or by narrow panels containing decorative forms or standing figures of saints, as in Sanminiatelli’s reconstruction of the predella of Beccafumi’s pala of San Bernardino in his eponymous Oratorio in Siena (the three predella panels in the Louvre; the four upright figures or groups in the Pinacoteca Nazionale, Siena), for example? If they were continuous, their combined length – and hence the approximate width of the altarpiece that stood above them – is about 160 centimetres. If some episodes are missing, or if the three existing panels were framed separately, then the width of the altarpiece could have been greater. 

The original  arrangement of the three panels cannot be established until and unless they are examined technically and their physical make-up determined. Glazed, embedded in their frames, their edges are concealed, so it is impossible to ascertain whether they were continuous on the same board, or whether they were separate panels individually framed; and, if so, by how many centimetres they were separated. One indication that favours, although does not entail, individual framing is that each panel has a plain painted internal frame. Another technical feature that needs to be determined is the medium with which these panels are painted. They seem to be executed in oil but the two predellas with which they have been associated are both stated by Vasari to have been in tempera.

No other panels whose size and subjects might have complemented these have been identified and no drawings by Beccafumi are known that might prepare complementary episodes – nor, indeed, are any drawings known for these panels, which might well have been improvised, so free is their execution. Until and unless others are discovered, it is probably best to assume that these three panels comprise a complete set. 

These issues are important in attempting to determine to which altarpiece these panels belonged. Of course, it is possible that this altarpiece is lost or unrecorded but, assuming it survives, can it be identified? There has been recent consensus that the altarpiece would be the Coronation of the Virgin (300 x 174 cm) painted in the early 1540s for the high altar of the Church of the Camaldolese nuns in Siena, and now in Santo Spirito there. However, it had earlier been suggested, by Sanminiatelli and Baccheschi, following Vasari, that they were placed below the Nativity of the Virgin (225 x 145 cm), painted by Beccafumi for the convent of the nuns of San Paolo and now in Siena’s Pinacoteca Nazionale, whose predella contained three scenes which Vasari called the Adoration of the Magi, the Marriage of the Virgin and the Presentation in the Temple.¹ If Vasari was describing the Mells panels, then he mistook the Marriage of the Virgin for the Visitation or confused the two subjects in his memory, a slip, of course, but not a disabling one, especially considering that there is no other surviving group of predella panels by Beccafumi that better fits Vasari’s description. The alternative view, that the Mells panels formed the predella of the Coronation also presents problems in relation to Vasari’s description: ‘Nella predella, similmente, sono alcune figurine fatte a tempera, molto belle’.²

In terms of subject, the choice of episodes, concentrating on the Virgin’s early life, while appropriate to a Nativity of the Virgin, would be less so to a Coronation of the Virgin where one might expect scenes from her later life such as the Procession to Calvary, the Transport of Christ’s Body to the Tomb, Pentecost or the Dormition. However, scenes from the Virgin’s early life form the predella of Raphael’s Oddi Coronation so this combination is not isolated. 

In short, either the Nativity of the Virgin or the Coronation of the Virgin would – to a greater or lesser extent – complement the subjects of the Mells panels. In favour of the Nativity of the Virgin is that the width of the main panel is only some 7 centimetres less than the combined widths of the Mells panels; but the width of the Coronation (174 cm) exceeds by only 14 centimetres that of the Mells panels in their present state, a shortfall that might easily be accommodated by suggesting that narrow decorative elements were placed between them. 

However, while it would be convenient to accept one or other possibility, there is a strong argument against connecting the Mells panels with either the Nativity or the Coronation. Both these altarpieces are illuminated from the left whereas the Mells panels, and their internal frames, are illuminated from the right. It is highly unusual for two elements in the same complex to be lit from opposed directions (there is no comparable disjunction of illumination in Beccafumi’s known oeuvre) and some special circumstances would have to be found to explain such an anomaly. Among Beccafumi’s later altarpieces only the Annunciation in the Church of S. Martino and Sta Vittoria at Sareano, probably delivered in 1546, and whose subject complements the Mells panels, is lit from the right. Beccafumi’s autograph note of 1548 mentioning a ‘tavola d’atare con più figure e suoi ornamenti’ commissioned some three years earlier and delivered the previous year, is generally accepted as referring to the Annunciation whose ‘ornamenti’ would no doubt have included a now unidentified predella.³ But at 222 centimetres the Annunciation is some 60 cm wider than the three Mells predella panels combined and, if they were continuous, one, or perhaps two, further episodes would be required. If they were framed separately, then one would have to accept gaps totalling some 60 cm, which is possible but perhaps not probable. The matter thus remains open and progress in reconstructing the original situation of these panels is unlikely to be made until and unless they are examined in a laboratory.

Graham exhibited the present painting, together with the other two predella panels, at the Royal Academy Winter exhibition of 1884, as ‘Adoration of the Magi’, ‘The Visitation’ and ‘The Virgin in the Temple’, by Domenico Beccafumi.⁴