Neil MacGregor, who is shortly to leave his post as director of the National Gallery, has been an outstanding helmsman, especially for his exhibitions and educational work, and for holding the line against admissions charges. But he has stifled the institution he so effectively nurtured. As director, he had only to uphold the role of the gallery, but instead he acquiesced to Nicholas Serota's misjudged ambition to make Tate Modern this country's national gallery of post-19th-century art.
So what has happened to the National Gallery? In Room 4 hangs one of its latest acquisitions. It's a little painting by Hans Wertinger (1465-1533). If you don't know who he is, do not worry; there is no reason why you should. The National Gallery has bought a bummer - a little panel barely a hand high and two hands wide, showing peasants scything corn, shearing sheep and carrying fruit, while gallants ride to the hunt against the backdrop of a lopsided hamlet perched on the bank of a river bouncing with outsize ducks. All very charming, no doubt, but there is not a millimetre of original observation or feeling in this travel-brochure confection of late-medieval life. And we're not even seeing the whole picture. The label tells us, in case we miss it, that "on the right is a cart with the lower part of a female figure seated on it". If you're thinking Wertinger is a proto-surrealist, I'm afraid you'll be disappointed; the painting has merely been cut down. It was probably part of a longer panel decorating a piece of furniture, and an unexceptional one at that. What is it doing in the National Gallery? It is certainly not a great painting. And if you've now become confused as to what that is, you need only glance to the right to see Hans Holbein's The Ambassadors hanging in the same room.
It would not matter so much if the Wertinger were just an endearing mistake by a great institution; the trouble is that it is a symptom of a much deeper malaise. The National Gallery is beginning to die, and the tragedy is that it is being killed off. It began to ail in 1998, when it was decided, without any public airing of the consequences, that the gallery's collection would no longer grow as the art of painting itself grew, but would be terminated at 1900.
The National Gallery was founded in 1824 when the government, with rare foresight, bought a small group of paintings to establish a gallery for the people. It took a while to find its feet, and could easily have become a closed collection, especially as many people were beginning to argue that, with the advent of photography, painting was dead. However, the keepers of the National Gallery were determined to make its collection one of the greatest in the world. Not even the Louvre, nor the National Gallery in Washington, nor the Hermitage, now rivals our National Gallery in the quality and breadth of its holdings.
This extraordinary flowering of the National Gallery's collection was partly a by-product of imperial wealth, but owed just as much to the vision and efforts of its staff. Almost all of the most famous pictures in the gallery were bought (never an easy option, even in Victorian Britain), such as Van Eyck's Arnolfini Portrait, in 1842, Leonardo's Virgin of the Rocks, in 1880, the Wilton Diptych, in 1929, Poussin's Adoration of the Golden Calf, in 1945, and Henry ("le Douanier") Rousseau's Tiger in a Tropical Storm, in 1972. The curators' aim was simple: to present to the public the greatest achievements in painting wherever they appeared. They diligently searched through the art of the past to discover works of genius that had been overlooked, but they did not just stick to the old masters; they continued to acquire the art of their time. So John Constable's Hay Wain entered the collection six years before Jan Vermeer's Young Woman Standing at a Virginal, acquired in 1892. No one could accuse the curators then of not being broad-minded in their taste or brilliant in their judgement as to what constitutes a great work of art. The same cannot be said of the curators at the National Gallery today.
Recent acquisitions have, for the most part, only added historical footnotes, rather than augmented the gallery's artistic standing. Three little early Italian panels have recently appeared in the Sainsbury Wing. The unknown Umbrian master's Virgin and Child and Man of Sorrows diptych merits the anonymity of its maker. The ambitiously attributed Cimabue is an extremely minor production by this most elusive of painters. The label extols its "new naturalism and tenderness", demonstrated by the "Christ Child touching his mother's hand". If you can make out this minute detail through the surface grime, it does not even begin to compare, aesthetically, with the Virgin and Child with Saints by Duccio hanging nearby (bought in 1857) - an extraordinarily natural and poetic conception. When the Clarisse Master's Virgin and Child was first exhibited - the third new acquisition in this group - the label justified its presence in the National Gallery by claiming that "it represents the type of Byzantising work from which Duccio developed". This label has now been replaced, and we are told that the picture is there because "Christ's becoming a man is recalled by the main image". The first label was more honest, for the picture has joined the National Gallery because it fills a gap in art history, not to promote Christian belief. No mention is made of its aesthetic merit because none can be: the panel is little more than sweet.
There is a worse example in the galleries of 19th-century work: a painting by Francois-Xavier Fabre. Although accepted as a gift in memory of the spirited art historian Peter Murray, this pleasant landscape, with its peasant buildings and paw-print leaves, does nothing for his reputation or the National Gallery's. At least it is little. The same cannot be said for Joachim Beuckelaer's whopping canvases, The Four Elements, bought in 2001, that now nearly fill Room 14. They are huge still lives, showing fish, birds, beasts and vegetables with pocket-handkerchief religious scenes behind - all energetically but repetitively painted, and not by any stretch of the imagination great art, unlike Pieter Bruegel the Elder's extraordinarily original and intense Adoration of the Kings (bought in 1920), now elbowed into a corner in the same room.
But can the National Gallery collect great paintings today? Are there any left to buy? Well, there are still byways of art history to be explored. The gallery has been right to add, over the past decade and a half, Caspar David Friedrich's Winter Landscape, Christoffer Wilhelm Eckersberg's View of the Forum in Rome, Thomas Jones's Wall in Naples and, more arguably, Theo van Rysselberghe's Coastal Scene - none of them great paintings of world class, but gems, nevertheless, which do enrich the gallery aesthetically, with the additional advantage that they do not take up much room. But all these pictures essentially fill gaps (except, perhaps, the Friedrich); they do not add new dimensions to the art of painting. Nor can the gallery do this easily now that it has voluntarily beheaded itself at 1900.
The collections of the National Gallery were acquired by people with a commitment to art, not art history. Several of the early directors were artists rather than academics. They knew great art when they saw it, and did not hesitate to acquire it. Many of their acquisitions were bold and inspired. Vincent van Gogh's Sunflowers was bought in 1924, only 36 years after it was painted - not bad judgement at the time. Edgar Degas's wonderful, unfinished Combing the Hair (1896) was bought as soon as 1937. It had been owned by Matisse. If only the National Gallery had been quick enough to acquire that artist's Red Studio when it was bought by the Museum of Modern Art in New York, what a fabulous pair those paintings would have made. But the National Gallery has no Matisse, nor a Picasso, so how can it hope to tell the continuing story of art? Some time in the 1980s, the gallery began to fossilise, and many now think it has missed the boat. But there are still great Matisses in private hands, and even more Picassos, not to mention superb works by other leading painters of the 20th century. Naturally, it would be a challenge to add modern masterpieces to the National Gallery, but what an invigorating one for curators and the public.
One short cut could be taken at once. The National Gallery used to take any paintings it wanted from the Tate after 50 years or so (for example, Van Gogh's yellow Chair, in 1961). This was good for both institutions, particularly in the field of British painting, where a permeable relationship needs to be maintained to allow both Tate Britain and, on a much smaller scale, the National Gallery to show great art from our native schools. This would have obviated the need for the National Gallery to buy George Stubbs's Whistlejacket (by far its biggest recent acquisition - in every sense), because there are better Stubbses in the Tate. This would be an even more beneficial arrangement in the international arena, because it would enable the National Gallery to select masterpieces as their status emerges, while leaving Tate Modern free to experiment and go forward.
Alfred H Barr, the first director of the Museum of Modern Art in New York, identified this problem when he began to build up its collections in the 1930s: how could MoMA, he asked, remain modern if it became filled up with the past? What has now been decided, without any public debate, is that Tate Modern will become our national gallery for post-1900 art. This is a mistake. There is room for both a great gallery showing peaks of achievement in painting, and another showing modern art in all its manifestations which keeps up to date. Tate Modern is designed for showing installations; paintings, particularly older ones, often look miserable and out of place there. The National Gallery could reassert itself as the fountainhead of painting at one stroke by taking back Monet's Waterlilies, which it handed over to Tate Modern (because it was painted after 1900), and with it Picasso's Three Dancers, as fitting partners for its Claudes and Titians.
Art is not made from art (no more than nails are made from nails), but artists always want to pitch themselves against the achievement of their predecessors. Picasso was briefly allowed to hang his paintings beside those of Eugene Delacroix in the Louvre: he wanted to see how they stood up. And in 1856, J M W Turner's estate bequeathed two of his paintings to the National Gallery on condition that they be shown next to pictures by Claude - comparing them is extraordinarily revealing: Claude's sea ripples rhythmically, Turner's is pulled by the moon; Claude's sky is a painted backcloth, Turner's a mountainous wave of light. That is art history where the art remains uppermost - an unfolding story in which the National Gallery can and needs to play a leading part. Temporary exhibitions such as the current one showing R B Kitaj's parody of the gallery's Paul Cezanne - Western Bathers, with provocatively dressed cowboys sporting their weaponry around a "camp" fire - are no substitute, but only serve to show how wide the gap is growing between the National Gallery's collection and the living world of art.
The National Gallery needs to continue to collect great paintings simply because artists have not stopped painting them. This should be the guiding principle for the incoming director, Charles Saumarez Smith, who has proved his commitment to modern art at the National Portrait Gallery. It is not too late for the National Gallery to reclaim its birthright and regain its status as the world's greatest living gallery of art.
Julian Spalding is a former director of Manchester City Art Galleries and of Glasgow Museums. His book The Poetic Museum: reviving historic collections is published by Prestel (£24.95)
