AAS 197, January 2001
Session 1. HAD I: Boners of the Century
Special Session Oral, Sunday, January 7, 2001, pm, SDSU, Zinner Collection

[Previous] | [Session 1] | [Next]

[1.01] J. B. Biot and Refraction Calculations

A. T. Young (SDSU)

The Auer-Standish (AJ 119, 2472, 2000) algorithm, recommended in the revised Explanatory Supplement for calculating refraction in an arbitrary model atmosphere, was derived and used by J.~B.~Biot (Conn.\ des Tems pour l'An 1839) a century and a half earlier, using Newton's (wrong) emission theory, and the clumsy notation of Laplace's Mécanique Céleste, which Biot had proof-read. Newton, Laplace, and Biot all describe refraction in terms of the trajectories of ``luminous molecules'' attracted by a central force exerted by the atmosphere; this explains why Laplace considered refraction a topic in celestial mechanics. Fortunately for these authors, the only optics required is Snel's law of refraction, which was discovered before Newton's birth, and which Newton's corpuscular optics was rigged to reproduce. Thus Biot's ``derivation'' of the refractive invariant nr \sin z by Laplace's method is a circular and unnecessary argument.

While Auer & Standish were reinventing Biot's method, the historian D.~T.~Whiteside (Centaurus 24, 288, 1980) noticed the mathematical similarity of the refraction theories of Newton and Biot to modern ones, and rashly concluded that ``working astronomers still find computational advantage in maintaining the fiction of a Newtonian emission theory'' --- which is absurd nonsense!

Despite being an emissionist, Biot understood atmospheric refraction much better than most astronomers do today: he knew why refraction is almost independent of atmospheric structure, except within a few degrees of the horizon, and that refraction at the horizon depends mostly on the local temperature gradient. His work --- together with that of Lord Rayleigh, who derived his eponymous scattering law from the elastic-solid theory of the luminiferous æther --- reminds us that a theory's correct results do not make it true.

This work was supported by NSF grant ATM-9714357.

[Previous] | [Session 1] | [Next]