14. Implications of infrared reflectogram

It has been argued for many years that the Virgin of the Rocks in the National Gallery in London is a copy done by Leonardo and one or more of his assistants to replace an original commissioned for the Chapel of the Fraternity of the Immaculate Conception in Milan and either sold or taken away to France by Leonardo himself.  That painting, considered the original and autograph version of the Virgin of the Rocks, is now in the Louvre Museum.

The angle of the head of the Madonna that
 lies  beneath the painting is similar to this
study for the Virgin and Child with St Anne 

Preparatory to the eighteen month restoration of the National Gallery's Virgin of the Rocks completed in 2010, it was examined using infrared technology by a team of experts from Florence.  To the surprise of the painting’s custodians, beneath the surface there appeared clear evidence of a very different composition to the one now visible.  This includes portion of a face and two sketchy hands.

The question has been asked “Why did Leonardo abandon this composition and paint instead a copy of a work that he had already painted before?” 

Perhaps this is the wrong question to be asking.  The implications of the discovery of the face beneath the painting  give support to a large body of evidence that, in fact, it is the Virgin of the Rocks in the National Gallery in London that is Leonardo’s original.  Even that it has been apparently worked on by painters other than the master himself supports rather than detracts from the opinion that the London Virgin of the Rocks is the earlier of the two.

See the National Gallery article, "The Hidden Leonardo"


Leonardo was commissioned to paint the Immaculate Conception.  What he painted and delivered was one or other of the compositions now known as the Virgin of the Rocks.  If the painting in London is the earlier version, then what lies beneath the surface is presumably an earlier version still.  Can the infrared image assist us in piecing together the evolution of the sublime works of art that we now know?

The infrared reflectogram presents us with a face, presumably that of the Mary the Mother of Jesus, turned at quite a different angle to that of the finished painting.  Her head is tilted away from us in the opposite direction and where that of the finished painting is inclined downwards, that of the infrared image is tilted upwards from the viewer’s perspective so that we see the undersides of her eyelids and nose.

The position of the Madonna is similar to
that of St Jerome in this unfinished work.
The right arm stretches out towards the border of the painting and the hand is reflexed, as if in surprise or awe.  Mary’s left hand, which under the infrared scan made its appearance so dramatically through the face of the finished Madonna, is touching the breast in a gesture of adoration.  There is an animation and intensity about this figure which is not present in the finished product.  This may be explained by the artist’s change of subject matter. 

The very existence of the earlier image beneath the present painting in the National Gallery alerts us to the fact that Leonardo, at the time he took up his brush, was still formulating the final subject matter.  The draught of the painting that was sketched onto the board was not, as it has been called, "a different painting".   It was the penultimate design in the development of the masterpiece that we see in the National Gallery.  The reason why it is so different in form to that which covers it, is that Leonardo had two very different concepts in his head simultaneously.  We can see the evidence for these ideas worked out in ink on a single sheet of paper in the Metropolitan Museum, New York.

When the panel was examined under the infrared scan, not only were the details of the earlier drawing found, but also the draft for the present painting, painted over a second prepared surface. Both the earlier composition of the Virgin in Adoration and the composition that was to become the Virgin of the Rocks showed the artist working in two ways. In both cases there were areas that were drafted in a precise manner as if from a prepared study, and there were other areas that had been loosely sketched. In the underlying composition, the head and left hand of the Virgin were carefully drafted while the outstretched right hand was sketched freehand, with a number of changes.  The backdrop of rocks had been sketched freehand.

It has been observed by the curator of the London painting, Luke Syson, that the finished work does not follow the second set of drafted outlines exactly, and in places, such as the composition of the rocks, deviates a great deal. In all these areas of deviation, the finished National Gallery painting is more like the Louvre painting than the underlying sketch.

One such difference is that the underpainting reveals long curls hanging beside the right cheek of the angel. These have been eliminated in the finished work.  In a publication by the National Gallery, it is  noted that it is "strange" that, in the finished painting, the artist has omitted the curls. In seeking an explanation, the conservators refer to the painter of the London version as "reverting to the earlier work in several details". The explanation continues: ''This may indicate that Leonardo's assistants, who helped complete the painting, found it quicker and easier to copy the composition of the existing Louvre version, rather than interpret the new elements Leonardo had sketched in.'' 

With regards to the changes to the rocky background, Syson writes: "Other parts, once again, are more freehand, including the rocky cavern setting, which Leonardo first drew as a free variation on his first idea, before delegating it to an assistant who brought it closer to its prototype." (i.e. the Louvre version)

These two statements can be feasible only if there is a total rejection of any notion that the National Gallery painting is the earlier. Logic indicates that the reason why the angel's curls were omitted from the National Gallery painting is that they were part of the initial drafted design, and that Leonardo himself decided to omit them. The different arrangement of the hair was not "copied" from the Louvre version, but invented as the artist saw fit. The pentimenti indicate plainly that the London painting progressed and evolved during its creation process. The hair of the angel in the Louvre is similar to that in London because it is copied from it, not the other way around. 

This detail of the London Virgin of the Rocks reveals an extraordinary
understanding of the process of disintegration of rock surfaces.  It is plainly
not the work of an assistant copying from a less-detailed painting.
Similarly, the reason why the sketchy design of rocks beneath the surface of the London painting is different from the finished composition is that it represents a stage of Leonardo's design process. The precisely-detailed and highly finished grotto of the London painting represents the end-product of this process and the much less detailed grotto in the Louvre painting represents a derivative version.

Moreover, to suggest that the London painting's superb depiction of the fractured, flaking, crumbling and eroded rock, exceeding even that of Mantegna in virtuosity, is from the hand of an assistant, copying the indistinct forms of the Louvre painting, is to do the Master a serious insult.

See National Gallery article: Who did the original drawing?

Documented: 4th November, 2005,  completed Sunday 6th November 2011
Copright:  Tamsyn Taylor