[Intermediate → Advanced] · Newton, Einstein, and the objects where the difference stops being academic.
Newton, and where he still works
Newtonian gravity is astonishingly good. It got humans to the Moon, runs every spacecraft trajectory in the Solar System, and predicted Neptune’s existence from Uranus’s orbital deviations before anyone saw it.
It is also, strictly, wrong — and the first place it visibly failed was Mercury.
General relativity
Einstein’s reframing: gravity is not a force. Mass and energy curve spacetime, and objects follow the straightest available paths through that curved geometry.
Wheeler’s summary is the one to keep: spacetime tells matter how to move; matter tells spacetime how to curve.
The predictions, and how each was confirmed:
Mercury’s perihelion precession — an anomalous 43 arcseconds per century that Newton couldn’t explain. GR gave exactly that, with no free parameters. A postdiction, and the more impressive for it.
Light bending — measured during the 1919 eclipse. Made Einstein famous overnight, and is now a working tool: gravitational lensing magnifies distant galaxies and maps invisible mass.
Gravitational time dilation — clocks run slower in stronger gravity. This is not a curiosity: GPS satellites gain ~38 microseconds per day relative to ground clocks, and uncorrected that would accumulate ~10 km of positioning error daily. Every phone with satellite navigation depends on GR being right.
Gravitational redshift — light climbing out of a well loses energy → note 03.
Gravitational waves — predicted 1916, detected 2015 by LIGO from two merging black holes. The signal moved the detector mirrors by a fraction of a proton’s width. It opened an entirely new observational channel.
Black holes
A region where escape velocity exceeds the speed of light. Not a hole and not a vacuum cleaner — a black hole with the Sun’s mass would leave Earth’s orbit unchanged.
The parts:
- Event horizon — the boundary of no return. At the Schwarzschild radius, . Not a physical surface; nothing special happens locally as you cross
- Singularity — where GR predicts infinite density, which is understood as the theory breaking down, not as a physical fact. Resolving it needs quantum gravity, which doesn’t exist yet
- Accretion disc — infalling matter, heated by friction to millions of degrees. This is what we actually observe; the hole itself emits nothing
The types:
- Stellar-mass (~5–100 M☉) — from core collapse → note 04
- Supermassive (10⁶–10¹⁰ M☉) — at galactic centres, including Sagittarius A* at ours. How they grew so large so early is an open problem
- Intermediate-mass — long elusive, now with growing evidence
The evidence is direct. Stars orbiting Sagittarius A* have been tracked for decades, giving a mass of ~4 million M☉ in a region smaller than Mercury’s orbit — that observation won the 2020 Nobel Prize. And the Event Horizon Telescope imaged the shadow of M87* in 2019 and Sgr A* in 2022 → interferometry.
Hawking radiation — quantum effects near the horizon should make black holes evaporate, slowly. A stellar-mass black hole would take ~10⁶⁷ years, so it is entirely unobserved and remains theoretical. It also generates the information paradox, which is unresolved and is one of the sharpest known conflicts between GR and quantum mechanics.
Neutron stars
What’s left when a core collapses but stops short of a black hole. Roughly 1.4–2 M☉ compressed into ~20 km.
The numbers are hard to take seriously and are correct: a teaspoon of neutron-star material would weigh about a billion tonnes. Surface gravity ~10¹¹ times Earth’s. Magnetic fields up to 10¹¹ tesla.
Pulsars are rapidly rotating neutron stars with beamed emission sweeping past us — natural clocks, some stable enough to rival atomic clocks. The first was briefly labelled “LGM-1” for little green men before the rotation explanation was found.
They are also physics laboratories we cannot build. Neutron-star matter is denser than an atomic nucleus, and its equation of state is genuinely unknown — constrained by measuring masses and radii, and now by gravitational waves from mergers.
Where GR is still open
Two genuinely unsolved problems, and both are honest:
Quantum gravity. GR describes the very large; quantum mechanics the very small. They are incompatible where both matter — singularities, the first instant of the universe. String theory and loop quantum gravity are candidates; neither has experimental support.
Dark matter and dark energy → note 07. Galaxies rotate as though they contain far more mass than we see, and the expansion is accelerating. Together these are ~95% of the universe’s content, and we don’t know what either is.
Modified gravity (MOND and successors) proposes changing the law instead of adding matter. It fits galaxy rotation curves remarkably well and fails on galaxy clusters and the CMB, which is why most of the field favours dark matter — while acknowledging that “most of the universe is something we’ve never detected” is not a comfortable position.
Related
- stars — what collapses
- cosmology — GR applied to everything
- control theory · numerical methods — the maths underneath
Source: [reference] — written Aug 2026.
Software engineering
Infrastructure & security
Computer science
Hardware & engineering