On the paper was written, in Ponder's very neat handwriting:

THE RULES

1 Things fall apart, but centres hold.

2 Everything moves in curves.

3 You get balls.

4 Big balls tell space to bend.

5 There are no turtles anywhere.

6 ... It's so depressing.

'Always been a bit of a one for rules, our Ponder,' said the Senior Wrangler.

'Number Six doesn't sound incredibly well formulated,' said Ridcully.

'You don't think he's going a bit bursar, do you?' said the Lecturer in Recent Runes.

'He always thinks everything has to mean something,' said Ridcully, who generally took the view that trying to find any deep meaning to events was like trying to find reflections in a mirror: you always succeeded, but you didn't learn anything new.

'I suppose we could simply heat the thing up,' said the Senior Wrangler.

'A sun should be easy,' said Ridcully 'A big ball of fire should be no problem to a thinking wizard.' He cracked his knuckles. 'Get some of the students to put Mister Stibbons to bed. We'll soon have his little world all warm or my name's not Mustrum Ridcully.'

TO THE WIZARDS OF UNSEEN UNIVERSITY, the heavens include two obviously different types of body: stars, which are tiny pinpricks of light, and the sun, which is a hot ball, not too far away, and passes over the Disc during the day and under it at night. It's taken humanity a while to realize that in our universe it's not like that. Our Sun is a star, and like all stars it's huge, so those tiny pinpricks must be a very long way off. Moreover, some of the pinpricks that seem to be stars aren't: they betray themselves by moving differently from the rest. These are the planets, which are a lot closer and a lot smaller, and together with the Earth, Moon, and Sun they form the solar system. Our solar system may look like a lot of balls whizzing around in some kind of cosmic game of pool, but that doesn't mean that it started out as balls or rock and ice. It is the outcome of a physical process, and the ingredients that went into that process are not obliged to resemble the result that comes out. The more we learn about the solar system, the more difficult it is to give a plausible answer to the question: how did it start? It is not the 'answer' part that gets harder, it's the plausibility. As we learn more and more about the solar system, the reality-check that our theories have to pass becomes more and more stringent. This is one reason why scientists have a habit of opening up old questions that everybody assumed were settled long ago, and deciding that they weren't. It doesn't mean that scientists are incompetent: it demon­strates their willingness to contemplate new evidence and re-examine old conclusions in its light. Science certainly does not claim to get things right, but it has a good record of ruling out ways to get things wrong.

What must a theory of the formation of the solar system explain? Principally, of course, the planets, nine of them, dotted rather randomly in space; Mercury, Venus, Earth, Mars, Jupiter, Saturn, Uranus, Neptune, Pluto. It must explain their differences in size. Mercury is a mere 3,032 miles (4,878 km) in diameter, whereas Jupiter is 88,750 miles (142,800 km) in diameter, 29 times as big, 24,000 times the volume, an enormous discrepancy. It must explain their differences in chemical composition: Mercury is made of iron, nickel, and silicate rock; Jupiter is made from hydrogen and helium. It must explain why the planets near the Sun are generally smaller than those further out, with the exception of tiny Pluto, out in the cold and the dark. We don't know a great deal about Pluto, but most of what we do know is strange. For instance, all the other planets lie pretty close to a single plane through the centre of the Sun, but Pluto's orbit is inclined at a noticeable angle. All the other planets have orbits that are pretty close to circles, but Pluto's orbit is much more elongated, to the extent that some of the time it is closer to the Sun than Neptune is.

But that's not all that a theory of the origin of the solar system has to get right. Most planets have smaller bodies in orbit around them, our own familiar Moon; Phobos and Deimos, the diminutive twin satellites of Mars; Jupiter's 16 satellites; Saturn's 17 ... Even Pluto has a satellite, called Charon, and that's weird too. Saturn goes one better and also has entire rings of smaller bodies surrounding it, a broad, thin band of encircling rocks that breaks up into a myriad distinct ringlets, with satellites mixed up among them as well as more conventional satellites elsewhere. Then there are the asteroids, thousands of small bodies, some spherical like planets, others irreg­ular lumps of rock, most of which orbit between Mars and Jupiter, except for quite a few that don't. There are comets, which fall in towards the Sun from the huge 'Oort cloud' way out beyond the orbit of Pluto, a cloud that contains trillions of comets. There is the Kuiper belt, a bit like the asteroid belt but outside Pluto's orbit: we know over 30 bodies out there now, but we suspect there are hun­dreds of thousands. There are meteorites, lumps of rock of various sizes that wander erratically through the whole thing ...

Each of these celestial objects, moreover, is a one-off. Mercury is a blisteringly hot lump of cratered rock. Venus has a sulphuric acid atmosphere, rotates the wrong way compared to nearly every­thing else in the solar system, and is believed to resurface itself every hundred million years or so in a vast, planetwide surge of volcanic activity. Earth has oceans and supports life; since we live on it we find it the most congenial of the planets, but many aliens would probably be aghast at its deadly, poisonous, corrosive oxygen atmos­phere. Mars has rock-strewn deserts and dry ice at its poles. Jupiter is a gas giant, with a core of hydrogen compressed so much that it has become metallic, and maybe a small rocky core inside that, 'small' compared to Jupiter, but about three times the diameter of the Earth. Saturn has its rings, but so do Jupiter, Uranus, and Neptune, though these are nowhere near as extensive or spectacu­lar. Uranus has an icy mantle of methane and ammonia, and its axis of rotation is tilted so far that it is slightly upside down. Neptune is similar to Uranus but without that ridiculous axial tilt. Pluto, as we've said, is just crazy. We don't even know accurately how big it is or how massive it is, but it's a Lilliputian in the country of the Gas Giants.

Right... all that is what a theory of the origins of the solar sys­tem has to explain. It was all a lot easier when we thought there were six planets, plus the Sun and the Moon, and that was it. As for the solar system being an act of special creation by a supernatural being, why would any self-respecting supernatural being make the thing so complicated?

Because it makes itself complicated, that's why. We now think that the solar system was formed as a complete package, starting from quite complicated ingredients. But it us took a while to realize this.

The first theory of planetary formation that makes any kind of sense by modern standards was thought up by the great German philosopher Immanuel Kant about 250 years ago. Kant envisaged it all starting as a vast cloud of matter, big lumps, small lumps, dust, gas, which attracted each other gravitationally and clumped together.