Apollonius of Perga (c262-c190 BC) - whose life story remains a mystery
as it is not even certain he was born in Perga, and his life dates are rough
estimates - is remembered for his only major work still extant,
*Conics*, an 8-book work (of which the first 7
survive) which summarises the knowledge of the time on the subject, and goes on
to introduce numerous major new ideas. The terms "ellipse", "parabola" and
"hyperbola" to describe conical sections, were coined in this work, and new
definitions of the shapes were found. Until then, they had been defined as
sections, perpendicular to the base, of different types of cone. Apollonius
redefined them all as sections, at different angles, of the same cone. He
credited Conon of Samos (c280-c220 BC), a collaborator of Archimedes of
Syracuse (c287-212
BC), and Euclid of Alexandria (c325-c265 BC) with the original work on conical
sections that inspired this work.

Of his other books, all, with the exception of
*Cutting off a Ratio* (a copy of which was found
in arabic translation in the
late
17th Century), have been lost, and we know their contents only through
the accounts of others. The descriptions show the breadth of the subjects he
tackled. The majority were on the subject of geometry, but he strayed into
optics (in *On the Burning Mirror*) and even
astronomy. Books V-VII of *Conics* had a limited
impact on European science, as they only became available in Europe in
1661, having survived only
in Arabic libraries before then. Book VIII is lost, though the outline of the
content has been reconstructed from the reports in others' books.

*On the Burning Mirror* showed that,
contrary to the prevalent belief of the time, parallel beams of light hitting a
spherical mirror do not converge at one point. He probably investigated
parabolic mirrors too (rays do come to a single focal point with parabolic
mirrors). He is also said by Eutocius (c480-c540 AD) to have extended Euclid's theory of
irrationals and improved Archimedes' approximation of 'pi' - though it is not
clear whether he did this simply by inscribing and circumscribing the circle
with polygons with more sides than Archimedes' 96-gons or whether he did
actually find a new method.

The extent of Apollonius' astronomical work is unknown; the only
novelty he is credited with is a study of epicycles that suggested a way of
predicting the "stationary" point in a planet's orbit. It is believed, however,
that his contribution to astronomy was in all probability considerably
greater.