On this page
You can predict where Earth will be in its orbit next year with great accuracy, because Earth and the Sun together behave almost like a two-body system. Add a third body with a comparable pull, and things change. There is no general formula for three bodies under gravity. The mathematician whose work on this problem uncovered chaos was Henri Poincare. Here is his story, including the mistake that made it famous, a hands-on example of chaos, and what a student can learn from it.
Who was Henri Poincare?
The MacTutor biography of Poincare, from the University of St Andrews, says he was born in Nancy on 29 April 1854 and died on 17 July 1912. His mathematics teacher at school described him as a "monster of mathematics". He worked across pure and applied mathematics and physics, from optics and thermodynamics to celestial mechanics, and is acknowledged as a co-discoverer, with Einstein and Lorentz, of special relativity. He was elected to the Academie des Sciences in 1887 and became its president in 1906.
What is the three-body problem?
For two bodies, Newton's law of gravitation gives orbits that are ellipses, parabolas or hyperbolas, the result behind Kepler's laws of planetary motion. You can write the position at any future time with a formula.
For three bodies, each pulls on the other two, and the pulls change as they all move. Special solutions exist, such as the points described in our post on Lagrange points. But a general solution in neat formulas, valid for any starting positions and speeds, has never been found.
What was King Oscar's prize?
MacTutor explains that Oscar II, King of Sweden and Norway, launched a mathematics competition in 1887 to celebrate his sixtieth birthday in 1889. Poincare won it with a memoir on the three-body problem. In it, MacTutor says, he gave the first description of homoclinic points, the first mathematical description of chaotic motion, and the first major use of invariant integrals.
What was the mistake?
When the memoir was about to be published in the journal Acta Mathematica, Phragmen, the mathematician editing it, found an error. Poincare realised he had indeed made a mistake, and Gosta Mittag-Leffler, who had been corresponding with him about the competition, worked hard to stop the incorrect version from being published. Between March 1887 and July 1890, Poincare and Mittag-Leffler exchanged fifty letters, most about the competition and, later, the error. Historian June Barrow-Green told the full story in a 1994 paper, Oscar II's prize competition and the error in Poincare's memoir on the three body problem.
The corrected memoir appeared in 1890 and fills much of volume 13 of Acta Mathematica, divided into chapters on periodic solutions, invariant integrals and asymptotic surfaces. MacTutor notes that the error "is now regarded as marking the birth of chaos theory".
Worked example: watching a small difference grow
The three-body equations are too heavy for a page, but a simple rule shows the same behaviour. It is called the doubling map: multiply your number by 2, and if the result is 1 or more, subtract 1 (keep only the part after the decimal point).
Start two runs that differ by only 0.001: x = 0.300 and y = 0.301.
Step 1: 0.600 and 0.602. Difference 0.002.
Step 2: 1.200 → 0.200 and 1.204 → 0.204. Difference 0.004.
Step 3: 0.400 and 0.408. Difference 0.008.
Step 4: 0.800 and 0.816. Difference 0.016.
Step 5: 0.600 and 0.632. Difference 0.032.
Step 6: 0.200 and 0.264. Difference 0.064.
Step 7: 0.400 and 0.528. Difference 0.128.
Step 8: 0.800 and 1.056 → 0.056.
The difference doubles every step, from 0.001 to 0.128 in seven steps (0.001 × 2⁷ = 0.128). At step 8 the two runs are in completely different places. If you had measured the start to only two decimal places, you would have lost track within a handful of steps. This sensitive dependence on initial conditions is exactly what Poincare met in the three-body problem, and what Edward Lorenz later found in a weather model in Deterministic Nonperiodic Flow (1963).
What mistakes do students make about chaos?
Thinking chaos means "no rules". The doubling map has a perfectly clear rule.
Thinking more computing power fixes it. Errors grow exponentially, so halving your starting error buys you only one extra step in the example above.
Assuming every system is chaotic. The two-body problem is not. Kepler orbits stay predictable.
Where does this sit in your syllabus?
The JEE (Main) 2026 syllabus lists the law of gravitation, Kepler's laws and satellite motion in Unit 6, all two-body ideas. The three-body problem and chaos are not on it; treat them as background that explains why your textbook stops at two bodies.
Gravity with more than two bodies is much easier to grasp when you can watch orbits change. The Fermi Simulation Lab has interactive simulations for ideas that are easier to see than read.
What can a student take from how Poincare worked?
Admit the error and dig into it. The corrected memoir is the one now linked to the birth of chaos theory.
Change the question. When exact formulas were out of reach, he studied the behaviour of orbits, such as periodic solutions, instead.
Have your work checked. An editor reading carefully is what caught the error. Swap solutions with a friend before an exam.
Next step: run the doubling map yourself starting from 0.3 and 0.3001, and count how many steps it takes for the difference to pass 0.1.
