Module 10: From smooth to lumpy · Lesson 10.2 · about 16 min
Gravity makes lumps: the Jeans scale
Why big lumps collapse and small ones ring like sound: the Jeans length, why dark matter clumps first, how lumps grow (δ ∝ a), and why dark energy stalls them.
Module 10 · about 16 minutes. Read first: 10.1 The seeds of structure. Gravity pulls a dense lump inwards; pressure pushes back. Which wins depends on the lump’s size. By the end you will be able to work out the dividing size, the Jeans length, explain why dark matter clumped first, and say how fast lumps grow in an expanding universe and why that growth is now stalling.
1. A tug of war
Squeeze a patch of gas a little and two things happen. Its extra mass pulls inwards a little harder (gravity), and its pressure rises and pushes back outwards. Pressure acts at the speed of sound: a push crosses a lump of size λ in a time λ/cs. Gravity acts on the free-fall time, about 1/√(Gρ), which does not depend on size at all. So for small lumps the sound wave gets across first and the lump bounces back; for big lumps gravity wins before the pressure can respond. In 1902 James Jeans worked out the balance point for a gas cloud in space:
- λJ
- the Jeans length: lumps bigger than this collapse, smaller ones oscillate
- cs
- the speed of sound, which rises with temperature (cs = √(γkT/μmp) for a gas)
- ρ, G
- the density doing the pulling, and Newton's constant
- MJ
- the Jeans mass: the mass inside a ball one Jeans length across. Books differ by factors of a few in how they define it; the size it gives is what matters
- Physical meaning
- Hotter gas (faster sound) means only bigger lumps can collapse; denser gas means gravity is quicker, so smaller lumps can. It is a race between a sound wave and a fall.
2. Before and after the light let go
Until recombination, the gas was locked to light, and the light’s pressure made the sound speed enormous: about c/√3, more than half the speed of light. The Jeans length was then about the size of the horizon, and the Jeans mass of the gas was larger than the biggest galaxy clusters. Lumps of gas could not grow; they rang as the sound waves seen in the CMB (Lesson 7.5).
When the light let go, the gas was left with its own feeble pressure. At about 3,000 K its sound speed was only about 6 km/s, a drop of more than 20,000 times. The Jeans mass crashed to a few hundred thousand suns, about the mass of a globular star cluster. From then on, almost any lump of gas could fall into the dark matter’s wells.
Cold dark matter never had this problem. Its particles do not feel light and move slowly, so it has essentially no pressure and no Jeans length. Its only limit is how far its particles drifted early on before they slowed down; for the most popular candidates that is so short that the smallest dark matter clumps could be as light as the Earth. That is why dark matter’s lumps had a head start of hundreds of thousands of years (Lesson 8.4), and why hot dark matter, whose particles stream out of small lumps, was ruled out (Lesson 8.5).
3. Growing against the expansion
In a static cloud, a lump bigger than λJ grows exponentially fast. In the expanding universe, the stretching of space drags against the collapse, and the growth is far gentler. For a lump much bigger than the Jeans length, the density contrast obeys
- δ̈, δ̇
- how fast the lumpiness speeds up, and how fast it changes
- 2Hδ̇
- "Hubble drag": the expansion pulls the lump's matter apart as it gathers, slowing growth. H is the expansion rate (Lesson 4.2)
- 4πGρ̄δ
- gravity: the extra pull of the lump's extra mass
- Physical meaning
- When matter rules the expansion, the two balance so that δ grows exactly in step with the scale factor: δ ∝ a ∝ t2/3. Double the size of the universe, double the lumpiness. When radiation rules (before about 50,000 years), the expansion is too fast and dark matter lumps grow only very slowly (the Mészáros effect). When dark energy rules, growth stalls again.
From recombination (a ≈ 1/1100) to today, a lump can grow at most about 1,100 times in this way, which is why 1 part in 100,000 in the gas was not enough and dark matter’s head start was needed.
4. Dark energy puts on the brakes
Once dark energy outweighed matter, about 3.5 billion years ago (the expansion had begun speeding up a little earlier, about 6 billion years ago), the Hubble drag won. Lumps that had not already collapsed stopped growing much. Today, lumps are only about 78% as big as they would be in a universe of matter alone. This growth suppression is a separate test of dark energy: it shows up in how many galaxy clusters there are at each redshift, and in how fast matter falls into structures (Lesson 10.8).
- 1902 James Jeans finds the critical size for collapse in a gas cloud, for the formation of stars and nebulae.
- 1946 Evgeny Lifshitz works out growth in an expanding universe with general relativity: slow, as a power of time.
- 1974 Peter Mészáros shows dark matter lumps hardly grow while radiation rules.
- 1980 Jim Peebles' book The Large-Scale Structure of the Universe sets out the whole theory of growth.
- 1992 Sean Carroll, William Press and Edwin Turner give the handy formula for growth with a cosmological constant used in the problem above.
“Dark matter clumps first because it is heavier.” Weight has nothing to do with it: everything falls the same way in gravity (Lesson 3.4). Dark matter clumped first because nothing pushed back. Ordinary matter was held up by the pressure of light until recombination.
- Lumps bigger than the Jeans length λJ = cs√(π/Gρ) collapse; smaller ones oscillate as sound waves.
- Before recombination, light gave the gas a Jeans mass bigger than galaxy clusters; afterwards it fell to about 10⁵ suns.
- Cold dark matter has no pressure and no Jeans length, so it clumps on all scales, from the start of the matter era.
- In an expanding universe, growth is slow: δ ∝ a while matter rules, stalled while radiation rules.
- Dark energy stalls growth again: today lumps are about 78% of their matter-only size.
Check yourself
Sources
- J. H. Jeans (1902), The stability of a spherical nebula, Philosophical Transactions of the Royal Society A 199, 1 https://doi.org/10.1098/rsta.1902.0012
- S. M. Carroll, W. H. Press and E. L. Turner (1992), The cosmological constant, Annual Review of Astronomy and Astrophysics 30, 499 https://doi.org/10.1146/annurev.aa.30.090192.002435
- P. Mészáros (1974), The behaviour of point masses in an expanding cosmological substratum, A&A 37, 225 https://ui.adsabs.harvard.edu/abs/1974A&A....37..225M
- P. J. E. Peebles (1980), The Large-Scale Structure of the Universe, Princeton University Press
Your ticks stay in this browser. Save them as a file on the course home.
Updated 2026-11-22. Numbers are rounded for teaching; measured values come from the sources listed in each lesson. Simulations run in your browser; nothing you type is sent anywhere.