An epicycle is a geometric construct used in classical astronomy in which a planet moves in a small circle (the epicycle) whose center itself travels along a larger circle (the deferent) centered on or near Earth. The model was developed within the framework of Ptolemaic astronomy to account for the observed irregular motions of planets - particularly retrograde motion, in which planets appear to reverse direction against the background of fixed stars. Epicycles were a central feature of geocentric cosmology for roughly fourteen centuries and represent one of the most sophisticated mathematical systems produced in the ancient and medieval worlds.
The formal mathematical treatment of epicycles is attributed primarily to Claudius Ptolemy (c. 100-170 CE), whose Almagest codified the system in its most complete form, though the underlying concepts draw on earlier work by Hipparchus (c. 190-120 BCE) and Apollonius of Perga (c. 262-190 BCE). In Ptolemy's system, each planet required its own combination of deferent, epicycle, and additional refinements - including the equant point - to bring predicted positions into agreement with observation. The model achieved considerable predictive accuracy for its time and served as the basis for astronomical tables used in navigation and calendar-making across the Islamic world and medieval Europe.
The Copernican Revolution of the sixteenth century displaced the geocentric model with a heliocentric one, but early versions of the Copernican system retained epicycles to account for residual discrepancies in planetary motion. Epicycles were not fully eliminated from planetary theory until Johannes Kepler replaced circular orbits with ellipses in the early seventeenth century, demonstrating that the apparent need for epicycles arose from the assumption that planetary orbits must be circular.
A mathematically significant point - noted most prominently by Hanson (1960) and subsequent historians of science - is that any smooth periodic curve can be approximated arbitrarily closely by a sufficiently large sum of circular motions, a property formalized in modern Fourier analysis.[5] This means epicyclic systems are not simply ad hoc errors but reflect a genuine class of mathematical decomposition, even if their physical interpretation was incorrect.
There is broad consensus in the history of science and in physics that the Ptolemaic epicyclic model was superseded by Keplerian orbital mechanics and, subsequently, by Newtonian mechanics and general relativity as descriptions of planetary motion. See Astronomy Consensus - Epicycles for a summary of that consensus. The role of epicycles as a historical and philosophical case study - particularly regarding model underdetermination and the relationship between predictive success and truth - remains a subject of active discussion in philosophy of science.