Author: Sean F. Johnston

Title: An unconvincing transformation? Michelson's interferential spectroscopy

Abstract: Albert A. Michelson's interferential spectroscopy, developed during the 1890s, was a technique that fitted jarringly with the established methods of its period. Its apparatus was unfamiliar to conventional spectroscopists, relying on an interferometer to produce 'fringes of interference', rather than on a dispersive instrument to produce a spectrum. Its observations were disturbingly indirect, based on visual estimates of 'visibility' (fringe contrast) rather than on precise angular measurements of refraction or diffraction. Its foundations were equally uncertain: Michelson articulated the relationship between the measured visibility curves and the optical spectrum inadequately for his contemporaries.

Interferential spectroscopy required a fundamental change in practice, as well. Its equipment was simultaneously simple and sophisticated. Knowledge of glass dispersion, slit design and photographic techniques were irrelevant. Taking their place were skills in screw design, vibration damping and visual estimation of brightness. The subject of research was both narrowed and deepened: a narrow-band technique, it was restricted to exquisitely precise measurements of near-monochromatic light sources

With few collaborators or indeed followers, Michelson shaped this technique to satisfy objectives distinct from those of most practising spectroscopists. Interferential spectroscopy served as an esoteric example of the general technique of interferometry. It was one of many applications that he investigated, including the detection of aether drift, the angular separation of double stars, and the verification of standards of length. Michelson focused on the implicit power of such methods to detect minute differences. In the case of interferential spectroscopy, this centred on distinguishing close spectral components.

This paper will explore the extent to which interferential spectroscopy sought to conform to, and succeeded in improving upon, the conventional dispersion-based methods of the period. Focusing on practice and discourse, it will examine how the scientific problems undertaken were employed to highlight a particular and unfamiliar aspect of spectroscopic information.