The Greek Mathematician Who Built the Universe Out of Twenty-Seven Nested Rotating Spheres

The Eudoxean model of homocentric spheres is the first mathematical model of the solar system in the history of human astronomy. It was proposed by Eudoxus of Cnidus around 365 BC, refined by his student Callippus of Cyzicus about 35 years later, and adopted (with further refinements) by Aristotle in approximately 350 BC as the canonical astronomical framework of the Lyceum.

The model proposed that the apparent celestial motions of the Sun, the Moon, the five known planets (Mercury, Venus, Mars, Jupiter, Saturn), and the fixed stars could be explained by a system of concentric (homocentric) spheres, all centred on the Earth, each rotating at its own constant angular velocity around its own axis of rotation. The Eudoxean Earth was stationary; the Eudoxean celestial bodies were each carried by their own dedicated nested sphere-system through their observed apparent motion across the celestial sphere.

The original Eudoxean system used 27 spheres to account for all observed celestial motion. The breakdown was:

The fixed stars rode on a single outermost sphere that rotated once a day around the celestial polar axis. (1 sphere)

The Sun rode on a system of three nested spheres: the outermost provided the daily east-to-west motion (shared with the fixed stars), the middle provided the annual east-to-west motion along the ecliptic, the innermost handled small observed variations from a perfect annual cycle. (3 spheres)

The Moon rode on a system of three nested spheres providing the daily motion, the monthly motion through the zodiac, and the lunar latitude variations from the ecliptic plane. (3 spheres)

Each of the five planets rode on a system of four nested spheres: the daily motion, the longitudinal motion through the zodiac, and a pair of mutually-coupled additional spheres that together produced the observed retrograde-motion loops (the periodic apparent backward motion of the planets that any geocentric astronomical model has to explain). (4 spheres × 5 planets = 20 spheres)

The total: 1 + 3 + 3 + 20 = 27 spheres.

What the model accounted for

The Eudoxean model was a substantive intellectual achievement. It was the first proposed mechanism to explain the systematic apparent motions of the planets without invoking ad hoc independent variation; everything in the model was reducible to constant-rate uniform circular motion of physical bodies (the spheres). It accommodated the major qualitative features of observed planetary motion: the daily east-to-west sweep of the entire celestial sphere; the slower annual east-to-west drift of the Sun, Moon, and planets through the zodiacal background; the retrograde-motion loops (when Mars, Jupiter, or Saturn appear to reverse direction against the stellar background for several weeks before resuming their normal eastward motion).

The retrograde-motion accommodation was the model’s most-elegant feature. The Eudoxean treatment used the geometric construction known as the hippopede (literally “horse-fetter” — the figure-eight pattern produced when two perpendicular circular motions of equal period are superimposed). Each planet’s pair of retrograde-loop spheres was geometrically arranged so that their combined motion produced exactly the figure-eight loop that the observed planet traced against the stellar background during its retrograde period. The hippopede construction is one of the elegant geometric results of the entire Greek mathematical tradition and a substantive demonstration of the explanatory power of pure circular-motion mechanics.

What the model could not account for

The Eudoxean model had three principal observational limitations.

It could not handle the observed apparent-magnitude variation of the planets. Each planet, on the Eudoxean model, was at a fixed distance from the Earth (the radius of its innermost sphere). Mars, Venus, and Mercury are observably brighter at certain points in their orbital cycles than at others — most strikingly, Mars at opposition is approximately five times as bright as Mars at conjunction. The Eudoxean fixed-distance assumption could not explain this. The Hipparchian and Ptolemaic later modifications would address the apparent-magnitude problem by introducing epicycles and equants — geometric devices that allowed the planet’s effective distance from Earth to vary cyclically without abandoning the underlying uniform-circular-motion principle.

It could not handle the substantially-precise observed planetary positions that became available with the Hipparchian generation of careful observers. The Eudoxean model was qualitatively correct but quantitatively imprecise; its predicted planetary positions could be off by several degrees on a year-long forward extrapolation. The Hipparchian observational precision of approximately 0.5 degrees was sufficient to reveal that the Eudoxean predictions were substantively wrong in detail.

It could not handle the observed variations in the apparent angular sizes of the Sun and Moon. The Moon’s apparent diameter varies by about 14% across its monthly cycle (the difference between perigee and apogee); the Sun’s apparent diameter varies by about 3% across the annual cycle. Both variations require a varying Earth-to-body distance and were therefore incompatible with the Eudoxean fixed-distance assumption.

Callippus and Aristotle

The model was first refined by Eudoxus’s student Callippus of Cyzicus in approximately 330 BC. Callippus added seven additional spheres (two each for the Sun and Moon to handle the apparent-magnitude variations and the seasonal-inequality observation; one for Mercury, Venus, and Mars each to refine the retrograde-loop geometry). The Callippic system used 34 spheres total.

The model was then taken over by Aristotle in approximately 350 BC for inclusion in the Aristotelian cosmology of On the Heavens and Metaphysics Book XII. Aristotle made a substantive ontological commitment that Eudoxus had explicitly avoided: the spheres were not just mathematical constructs but physical entities, made of a fifth element (the aether) different from the four terrestrial elements, with their motion driven by a hierarchy of unmoved divine movers operating from the outermost sphere inward. To make the physical model coherent — each sphere had to be physically driven by the sphere immediately outside it, so each subordinate sphere needed to be carrying not only the celestial body but also “counter-rotation” spheres to cancel the residual motion of the outer spheres — Aristotle added 21 additional counter-rotation spheres. The Aristotelian system used 55 spheres.

The Aristotelian 55-sphere model became the canonical Greek-Roman cosmological framework through the Hellenistic and early Roman periods. The Hipparchian and Ptolemaic later astronomical traditions abandoned the homocentric-sphere model in favour of the more-flexible epicycle-deferent geometric framework — but the Aristotelian counter-rotation-sphere cosmology persisted as the philosophical-theological underpinning of medieval Islamic and European cosmology through approximately 1600 AD, until the Galilean discovery of the moons of Jupiter (1610) and the Brahean astronomical observations of the late 16th century progressively demolished the underlying assumption of perfect uniform circular motion.

Why this matters

The Eudoxean homocentric model is the founding example of a mathematical model in the history of human science. The model takes a complicated set of observational phenomena (the apparent motions of the celestial bodies) and shows that they can be accounted for by a small number of simple underlying assumptions (concentric spheres, uniform rotation, geometric combination) operating in a precise geometric framework. The model is the first time anyone had ever applied the principle “if the underlying mechanism is simple enough, the complex observed phenomena will follow from it as theorems” to a domain of empirical phenomena.

The principle is now the standard method of all of physical science. The Eudoxean 27 spheres are the ancestors of the Keplerian elliptical orbits, the Newtonian gravitational equations, the Einsteinian geodesics, and the modern quantum-mechanical wave equations. Each successive framework has replaced the previous one with a substantively better mathematical model — but the underlying methodology, of building physical reality out of a small set of geometrically-precise mathematical assumptions and then deriving the observed phenomena as consequences, is substantively Eudoxean.

The model itself was eventually superseded. The Aristarchian heliocentric proposal (c. 270 BC) had argued for a substantively different cosmological framework; the Hipparchian and Ptolemaic geocentric-epicyclic frameworks (c. 150 BC and 150 AD) substantively replaced the Eudoxean homocentric framework as the working Greek astronomical model; the Copernican rediscovery of 1543 substantively restored the Aristarchian heliocentric position and substantively eliminated the Eudoxean spheres from the astronomical mainstream.

The hippopede figure-eight construction is the surviving legacy. Modern undergraduate geometry textbooks still teach the hippopede as the elegant example of a curve constructed by the superposition of two perpendicular circular motions of equal frequency. The 2,400-year-old Eudoxean construction is the founding case.