In the middle room of the Leonardo show at the National Gallery you can swivel on one heel and see, almost simultaneously, the two versions of his Virgin of the Rocks. They face one another across 15 yards or so. There is no reason to think the two paintings will ever share the same space again, at least in my lifetime, and maybe they never have before. For the longer one looks at the pictures and puzzles over what scholars have to say about the scrappy documents that mention them, the less likely it seems that Leonardo painted the one in sight of the other. The story of the two paintings is typical of his life. In style and atmosphere, the version owned by the Louvre – hats off to the French, with just a touch of incredulity, for having let the much abused and vulnerable object cross the Channel – looks to have been done no later than the mid-1480s. Kenneth Clark even thought Leonardo might have brought the panel with him from Florence to Milan (though this cannot be right). Let’s say it was finished by 1485. Then something went wrong between Leonardo and his patrons.
The picture was originally intended as one element, central but not necessarily dominant, in a gilded and sculpted altarpiece ordered by the Confraternity of the Immaculate Conception in Milan. This club of notables had only recently been founded, and the elusive dogma it was supposed to celebrate was still, shall we say, emerging from the theologians’ mist. Sermons grew longer, things barely clarified. The Louvre painting never reached the Brothers’ chapel. Documents from a few years later speak to Leonardo haggling over price. His stock had risen in the 1480s, the painting seems to have been seen from the start as something unprecedented and wonderful (the Louvre’s is the version copyists soon fastened on), and it may well be that a courtier or emissary had made him a better offer. Some scholars wonder if the Confraternity balked at the picture, considered as an exposition of its great Immaculate theme. It does seem to make more sense – though nothing, in my view, is ever going to edge the fantastical world of grotto and meteora into a safe Marian frame – as an illustration of a popular story from the Apocrypha, which has the infant John the Baptist, in flight from Herod’s executioners, come across Christ on his flight into Egypt. Though never was Egypt like this. Maybe the trigger for the temple of boulders was a line from the Song of Songs: ‘My dove in the clefts of rocks (in foraminibus petrae), in the cavities of walls (in caverna maceriae), reveal your countenance to me.’ This reads to me as a high prefiguration of Yeats’s ‘I offer to love’s play/ My dark declivities’ (and I dare say did to its first audience) but I accept that by Leonardo’s time it had been thoroughly allegorically cleansed. Whether the clefts and cavities are given back in the painting something of their Old Testament charge – erotic, portentous, delicate, not quite real – is another matter.
The painting eventually delivered to the Confraternity is the one now owned by the National Gallery. ‘Eventually’ here is an understatement. Twenty or more years after the first Virgin of the Rocks was completed (and we know nothing of that painting’s early whereabouts until it is mentioned in France a century later) we have a document dated April 1506 putting it on record that the Confraternity’s Virgin was still not done, and that Leonardo was obliged to hand the finished article over within the next two years. He did, in 1508. There is an unimpeachable later paper trail for the London picture – Gavin Hamilton, Lord Lansdowne, the Earl of Suffolk – and there can be no doubt it came from the Immaculate altarpiece. But was it Leonardo who painted it, finally? The London panel is a formidable object, brilliant, ice-cold, entirely conscious of its power; and connoisseurs, almost as much as the general public, have always found it hard to like. When Clark toyed with the idea of the Louvre version coming literally along the road from Florence he was only dramatising a widely shared intuition. The painting in its original form is the last masterpiece of the Quattrocento. And ‘Quattrocento’, however hard the experts work to dismantle the category, remains a magic value – maybe particularly in these pre-Raphaelite islands. Those of us who go on wondering what Leonardo might have been if he had not become a court artist, a peripatetic (nostalgic counterfactual, but did not Leonardo have the same regret?), look from the Paris to the London Virgin and consider that art here took a fatal turning. Maybe inevitable, but chilling. ‘The Louvre version of the Madonna of the Rocks is so superior to the London one,’ Wölfflin wrote in 1899, ‘that it seems inconceivable that its originality could ever have been doubted.’ (This was before the archives coughed up evidence, such as it is, for why we have a second painting at all. And in 1899 the Paris picture presumably looked even better, and the London almost certainly worse. The whole story is exceedingly strange: it is as if Shakespeare at the time of The Tempest had been forced by an old contract to rewrite Richard III.)
One way out of this quandary – for it does not feel right to historians that the golden years of the High Renaissance should leave behind such a central, equivocal instance of ‘progress’ – has been to doubt that what we see in London is really Leonardo at all. But this will not do. The issues are complicated. X-rays, for example, show an underworld to the London panel that has nothing in common with studio copying procedure; none of Leonardo’s possible collaborators in Milan seems remotely capable of, say, the modelling of the Virgin’s or new angel’s face (and seeing Boltraffio and de Predis close by in the present show confirms this); there are episodes of outright sketchiness and unfinish in the new Virgin which surely a studio team would not have dared pass off. But essentially technical questions cede to an overall intuition, which having the paintings together at last confirms: the London painting is Leonardo. No doubt there are passages in it done a bit mechanically by assistants, but the overall conception and execution is under one man’s steely control. It is what the control was intended to do, to the scene and our access to it, that is the difficult question.
I take my cue from the top of the two altarpieces, and what happens in the London one to the rocks in relation to the sky. The sky is edged out of sight, largely, and the confetti of leaves and grasses disappears. The cave becomes a lid to the illusion. Everything in 1508 is pitched forwards, nudging against the picture plane – claiming proximity to the viewer, enfolding or almost touching us, insisting on presence. The other world is here in the room.
The Paris painting has yellowed and darkened, but it is surprisingly easy in the National Gallery show – the lighting somehow cuts through the protective glass cage – to see and imagine it as it once was. Even when the picture’s play of light was stronger, the lit areas of flesh and drapery would have been decisively smaller than in the London revision, and put in a different kind of tonal relationship to the rocks and even to the drapery in shadow. The forms in the sun are not yet polarised, taking on London’s ghostly pallor; they are less spot-lit, less front-stage. The London picture reaches towards us, aims itself at us: the Virgin’s hand is the emblem of that. Internal to and fro between actors and places in the scene (grounds, distances, caverns) still goes on in 1508, but cedes to the general turning in our direction, the tremendous forward reality effect. The leaving out of the angel’s strange pointing finger is another sign of this: the sense one has in Leonardo’s first set-up of the figures still orienting themselves in relation to one another, still feeling their way to an understanding of the miracle they take part in, gives way in London to sheer presentation. The Virgin’s hand in the repetition releases the head of the Christ child, almost like a rabbit from a hat. This is not in the least true of the passage (of two hands and head) done two decades before.
The Christ in London teeters, just a trifle too dramatically, on the edge of a cliff. (The whole bottom precipice is higher and showier, and has less of a flat supporting rim of bare stone.) The light and shade project Christ into our space. From being a child with a child’s body, canopied by two grown-up hands, he has become a miniaturised Michelangelesque athlete.
Everything concerning drapery – the word sounds workaday and technical, but it is the key to Renaissance painting’s sense of the body expanding and luxuriating in the world – is hardened and lost hold of in London. The most striking case is the way the Virgin’s robe meets the ground. In Paris it crumples slowly, heavily, elaborately: one feels the material falling and spreading as the result of its own weight and consistency. And it is gently analogised with the folded rock strata below. All of this is gone in version two. An inflated blue cylinder puffs out suddenly at ground level, like a table decoration. Analogies between it and the earth it rests on have become unthinkable: if it rhymes with anything in the picture now, it is with the central rock column above (which has likewise been fine-chiselled and dramatised in comparison with Paris, turning into something from a Wunderkammer). Looking at the Paris picture, after a while you realise you are being shown the kneeling angel’s naked foot, bearing weight, resting on the grass just visible behind Christ’s back. It is an odd-looking extremity: one scholar thought it had harpy’s claws. But it is a foot; and the way it completes a possible body for the angel, however unlike bodies we are used to (the way we are shown the angel feeling its way across the grass), is, well, Quattrocento. In London blue drapery just muffles the point.
The flowers in London are paper – or anyway, too preserved and immaculate, with not a trace of natural weakness. In the Paris painting – we are sometimes advised to despair in the face of its dirt and discoloured varnish, but again, the present showing coaxes the object back to life – there is a time of day, I think, a sunset touch; and the condition of the atmosphere is worked out in the detail of light in the foreground. Experts talk about the later version going further with Leonardo’s mature investigations of modelling and materialising the illuminated body – the contours and orientation of the Virgin’s face are good examples – and they are right; but the substance, the cold three-dimensionality, is abstract. The light is not of this world. No doubt this was deliberate and in a sense appropriate (maybe the Immaculists approved), but what goes by the board in London is the very thing we may value most in Italian painting: the sense of the sacred belonging to a reality we recognise, and one whose strangeness is built from the strangenesses of nature. Look at the river and mountain landscape on the left. In the Paris painting it is otherworldly, yes, but for that very reason familiar. Its space and atmospherics are those of the Alps going north from the Lakes, the wild Chiavenna we know Leonardo delighted in; and the onset-of-sunset glow reminds us of everything in non-supernatural experience that regularly transfigures things and makes them mysterious – makes nature not ours. Substituting glacial blue for pale yellow, as I reckon Leonardo did in 1508, is putting a (marvellous) lantern slide in place of a true act of memory and imagination.
I go too far. The London picture is prodigious. The point of description cannot be to demote or belittle it – in a sense I think we have never, since Wölfflin, given the painting its due – but to grasp what kind of prodigy it is. Let me start from a typical transfixing Leonardo detail, the unfolded yellow lining of the Virgin’s cloak. In the Paris picture (whatever the changes brought on by time) there was a connection, I am sure, between the yellow fold and the light in the sky. The lining, spellbinding as it is – separate from and superior to its being a condition of some stuff in the sun – is a dream condensation of the yellows of late afternoon. In Paris the lining still has softness: it is conceivable as folded, touched by human fingers. The yellow emerges, at first quite gradually, from under the stretched canopy of blue. It is the inside of a garment spilling out. We may see it as a concentration of the landscape light, but also as a way of bringing that far light closer – optically, seemingly accidentally – in a manner that viewers can accept as actually happening, here among the rocks.
Does any of the above apply to the fold in London? I doubt it. The lining in 1508 participates in the painting’s overall crystalline abstraction – its nearness without tangibility. It is of the essence of the London panel – the key to its invention of a new kind of pictorial space – that everything in the foreground, yellow, blue, marble grey, ghostly grey-green, takes on from the world in the distance only the river’s bleak cold. The light on the babies’ puppy fat seems to issue, if it comes from anywhere, from a source close to where we (and the worshippers) are standing, here on the other side of the picture plane. The scene, to repeat, has the sacred be our possession. And that rule of nearness, of abstract proximity, applies even to the landscape. The rocks are a niche from which figures come forward – a niche with glimpses into a grotto – and their substance and architecture have to be as sharp-focused as the light on the Christ child’s arm. Of course things far away are somewhat subdued: they are meant as a foil or a backdrop; but not for a moment now are they indistinct, or likely to elicit a dream movement into a ‘beyond’.
Behind all this, I suppose – I mean my recoiling from London clarity – there is the shade of Freud’s Leonardo, which is still essentially Walter Pater’s. I follow the two grandees in thinking the key to Leonardo’s continuing hold on us is his combination of dreadful curiosity with continual visual parapraxis, unstoppable reverie and over-determination. Just round the corner from the two Virgins in the exhibition there is an appalling, exquisite sheet from a Leonardo notebook in Windsor, with a rare date inscription: 2 April 1489. On the page are two pen studies of a skull, with a pulpy eye still in the lower one’s socket, and what appear to be arteries still clinging loosely to sockets and temple; and then to the left there are six and a bit lines of mirror writing, carefully balanced, as a block of script, between the observations top and bottom. Ethereal pen lines spread out from the skull below, going as far as the writing: a halo, an aura, an atmosphere, an act of pictorial hallucination. This is Leonardo, surely – inquisitive, cunning, ruthless, fragile, impenetrable. (Maybe the page was dated because the eye in the socket was already putrefying.) And this Leonardo is continuous, I think, with the Leonardo of the best paintings. What I miss in the London painting, it follows, is a sense of encounter with the world still going on; or an encounter still capable of destabilising the artist’s immense gifts (his control, his ruthlessness). Look again at the angel in Paris with the pointing finger. There had never been a figure like it before in painting, and there never would be again. It is not just the foot on the grass and the pointing finger that are uncanny: it is the pose as a whole, spreading out laterally, half turning towards us, but meeting our eyes from an utter remoteness; and the roll of the green sleeve and the long pale arm inside its diaphanous puff of drapery; and the astonishing – unthinkable – explosion of rich red, tying the figure to a world of flesh and blood but spreading and unfolding far beyond (it feels like) the mere frame of the illusion. No wonder all this – this overflowing wish-fulfilment – had to be corrected in 1508. The London angel’s drapery is still elusive, but at least now it adheres to a possible anatomy: it does not just seem to occur as the brush tries out a new colour. Colour drains away. The angel’s shoulders come out of their carapace. The figure is clear and coherent (comparatively) but also (comparatively) unfelt. It is as if the invention had been put back inside a box – robbed of its first fairytale unfolding.
Also around the corner in the show, on the wall facing the skulls, is Leonardo’s desiccated Saint Jerome. It is, in its abandoned schematic state, a difficult picture to warm to. But then it dawns on one that Jerome and the Paris Virgin of the Rocks – they cannot be more than a few years apart in date – essentially repeat the same fantasy. They envisage a wilderness, a place the everyday world will not reach, and which therefore can offer abiding shelter, safe surroundings, protection. I go along with Freud in thinking this a central, overdetermined wish. And the ebbing away in London of the Quattrocento’s palliative, reparative atmosphere – its hortus conclusus softness and naivety – has truly about it, I think, the chill of disillusion. The world in the painting’s redoing has become all outwardness: there is no place of shelter in it: light makes everything frigid and unyielding. Just as much as ever – Leonardo will never stop dreaming – this is a fantasy world, only now the imagined (maybe wished for) pictorial state is of endless stillness, perfect balance, invulnerability. The little naked boys are already tomb sculpture. From being a dream of freedom their nudity has become a kind of ghostliness. The Virgin seems to want to protect them from an invading ice age, but her blue is the carrier of it. In Paris she had been a participant in a play of infantile homage – young, frail, doing without a halo. In London the lost dream-mother cedes to a Mary already seeing (cradling) the cross. The folds of yellow at her midriff are like festival armour buckled in a joust.