That the earth is a sphere was already known to the Greeks by the time of Plato; and most of them were agreed that the sky too is spherical, as suggested by the daily wheeling motion systematically shared by all but seven of the stars. So far so good, but with the seven exceptions, Sun, Moon, Mercury, Venus, Mars, Jupiter and Saturn, the attempt to understand the structure of the universe ran into difficulties. These seven circled the earth every day just like the ‘fixed’ stars, but in addition they moved against the background of the fixed stars without, it seemed, rhyme or reason. However, as the lunar month and the seasonal years showed, the apparently random movements of the ‘wanderers’ or ‘planets’ concealed regularities that were not always evident at first sight, and this raised the question of ‘saving the appearances’ of the planets, of explaining their movements by displaying the geometrical patterns underlying them.

This was the problem that dominated astronomy in the two thousand years that separated Plato from Kepler: a highly technical challenge, of such complexity that in the thirteenth century, Alphonso X of Castile, exasperated by the difficulties of preparing the Alphonsine planetary tables, said he wished God had consulted him when creating the world. Yet, despite its technicality, the problem of the wandering stars is one of the most significant issues in the whole history of science; for when in the sixteenth century Copernicus proposed a solution according to which the earth was itself merely one of several planets which circle round the sun, his specialist monograph sparked off one of the most profound revolutions ever to occur in human thought.