<P> The classical Greek scholar Aristotle (384--322 BC) was first to devote serious attention to the rainbow . According to Raymond L. Lee and Alistair B. Fraser, "Despite its many flaws and its appeal to Pythagorean numerology, Aristotle's qualitative explanation showed an inventiveness and relative consistency that was unmatched for centuries . After Aristotle's death, much rainbow theory consisted of reaction to his work, although not all of this was uncritical ." </P> <P> In Book I of Naturales Quaestiones (c. 65 AD), the Roman philosopher Seneca the Younger discusses various theories of the formation of rainbows extensively, including those of Aristotle . He notices that rainbows appear always opposite to the sun, that they appear in water sprayed by a rower, in the water spat by a fuller on clothes stretched on pegs or by water sprayed through a small hole in a burst pipe . He even speaks of rainbows produced by small rods (virgulae) of glass, anticipating Newton's experiences with prisms . He takes into account two theories: one, that the rainbow is produced by the sun reflecting in each water drop, the other, that it is produced by the sun reflected in a cloud shaped like a concave mirror; he favours the latter . He also discusses other phenomena related to rainbows: the mysterious "virgae" (rods), halos and parhelia . </P> <P> According to Hüseyin Gazi Topdemir, the Arab physicist and polymath Ibn al - Haytham (Alhazen; 965--1039), attempted to provide a scientific explanation for the rainbow phenomenon . In his Maqala fi al - Hala wa Qaws Quzah (On the Rainbow and Halo), al - Haytham "explained the formation of rainbow as an image, which forms at a concave mirror . If the rays of light coming from a farther light source reflect to any point on axis of the concave mirror, they form concentric circles in that point . When it is supposed that the sun as a farther light source, the eye of viewer as a point on the axis of mirror and a cloud as a reflecting surface, then it can be observed the concentric circles are forming on the axis ." He was not able to verify this because his theory that "light from the sun is reflected by a cloud before reaching the eye" did not allow for a possible experimental verification . This explanation was later repeated by Averroes, and, though incorrect, provided the groundwork for the correct explanations later given by Kamāl al - Dīn al - Fārisī (1267--1319) and Theodoric of Freiberg (c. 1250--1310). </P> <P> Ibn al - Haytham's contemporary, the Persian philosopher and polymath Ibn Sīnā (Avicenna; 980--1037), provided an alternative explanation, writing "that the bow is not formed in the dark cloud but rather in the very thin mist lying between the cloud and the sun or observer . The cloud, he thought, serves simply as the background of this thin substance, much as a quicksilver lining is placed upon the rear surface of the glass in a mirror . Ibn Sīnā would change the place not only of the bow, but also of the colour formation, holding the iridescence to be merely a subjective sensation in the eye ." This explanation, however, was also incorrect . Ibn Sīnā's account accepts many of Aristotle's arguments on the rainbow . </P>

The arc shape of rainbow is mainly due to