It appears that from the times of Eratosthenes, in the 3rd century BC, a representation of the celestial sphere was widely used by natural philosophers of the Alexandrian school, consisting of graduated metallic rings representing the equator, the ecliptic and certain meridians and parallels. This "skeleton" of metal rings (armillae) was supported by a fixed ring representing the horizon of the observer and could be adapted to the latitude and longitude of the place.
At the centre of this armillary sphere the Earth was represented together with, in the more complicated models, the Sun, Moon and planets. By correctly placing bodies inside the sphere and the rings, it was possible to solve problems of spherical astronomy and compute the coordinates of the stars on the celestial sphere. Obviously, the lack of precision dictated by the dimensions of the instrument favoured its use for more didactic and explicative purposes.
It was Egnazio Danti, in his Dell’uso et fabbrica dell’astrolabio et del planisfero of 1578, who gave a complete description of what he referred to as the "Armillary astrolabe of Ptolemy".
With the development of the heliocentric system, armillary spheres began to be built with the Sun in the centre, providing a complete description of the Copernican system.
Armillary spheres ranged in size from the three metres of the biggest exemplar ever built - made in wood engraved and gilded by Antonio Santucci delle Pomarance at the end of the 1500s and on display at the Istituto e Museo di Storia della Scienza in Florence - to the 15 cm diameter sphere built in 1720 by the Londoner John Rowley .
The finest exemplars date from the end of the 1600s when they were built in brass gilded with splendid engravings of zodiacal signs and important positions on the different rings. Their style and elegance meant they became more objects of fashion than of study. At the end of the XVIIIth century, especially in France, armillary spheres were mass-produced, in wood covered with colored paper.