Aether is as ancient as ancient Greece. The Greeks considered aether the fifth and the most subtle element of nature, the matter of stars. In modern times, Descartes explained the motions of planet by aethereal whirls around the sun. When the existence of electromagnetic waves, announced in Maxwell's electrodynamics, was experimentally corroborated (Hertz, 1888), one the most important questions became that of the environment in which the waves propagate. Indeed, all the waves hitherto known required some environment for their propagation. The natural candidate was - what else? - cosmic aether, mentioned already by Huyghens as the material basis for the phenomena of light. Aether was also used to determine the inertial reference system: it was the one in which aether remains at rest. Because of so many extraordinary features (it was also believed to be transparent and weightless), the experimental discovery of aether meant a serious challenge for the physicists that could only be coped with using sophisticated interferometric methods. Electromagnetic waves were thought to be elastic strains of aether. According to the theory of aether at rest (Fresnel), ponderable bodies could pass through aether without any resistance, whereas by the theory of convected aether (Stokes and Hertz) they would drag it along. With respect to slow motions, like that of the planets, aether would behave like a viscous liquid; with respect to rapid motions, like field vibrations, it would behave like an elastic body. The experimental problem consisted in establishing whether we are surrounded by an aethereal wind or not.
In spite of Maxwell's electrodynamics being a non-mechanical theory, it used to be interpreted in terms of mechanics. Maxwell himself thought of the field lines as very thin tubes filled with incompressible liquid: aether. This analogy was undoubtedly inspired by the formal similarity between the formulas of electrostatics and the description of irrotational motion of incompressible liquids. Lord Kelvin himself was the author of an interesting idea which used aether to explain the structure of matter. Atoms were for him but knotted aether whirls and he hoped that a classification of knots would recreate Mendelyeev's periodical system of elements. In Kelvin's theory chemical particles were chains composed of such links - knots. Stability of atoms was certainly the weak point of this theory. Whirls moving in the aether would not be stable.
In the meantime optical experiments refuted the theory of convected aether. Even if some of them remained consistent with the theory of aether, the idea of aether was rejected. Michelson-Morley's experiment proved that the velocity of light was the same in all inertial reference systems. This could be made to agree with the existence of aether, but the theory would become very unclear and full of new difficulties. The story bears some similarity to that of Ptolemy's theory. As the precision of observation became greater, the gap between the model and experimental data grew wider. Ptolemy's model of the planetary system could be saved by adding more epicycles, but this would only initiate an endless process, the model itself becoming more and more absurd and deprived of its fundamental value. It would be of no use in explaining the known phenomena and predicting new. The corrected theory of aether would have suffered the same fate.
The idea of the non-existence of aether yielded a much simpler and efficient description of the world. Electromagnetic waves can move in the vacuum and the structure of time and space is different from that of the Newtonian model. The new image of the world, born together with the relativity theory, offered science much more than the theory of aether could offer. Although not free of paradoxes, it proved to be simpler and more elegant.