Exhibitions: Gorgeous by name, gorgeous by nature

Portraits by Ingres

National Gallery, WC2

What a resounding name was that possessed by Jean-Auguste-Dominique Ingres. There is a touch of the roll-call about it. His parents must have hoped that he was destined for great things and it seems that his father, who was himself a painter in a minor way, encouraged his son. One of the pleasures of the National Gallery's "" exhibition is to see the way that a provincial artist suddenly appeared at the centre of French revolutionary events. The earliest pictures in the show are pleasant, but not remarkable. They contain a lot of awkward passages, so that one would hesitate to attribute them to Ingres without documentary evidence. And then - we find a blazing and domineering surprise, the painting Napoleon I on His Imperial Throne, faultless in execution, full of power and glory.

This painting is well known in reproduction, but much less familiar in reality (it belongs to the Musee de l'armee in Paris). Its effect is astonishing. The painting is direct and forceful, as one would expect in any portrait of Napoleon. But it is also packed with significance, in the way that religious paintings are full of emblems. Yet the purpose and meaning of this image is entirely secular.

We feel that Ingres, with instinctive understanding of Napoleon's character, had wiped out and dismissed all previous religious art. Ingres was then only in his mid-twenties. Obviously the painting respects and honours many pictorial traditions. None the less, it was the work of a radical.

He was never again to be a radical in the usual sense of the word, and, of course, he is often regarded as an arch-conservative. Partly this is due to his longevity. Like Claude Monet, he attained a good old age, so that he lived into a new epoch. Ingres did not die until 1867, at just the time when Monet's early career was taking flight. To this day, it seems odd to think that Ingres had an effect on such a modern master as Edgar Degas, who revered him, met him and asked him for advice. There are different stories about the things Ingres had to say to his admirer. On the whole it seems that he told Degas to concentrate on line, to copy the antique and the old masters, and not to study nature.

Degas loved Ingres's portraits, and was influenced by them. He did not care whether they were old-fashioned or not. It was enough that they were so good. One has similar feelings about this National Gallery exhibition. The famous disputes about classicism versus romanticism, or the contrasts between Ingres and Delacroix, seem not to matter. We are in the show for one purpose alone: to look at the ways that an artist with consummate skill dealt with the way that his contemporaries looked. Their behaviour, allegiances or social position are of less account. We are interested in the ways that paint and pencil represent faces, figures and clothes. The Napoleon portrait is the only picture in the show that has dramatic political meaning.

For many people, a favourite part of the exhibition will be Ingres's drawings. There are 50 of these remarkable works. The earliest are profile medallions, done when Ingres was a student. Then comes a quite large portrait of the adventuress Barbara Dansi (we guess that Ingres liked her), which is in black chalk with white highlights. This is also a precocious student work, from the late 1790s, when Ingres worked in the studio of Jacques Louis David. Then, around 1806, we find a new style of portrait drawing that continues, with little real change, for half a century. The size of the paper is smaller (and would always be small); the subject or subjects are seen frontally; there is rarely any background, and the medium is always pencil.

The informative catalogue has much to say about the people who sat for the pencil portraits. Every one of them is identified. We learn little about the nature of the genre which Ingres made his own and obviously loved. The drawings are miracles of artificiality. The way that Ingres saw a person's features, then caught those features with a fine, uncorrected line, has few parallels in all art, including old-master art. Historically, these sheets update the "presentation drawings" of the high Renaissance. They also owe something to the tradition of tourist portraiture in Rome, where Ingres lived when his style was formed. I suspect that he was also impressed with the shrewd portrait busts of the later 18th century by Jean-Antoine Houdon. In such ways, then, the drawings are academic.

Usually, he did not draw to the edge of the paper. Furniture, musical instruments and clothes are outlined with summary and even mocking elegance. The effect is curiously modern. It is important that these portraits, though never disrespectful, have a sense of humour.

It's sad that Ingres's drawings did not translate into small paintings, say, a foot high. Then Ingres might have had a fruitful over-the-centuries dialogue with his beloved Raphael, who made lovely pictures with just these dimensions. Most of Ingres's painted portraits are of a standard, what one might call a bourgeois, size. As, of course, was appropriate, socially and aesthetically. There are still one or two paintings that strike me as having a wrongish size, and therefore give an eccentric scale to the figure.

One such work is Queen Caroline Murat, which is reproduced here. This is the first time that the picture (from a private collection) has been exhibited. Here we see Napoleon's youngest sister, who had married one of his generals. The Emperor made them King and Queen of Naples, thus extending his rule to smoking Vesuvius, which we see in the background. The Queen didn't like this bizarre portrait. Certainly, in academic terms, a number of things are wrong. For what reason, when we have such a professional academic painter? I think that Ingres approached Caroline in the way that he sat down with the subjects of his portrait drawings. His mistake was to paint her in oils.

He turns out to have been a more adventurous colourist than we imagined. I do not believe the old story that he was happier when painting unintelligent women. A portraitist is by nature a servant. Ingres was none the less manly with his male subjects, as a painter should be. He was an accurate, also suggestive painter of costume, better in this department than even the impressionists were. His stricter sense of deportment helped the dignity of his art. And he needed dignity, for he had known both the terrors of the French Revolution and the comparatively settled arrangements of the French middle-class state of the 1860s. We should class him as a realist more than a classicist.

National Gallery, WC2 (0171 747 2885), to 25 April.