Cicero (106-43 B.C.) relates that after the conquest of Syracuse in 212 B.C., Roman consul Marcellus brought to Rome a celestial globe and a planetarium built by Archimedes (287-212 B.C.). The planetarium was an extraordinary object that, at each rotation, showed the Moon rise after the Sun above the immobile Earth, the eclipses of Moon and Sun at proper time intervals, as well as the motions of the other five known planets: Mercury, Venus, Mars, Jupiter and Saturn (On the Republic , I, 14, 21-22; Tusculan Disputations , I, 63). This planetarium is also mentioned by Ovid (I century B.C.) in The Calendar (VI, 263-283), by Lactantius (IV century A.D.) in his Divine Institutions (II, 5, 18) and in an epigram by Claudian (IV century A.D.) entitled Archimedes' Sphere . Claudian adds that the instrument was enclosed in a star-studded glass sphere.

Unfortunately, there is no detailed description of the mechanisms that animated Archimedes' planetarium. In 1974, science historian Derek J. De Solla Price presumed that the instrument functioned with gear trains similar to the ones present in the Antikythera mechanism, dating to the I century B.C. In 1975, astronomy historian Otto E. Neugebauer concluded, on the contrary, that no mechanism of the kind could combine the diurnal motion of the asters around the celestial poles and the motions of the Sun and Moon and of the other planets along the ecliptic, as seems to emerge from Cicero 's words.

The planetarium exhibited here does not intend to advance a new hypothesis on the real aspect of the instrument by Archimedes, but instead and more modestly, to point out the complexity of a mechanism capable of simulating, inside a star-studded glass sphere: 1) the diurnal motion of the asters around an immobile Earth, 2) the monthly and annual motions of the Moon and Sun along the ecliptic, 3) the eclipses of the Sun and Moon at proper time intervals.