For nearly two thousand years, educated Europeans and scholars across the Islamic world inherited a cosmos designed by Aristotle in the fourth century BCE. In that cosmos, Earth sat motionless at the center, and every celestial object was carried around it on a nested, transparent sphere. Aristotle’s argument was not arbitrary. He reasoned that the heavens were composed of an unchanging fifth element and that the only motion appropriate to a perfect, eternal region was uniform motion in a perfect circle. He also offered an empirical defense: if Earth were racing through space, he argued, the nearby stars should appear to shift against the more distant ones over the course of a year. No such shift — no stellar parallax — could be detected, so Earth must be standing still.

The argument was logically sound. Its hidden flaw was a badly underestimated scale: the stars are so remote that their parallax is far too small to measure without a telescope, and it was not successfully observed until 1838.

The model faced a more immediate embarrassment. Most nights, a planet such as Mars drifts slowly eastward against the background stars. But for several weeks roughly every twenty-six months, Mars reverses course, loops westward, and then resumes its eastward march. This backward drift is called retrograde motion, and a single uniform circle centered on Earth cannot produce it.

Around 150 CE, the Alexandrian astronomer Claudius Ptolemy engineered a solution in a work later known as the Almagest. Rather than abandon circles, he multiplied them. Each planet rode on a small circle, an epicycle, whose center itself traveled along a larger circle called the deferent. When the planet swung along the inner portion of its epicycle, its apparent motion reversed. Ptolemy added further adjustments — an offset center, and a point called the equant about which motion appeared uniform — until his tables reproduced the sky to within a degree or two. That accuracy was good enough to survive for fourteen centuries. It was also, as later critics complained, purchased at the price of a machine that predicted planetary positions without ever explaining why the planets should behave that way.

Nicolaus Copernicus, a Polish cleric and astronomer, spent decades circulating a radical proposal before allowing it into print in 1543, the year of his death. In De revolutionibus orbium coelestium, he moved the Sun to the center and set Earth in orbit as the third planet. The payoff was immediate and elegant: retrograde motion ceased to be a real reversal and became a matter of perspective. Earth, on a shorter and faster orbit, periodically overtakes an outer planet the way one car passes another on a highway, and the slower planet appears to slide backward. Copernicus could also derive, for the first time, the relative distance of each planet from the Sun rather than simply assuming an order.

Yet the heliocentric model of 1543 was not more accurate than Ptolemy’s. Copernicus kept the ancient commitment to uniform circular motion, so he still needed epicycles to force his predictions to match the sky. Critics raised a familiar objection as well: if Earth truly moved, where was the stellar parallax?

The answer required better data, and better data required a better instrument-maker. Tycho Brahe, a Danish nobleman, received the island of Hven from the king and built Uraniborg, an observatory-palace stocked with enormous quadrants, sextants, and armillary spheres of his own design. Working entirely with the unaided eye — the telescope was still decades away — Tycho pushed positional accuracy to roughly one to two arc minutes, several times sharper than any predecessor. He also documented two events that damaged Aristotle’s unchanging heavens: a brilliant new star in 1572 that showed no measurable parallax, and a comet in 1577 that clearly traveled beyond the Moon and through the supposedly solid crystalline spheres.

Tycho was not, however, a Copernican. Unable to detect parallax himself, he proposed a compromise: the planets circle the Sun, while the Sun and Moon circle a stationary Earth. His hybrid model failed. His observations, particularly the twenty-year record of Mars, did not. In 1600 he hired a brilliant, argumentative German mathematician named Johannes Kepler, and when Tycho died the following year, that record passed into Kepler’s hands.

In 1609, hearing of a Dutch spyglass, Galileo Galilei built his own improved version and did something no one had thought worth doing: he aimed it upward. Within months he had assembled a body of observational evidence that no version of Aristotle’s cosmos could absorb. He published the first installment in 1610 in a slim, urgent pamphlet titled Sidereus Nuncius, or Starry Messenger.

The Moon, magnified, was not a polished sphere of celestial material. It carried mountains, valleys, and craters, and Galileo estimated the height of lunar peaks from the length of their shadows. The Milky Way resolved into countless individual stars. Around Jupiter, four points of light shifted position from night to night, always staying near the planet. Galileo recognized them as moons in orbit — and their existence destroyed a favorite objection to Copernicus, namely that Earth could not possibly drag the Moon along while it circled the Sun. Jupiter was doing exactly that, four times over.

His most decisive observation came late in 1610. Venus, watched over many months, displayed a complete cycle of phases: a thin crescent when it appeared largest, a half disk, a fat gibbous shape, and finally a small, fully lit disk. In Ptolemy’s arrangement, Venus rides an epicycle positioned permanently between Earth and the Sun, so its sunlit hemisphere always faces mostly away from us. Ptolemaic Venus can be new and it can be crescent, but it can never be gibbous or full. Only a Venus that travels all the way around the Sun — passing behind it as well as in front of it — can show the full sequence, and the apparent size must shrink dramatically when Venus is full and far away.

This was falsification in the strict sense. Galileo had not proven that Earth moves; the observation is equally consistent with Tycho’s hybrid model. But he had demonstrated that the pure Ptolemaic system was wrong, and he had shown that a telescope could settle an argument that logic alone had left open for eighteen centuries.

Johannes Kepler inherited Tycho Brahe’s Mars observations in 1601 and expected to finish the orbit in eight days. It took him closer to eight years. He tried circle after circle, including offset centers and equants, and eventually produced a model that matched Tycho’s Mars data to within about eight arc minutes. By every earlier standard that was a triumph. Kepler refused it. He knew Tycho’s instruments did not make eight-arc-minute mistakes, and he wrote that those eight minutes alone had pointed the road to a reformed astronomy.

So he surrendered the assumption that had governed the subject since Aristotle. In Astronomia Nova (1609) Kepler announced that Mars travels on an ellipse, with the Sun at one of the two foci rather than at the center. The degree of stretch is measured by eccentricity, the distance between the foci divided by the length of the major axis; a circle has an eccentricity of zero. Mars, at 0.093, is only slightly out of round — which is precisely why the error had stayed small enough to hide behind pre-Tycho data.

The same book contained a second discovery. A planet does not move at a constant speed. It accelerates as it approaches perihelion, its closest point to the Sun, and slows toward aphelion. Kepler captured this with a geometric statement of startling economy: an imaginary line joining a planet to the Sun sweeps out equal areas in equal intervals of time.

A decade later, in Harmonices Mundi (1619), he published the relationship he had hunted for most of his life — a rule connecting one planet to another. The square of a planet’s orbital period, measured in Earth years, equals the cube of its mean distance from the Sun, measured in astronomical units: T² = a³. Kepler could describe planetary motion with three compact laws, but he could not say why they held. That explanation arrived in 1687, when Isaac Newton derived all three from a universal inverse-square law of gravitation, converting Kepler’s empirical rules into consequences of a single physical cause.