Mauritshuis audio guide,
five stops, free.

Count the animals before you decide what this is. There are about a hundred of them, and they come from continents that have never shared a clearing: dromedaries, an elephant, big cats, a peacock with its tail out along the bottom edge, macaws in the top left corner, birds of paradise in the branches.

Two men made the panel, around 1615, and they divided it by subject rather than by half. Peter Paul Rubens did the figures. Jan Brueghel the Elder did everything that grows and everything that breathes.

The order they worked in has been read off the paint, and it is the opposite of what you would guess from the result. Brueghel was responsible for the composition, but Rubens went first: Adam and Eve sketched in thin paint, then the tree, then the horse at the left edge and the serpent in the branches. Brueghel came afterwards and built the rest of the world around them, in finishing paint, on top. Infrared shows his underdrawing underneath it all, done in pencil or silverpoint.

Brueghel could paint a macaw correctly because he had looked at one. Princely menageries kept exotic animals alive, he studied them there, and that is why the birds here still hold up to an eye that knows what a bird of paradise looks like.

The two were not simply business partners. Rubens was godfather to Brueghel's children.

Now the thing that makes the picture work. In a panel crammed with a hundred creatures, exactly one event is taking place: an apple is passing from Eve's hand to Adam's. Everything else is grazing, preening, drinking or asleep. The horse has not noticed. The peacock has not noticed. The whole world is a second away from changing and the only witness paying attention is the snake.

The panel belonged to Pieter de la Court van der Voort of Leiden from 1766, then to William the Fifth, Prince of Orange, and in 1816 it went into the royal picture collection that opened in this house as a museum six years later.

It is smaller than you were expecting. Thirty-three and a half centimetres by twenty-two point eight, a panel you could cover with two hands, holding a wall in a room otherwise full of Jan Steen's noisy households.

Start with the chain. It runs from a ring on the feeder up to the bird's ankle, thin, catching one line of light. Goldfinches were kept as pets in the seventeenth century and trained to perform a trick: hauling their own drinking water up in a thimble-sized bucket on a chain like that one. The Dutch name for the bird, puttertje, is a diminutive meaning draw-water, and the same idea survives in an old Norfolk name for the goldfinch.

Now step back and look at the shadow on the wall behind it. The light is coming from below. That is a clue about where this panel used to live: the perspective suggests it was meant to be mounted high, above eye level, and one scholar has proposed that it was fixed to the inner jamb of a window, or used as a hinged cover over something else. Whatever it was, it was not made for a wall at chest height. It was made to be looked up at.

Carel Fabritius signed it along the bottom, C fabritivs 1654, and that date is why this small panel carries more weight than its size. On the twelfth of October 1654 about thirty tonnes of gunpowder stored in a former convent in Delft went up. The blast flattened the middle of the city, left a deep crater, killed at least a hundred people, and was heard on the island of Texel, a hundred and ten kilometres away. It took Fabritius's studio, his pupil Mattias Spoors, the deacon Simon Decker, and Fabritius himself, who was 32. Few of his works are known to have survived.

Then the bird disappeared for two centuries. The critic Théophile Thoré-Bürger found it in Brussels in 1859, in the collection of a retired Dutch army officer, and Abraham Bredius bought it for this museum at auction on the twenty-seventh of February 1896, for 6,200 francs.

One more thing, from the restoration of 2003, which had to use a CT scanner because the lead-based paint defeated X-rays and infrared. This panel was once part of a larger piece of wood. The remains of a wooden pin are still in it.

She is small. Forty-four and a half centimetres by thirty-nine, and the crowd in this room is usually bigger than the painting.

She is also not a portrait. She is a tronie, the Dutch word for a painted head that was never meant to be a likeness of anyone in particular: a face, a costume, an expression, sold as a thing in itself. So the question everybody asks in this room, who was she, is one the picture was built not to answer.

Look at the background first, because it is the part that has changed most. It reads as flat brown-black. It was not. When she was restored in 1994, conservators found that it had originally been a deep, enamel-like green, laid on as a glaze whose two organic pigments, indigo and weld, have since faded out of it. Stand close: the surface is mottled and uneven rather than dead flat, and that is what is left of the green.

In 2018 the museum built a glass workshop in these galleries and studied her in it for two weeks, in front of visitors. The project was called The Girl in the Spotlight, the conservator Abbie Vandivere led it, and the results came out in April 2020.

Two of the findings change what you are looking at. She has eyelashes: tiny painted hairs around both eyes, invisible at this distance, found by macro X-ray fluorescence scanning. And behind her head there was once a green curtain. Because she had no eyebrows and stood against nothing, people had decided Vermeer was painting an idea of a face rather than a face. He was painting a real person in a real space.

The blue on her head cost a fortune: natural ultramarine, ground from lapis lazuli brought from what is now Afghanistan, worth more than gold in the seventeenth century. The recent work suggests the stone was heated first, which made it easier to grind and pushed the blue further.

Now the pearl, which is the reason you know this painting, and which may not be there at all. The researchers describe it as an illusion, on the grounds that it has no contour and also no hook to hang it from the girl's ear. In 2014 the astrophysicist Vincent Icke went further and argued that it is not a pearl at all but something closer to polished tin, on the evidence of the specular reflection, the pear shape and the sheer size of it.

In 1881 Arnoldus Andries des Tombe bought her at an auction here in The Hague for two guilders, plus thirty cents of buyer's premium, roughly twenty-four euros in 2015 money. The paint was lifting off the canvas. He had no heirs, and in 1902 he left her to this museum.

Under the German occupation she spent two years in a bombproof bunker directly beneath this house, then went to the caves near Maastricht until the liberation in May 1945. And in 2023, when she travelled to Amsterdam for the Rijksmuseum's Vermeer exhibition, the museum asked the public for stand-ins to hang in the gap. Three thousand four hundred and eighty-two people sent one, aged three to ninety-four, and five were printed and hung. One of the five had been generated with an artificial-intelligence tool, and the choice caused a public row. The Dutch artist Iris Compiet called it an insult to the legacy of Vermeer and to any working artist. The museum's Boris de Munnick answered that they were not the museum to discuss whether artificial intelligence belongs in an art museum.