A friend and I bought this portrait of King Richard III from Christie’s South Kensington last summer.

It was catalogued as Eighteenth Century,  and relatively cheap at £1,000 – 2,000.  The cataloguer suggested that the Christie’s painting was a copy of the National Portrait Gallery’s King Richard III (NPG148) – a mid-Tudor copy of a lost original. The flat handling of the coat and background certainly made it look a bit of a pub sign. The inclusion of twee medievalising details didn’t help. The weakly-painted coat of arms and the ivy twining round the RICARDVS III inscription reinforced the sense that you were looking at a later pastiche.

Several things jumped out, though. For one thing it was painted on an oak panel, and this gave us hope. Also, where the split in the panel to the left of the King’s head passed through the chain there was obvious repainting. As soon as you knew that there were several periods of work involved, you could start to unpeel the layers, and start to reconstruct what might be underneath. The ivy and the coat of arms looked the most recent, perhaps as new as c.1900. But beneath the ivy, the style of the RICARDVS III letters was entirely convincing, compared with this example of a portrait c.1575 formerly with Philip Mould and Co.

Historical Portraits (c)Philip Mould and Co.

The dimensions of the Christie’s picture were also encouraging. The panel measured 18 1/2 x 14 3/4 inches. The standard English panel size by the second half of the Sixteenth Century is around 23 x 19 inches. If our painting had lost about four inches at the bottom and two on each side the original composition might have been closer to the portrait shown above, showing the whole of both hands.

We couldn’t get to London to view, but Christie’s website lets you blow images up quite as closely as you could ever see them in real life. This isn’t infallible  – some aspects will only be apparent to you viewing in person – but it seemed that we were dealing with an older painting that had been repainted after a cut-down job. Details like the translucent highlight on the thumbnail, and the impastoed jewellery of the chain smacked of authentic Tudor technique, but there seemed to have been an extensive campaign of overpainting, probably in the Eighteenth Century, as the cataloging suggested, which may have been when the panel was cut down. There was thick, discoloured lead-white overpainted on the fur lapels, and at the time I thought that the face might be completely repainted, as well as the hat. Taking it to the restorer and ‘having a go’ was irresistible.

Here it is in the restorer’s studio, after an initial scrub. The portrait was painted over a light ground on a gesso-prepared panel, so any remaining anxieties about date were quickly dismissed. Only a fanatically antiquarian pasticher would go to such lengths, and it was clear that when this work was painted in the living tradition of early English panel painting.

As we expected, the ivy and coat of arms were basically modern, and came off with solvents immediately. The opaque background was also modern, and our restorer was excited to find a beautiful cool grey underneath it, which gave the painting depth and pushed the figure forward in the picture space. The earliest campaign of overpainting did seem to be Eighteenth Century. The glimpse of red above the King’s gold tunic was a bold slashing stroke of thick pinky-red impasto, like a fat Z, which looked pure 1700. It was done quite casually and overlapped the gold tunic, proving that it was a later addition. I was quite in favour of keeping this detail. I love a good bit of impasto. It was a beautifully physical piece of brushwork, and part of the archaeology of the painting, something that proved its age, like the blocked windows and remodelled fireplaces that give an old house character.

Comparing our portrait with other versions of the same composition showed that the left hand ought to be emerging from a fur cuff. I was convinced that there must originally have been a similar cuff on our portrait, and I was thrilled when I saw the portrait on the restorer’s easel. You can see the semi-circular shape at the bottom of the panel-split, showing through the repainted coat like the foundations of old buildings showing up in aerial photographs of a field. The disturbance to the paint surface suggested that this detail had been straightforwardly overpainted, and might still be preserved intact underneath.

As you can see from this photo of the restored portrait, the cuff was there and came out beautifully. Through being covered up, rather than scoured and repainted, the fur of the cuff survives much better than on the lapels, and has a real softness and three-dimensionality. There are even little hairs standing out from the edge of the fur and visible against the dark of the coat.

Rediscovering the fur cuff confirmed that our restorer had got down to the original surface of the King’s doublet.  When the split was first mended, the Eighteenth Century restorer used a thick plaster filler and then repainted on top. Perhaps this was also when the panel was cut down at the bottom and sides. At this date, the coat was brushed over in a single matt colour. Our restorer showed that the purple velvet coat was originally painted in a mulberry glaze over a dark red under paint, and returning this translucence to the painting was one of the last stages in restoring it to life.

There were some further surprises. We were initially unsure how to restore the neckline, where the later impasto overlapped the gold. The Eighteenth Century restorer had added a further line of gold tunic between – what I cannot help thinking of as – the King’s white t-shirt and the red band of the under tunic. This would have to go, because it was a misreading of the costume, and so the restorer removed all the later paint from that area. Extraordinarily, he discovered a perfect, beautifully painted piece of Tudor scarlet, as seen in the above image. Who knows why it was ever overpainted. Perhaps it had gone dark with dirt and old varnish, and the old restorer had slapped on a few strokes of bright new paint. This had gone pale in time, whilst the original looked as fresh as when it was painted.

After dirt and old varnish were removed, it turned out that the face was original after all, and came to life again, with traces of old underdrawing visible in the features. The only area of later overpainting that we left was the jewel in the King’s hat. The pearls are substantially original, with red reflections cast in them, but the filigree gold details must be a reinterpretation of the original rhomboid jewel, where the black spaces at 12, 3, 6 and 9 o’clock would originally have been diamonds, which were painted as black at this date. The original details may be there underneath, but they might equally be lost through abrasion. We decided to keep the most accomplished, sensitive and beautiful part of the historic restoration rather than replace it with modern work, and this feels like a proper way to show our respect to the early restorer, and to the history of the portrait itself.

The date of the portrait was an important question. There two oldest-known portraits of King Richard III are a distinct type in reverse to our version, dated c.1520 and belonging to the Society of Antiquaries, and the version in the Royal Collection. Both are believed to be a copy of a lost portrait – or portraits, perhaps – painted from life. The Royal Collection panel is datable to c.1504 – 1520, and it is first recorded in an inventory of King Henry VIII in 1542, along with portraits of King Henry V, King Henry VI, and King Edward IV. This set commissioned by King Henry VIII or his father King Henry VII is probably the first example in this country of the corridor set of Kings and Queens that would become a feature of country houses until early in the following century.

Royal Collection Trust/© Her Majesty Queen Elizabeth II 2014

The NPG’s portrait is a very close copy of the painting in the Royal Collection. Like all Tudor portraits, different workshops would produce their own variant on the established pattern. This illustration of four known examples illustrates some of the different types (from The Trial of Richard III Richard Dreweatt and Mark Redhead, Sutton Publishing for Channel 4, verdict: Not Guilty). Our portrait is closest to the painting indicated by the mouse top right, a portrait in the collection of the Duke of Northumberland at Syon.

The Duke of Northumberland’s painting is a superb example of a corridor portrait. Pamela Tudor-Craig, King Richard III’s iconographer, dates it to ‘probably 1600, maybe 1620’ (Lady Wedgwood in The Trial of Richard III p.108). The Duke’s archivist very kindly sent us a better image of this painting for study. It was interesting to see that whilst the Syon painting represents the same type as our painting, the characterisation is more generalised and the sense of detail, the ‘reality’ of the objects depicted is less sharp. The fur of the Syon portrait is broader and flatter – there are no delicate hairs overlapping the dark of the coat – and the relation of the fur trim at the front of the gown to the cuff is interpreted differently; they are treated as a single length as though the gown loops up over the King’s arm like a toga rather than forming a separate sleeve. The square diamond with drop pearl in the hat of the Syon portrait is a very Jacobean jewel. It compares with the one in Lady Mary Wroth’s hair, or on the breast of the unknown lady by Cornelius Jonson. It could be a later repainting, as with our portrait, but the jewelled collar of the Syon portrait is too delicate, and this must be part of the portrait’s original conception. The form is accurately recorded, but there is no sense of the sheer heft of it as it falls across the King’s shoulders.

By contrast, the encrusted collar with its engrossed gold detailing is one of the most impressive passages in our portrait. It was very exciting when the restorer passed me his magnifying goggles so I could make out the the little crown shapes beautifully worked in the angles of each gold panel of the chain. Whoever painted it had a sense of how a collar like that looked and felt in reality. My friend wondered if our painting might be original source of the Syon portrait. It might never be possible to construct a precise chain of descent for these workshop portraits, but I have no doubt that it represents an earlier generation of the type, and details such as this argue for a date in the late Sixteenth Century.

And lastly, the King’s thumb…

You’ll see from the portrait in the Royal Collection, our portrait and the other four illustrated in The Trial that the King is gripping the front of his robe with the fingers of his left hand, and the thumb of that hand disappears behind the other side of the fur trim, or into the shadow under his jewelled collar.

All except the NPG’s portrait in which the King’s thumb stands out as a sort of skeletal thimble shape. I’ve always been bothered by this, ever since I hung a poster of NPG148 on the wall of my first London bedroom in 1994. Even allowing for the imperfect anatomy of medieval painting it struck me as badly drawn, and a rather glaring error in such a beautiful picture.

(c)National Portrait Gallery, London

I went to see the original and wondered whether this was a later repainting, to emphasise the alleged weakness of King Richard’s right arm. But the answer was in my Grandmother’s copy of Through the Looking Glass. If Dally liked a postcard she would put it in the book she was reading. She was a great fan of King Richard III, as I’ve mentioned before, and in 1957 Con her eldest sister sent her this postcard from the NPG.

It’s interesting to see that at this date NPG148 had broad framing gilt corners like the Royal Collection portrait. But at some date between 1957 and 1994 the painting was cleaned. Rediscovering the delicate floral traceries was definitely a good thing; but it’s a shame that the shadowing at the end of the King’s thumb was cleaned off as well. In Aunty’s postcard you can clearly see how this dark glaze causes the tip of the thumb to recede, into the shadow under the chain as if the thumb is just reaching behind the fur trim of the doublet. It’s a superb gesture, this clutch of the gown in King Richard’s portraits. It anchors the gesture of his hands in taking off, or putting on the ring, and gives his pose a mobile dynamism, a sort of coiled-spring sense of oomph. Previous kingly portraits tried to make the hands busy and lifelike, but they never quite seemed to belong to the sitter; the chain of gestures in King Richard’s portrait solves this problem brilliantly. They are emphatically his hands – the hands of a man of action, with a restless motion that reflects the working of his mind.

One of the most remarkable things you learn working with pictures is the effect that a single shadow, lost or replaced can have on the reading of an image. How it turns flat into real and makes a portrait come alive in the frame. I would love it if the NPG could put this missing glaze back again. I think they would be amazed at the result.

Update: since writing this post I am very grateful for correspondence with the National Portrait Gallery, which reveals that the thumb has been foreshortened to suggest a bend at the joint, and that the shadowing visible in the 1957 reproduction was not original paint but the result of old restoration.