The great inequality
By the eighteenth century, astronomers comparing ancient and modern observations had noticed that Jupiter seemed to be speeding up and Saturn slowing down. Taken at face value, Jupiter was slowly falling towards the Sun and Saturn drifting away. Newton himself had suspected that the solar system might need occasional divine correction.
In 1785 Laplace showed that the drift was not permanent. Two orbits of Saturn take almost exactly as long as five orbits of Jupiter, so the two planets meet at nearly the same places again and again, and their small tugs add up instead of cancelling. The result is a slow swing, with a period of about 900 years, in which one planet runs ahead of its expected position while the other lags. Over a full cycle it averages out.
Is the solar system stable?
The great inequality was one piece of a larger argument. Working with Lagrange, Laplace showed that, to the accuracy of their approximations, the average sizes of planetary orbits do not grow or shrink over time; their shapes and tilts only oscillate. He collected this and much else into the five volumes of the Traité de mécanique céleste (1799–1825).
The answer turned out to be more subtle. Henri Poincaré showed in the 1890s that the series they used do not converge in general, and computer integrations since the 1980s have found that the inner solar system is chaotic over hundreds of millions of years. For any span of time Laplace could have cared about, though, his conclusion holds.
The nebular hypothesis
All the planets orbit the Sun in the same direction and nearly in the same plane. Laplace thought that was too orderly to be chance. In the Exposition du système du monde (1796) he proposed that the solar system formed from a hot, slowly rotating cloud that contracted, spun faster and shed rings of material which then gathered into planets. Immanuel Kant had sketched a similar idea in 1755. The details have changed, but the modern picture of planets forming in a disc around a young star is a direct descendant.
When Napoleon remarked that the book did not mention the Creator, Laplace is said to have replied, “I had no need of that hypothesis.” The story was told years later and may have been improved in the telling.
Dark stars
The first edition of the Exposition contains a remarkable paragraph. Light, Laplace reasoned, is pulled by gravity like anything else. A star of the same density as the Earth but 250 times wider than the Sun would have such a strong pull that its light could not escape: “the largest luminous bodies in the universe may, through this cause, be invisible.” John Michell had made the same argument in England in 1783.
Their reasoning used Newton's theory and a particle picture of light, both later replaced. Yet the critical radius it gives for a given mass, R = 2GM/c², is exactly the Schwarzschild radius of general relativity. Pick a density and a size below and see whether light gets out.
For a uniform ball the escape speed is R·√(8πGρ/3), so it grows in proportion to the radius at fixed density. With Earth's density the critical size comes out at about 245 solar radii, close to Laplace's figure.