Chapter 1. On Early Theories of Light and The Ether
1.1 Space, ether, light waves and corpuscles
Our natural perception of space by itself is that of an emptiness. Objects, like ourselves and the surrounding bodies of atomic matter, the Earth, the planets, etc., are considered to fill with their volumes appropriate bites of the empty space. These objects are naturally conceived as consisting of continuous dense matter. The emptiness of space is also deduced from the fact that space does not resist the motions of these bodies, that space is transparent to them. Hence, apart from the geometrical property of being naturally three-dimensional, empty space should not have any physical properties.
The wave theory of light was developed by Christian Huygens (1629-1695) in his Traité de la Lumière (presented to the French Academy in 1678 and published in 1690). The theory had to specify the substance carrying the light waves. This substance must fill the whole space, because "holes" for the light were not known. It must also be transparent to the motions of bodies, thus highly diluted of matter. This thinness was expressed in the name "ether", given to the carrier of light waves.
The velocity of light was first calculated by Olaus Roemer (1644-1710) in 1676 from the observed eclipses of Jupiter's satellites. The obtained value was very close to 300 Mm/s. The wavelengths of the light-waves were not yet measured then, but they were known to be short. With the velocity of light, exceeding the highest known velocities of waves in matter by a factor of , the frequencies of the light waves must have been considered tremendously high, as compared with the known frequencies of sound. Therefore, the elastic moduli of the etherous carrier of light had to be assumed to exceed the elastic moduli of steel by thousands of times. Hence, from its very inception the ether was carrying the intolerable contradiction between the requirements to be as empty as the vacuum space, and much stiffer than steel.
The corpuscular theory of light had no need for the ether, considering light as streams of particles, emanating from the sources of light. This theory was developed by Isaac Newton (1643-1727) in his book Optiks (1672). The corpuscular theory reigned for 140 years, until 1815, when the works of Augustine Jean Fresnel (1788-1827) begun on the diffraction and interference of light. Fresnel showed that light beams, meeting in space, can undergo destructive interference, suppressing or even cancelling each other. He proved, therefore, that these beams consist of waves.
Diffraction and interference of light were known to Newton. He repeated F. M. Grimaldi's experiments (1665, postmortem publication) on the diffraction of light on rods and in gratings. Newton also repeated Robert Boyle's experiments (1663) with color rings on thin films. He then derived the dependence of the colors on the thickness of the film, and the phenomenon is known since as "Newton rings". Undoubtedly, Newton must have known that these phenomena indicate some wave properties of light. Unlike so many of his followers, Sir Isaac was not fanatic and did not blindly deny experimental results unfit for his theories. We may therefore believe that Newton was the first the realize, that in some phenomena light exhibits corpuscular properties, while in other phenomena undular properties are revealed. This is the meaning of the "particle-wave duality" concept in quantum physics.
1.2 Transverseness of light and the inconsistency of the ether
The polarization of light was discovered by E. L. Malus (1775-1812) in 1809. Althought a wave phenomenon par excellence, polarization was used by Malus, amazingly enough, to disprove the wave theory of light. In the wave theory, polarizability means that light waves are transverse. But bulk transverse waves can propagate only in solids. The wave theory is contradictive more than enough in forcing us to believe that the thinnish ether can carry longitudinal waves, as do gases and bulks of liquids. Malus reasoned that to assume the zero-density ether as a solid, carrying transverse waves, will be beyond the ability of human reasoning, hence the wave theory must fall. Malus did not know how easily we adjust to anything which contradicts normal reasoning, provided that it serves our purpose and that necessary time is given for the adjustment.
The transverseness of light-waves was first pronounced in a hypothesis by Robert Young (1773-1829) in 1817. The transverseness was experimentally proven by Fresnel and Arago. They showed that two coherent light beams, polarized in mutually perpendicular planes, do not interfere with each other. Such a result is only possible when the two beams consist of transverse waves. Light waves were since understood as elastic transverse waves of the ether. The ether had to be considered then as a solid, with stiffness exceeding that of the densest solids by many thousands of times. And, as the search of ultraviolet radiation (since 1801) continued towards shorter wavelengths, the assumed stiffness of the ether had to increase. Thus, the contradiction with the required vacuum-like thinness of the ether also increased. The exclusive transverseness of light waves, i.e., the fact that the solid ether does not carry longitudinal waves, too, as do all solids, added one more contradiction to the ether concept.
In spite of the inconsistencies of the ether model, it provided the expected service to the explosive development of optics and electromagnetism in the XIX century. It survived in full glory until the publication in 1887 of the results of Michelson-Morley's experiment, which denied the existence of the ether. Even after that the ether concept was used half a century, though decaying in value, and many attempts were made to revive it. So lively is the concept, that the words "ether" and "the waves of ether" are still in use.
1.3 Lack of ether winds around Earth and the fiasco of the ether concept
The velocity of Earth in her motion around the Sun was first calculated by James Bradley (1693-1762). In 1728 he discovered the aberration of starlight and interpreted it as due to this motion. From the value of the aberration (up to 20.5 seconds of arc per year) Bradley found that the orbital velocity of Earth should be of the velocity of light, i.e., ~30 km/s. It was therefore expected, that the motion of Earth in the ether should cause measurable ether currents or winds flowing around the Earth. Attempts to detect ether winds, also with the use of interference methods (e.g., by Jacques Babinet, 1839) were unsuccessful.
The first experiment able to prove decisively that ether winds do not exist was the experiment by A. A. Michelson (1852-1931) and E. W. Morley (1838-1923), published in 1887. In their experiment, light from a monochromatic source was split into two beams, parallel and perpendicular to any direction of Earth's motion. After passing adjustable, slightly different optical paths, the two beams were brought together to produce their interference fringe pattern. Then the whole apparatus was turned by 90°, so that the two beams exchanged the directions of their paths. If there were a difference between the velocities of light along the two paths, then during the exchange of orientation of the two beams their interference fringe pattern should have been shifted. Although the apparatus was capable of measuring a fringe-shift smaller than the expected one, none was found. The experiment was performed at different seasons and locations, and no difference could be found in the velocity of light in directions along the expected ether wind, opposite to it, or perpendicular to it. Therefore, there are no ether winds or currents; hence, there is no ether.
1.4 Trials to revive the ether concept
The denial of the ether was not easily accepted by the scientific community, and much work was done to defend the ether concept and to explain the result of Michelson-Morley's experiment in ways, compatible with the existence of the ether. One way was to explain the absence of an ether wind by assuming that the moving Earth drags the ether along. Other ways were to invent effects, funny for those days, which could save the ether by twisting established physical definitions and laws. For example, G. Fitzgerald suggested (1893) that fast moving bodies shorten their dimension in the direction of motion. Therefore, in Michelson-Morley's experiment the path of light in the direction of Earth's motion would become shorter. Fitzgerald was later joined by H. A. Lorentz (1895), the mathematics was worked out, and the alleged effect became known as the Fitzgerald-Lorentz contraction of moving bodies.
Instead of the contraction of the path along the direction of Earth's motion, a time-dilation effect could also save the ether. The time dilation, derived from Lorentz transforms, stated that to an observer at rest a clock appears to tick slower when it moves, than when it is at rest. If we observe an event in a moving frame of reference. then the clock which is resting with us will record for the event a longer time-interval, than an identical clock positioned in the moving frame.
Yet another way of trying to save the ether was to question the accuracy of Michelson-Morley's experiment. Therefore, the experiments were repeated in 1904 by Morley and D. C. Miller. They could detect a hundred times smaller effect than the one expected from an ether wind, but found none. Then came a series of measurements on high altitudes: on the top of a mountain (1920), and in a stratostate balloon (A. Picard); G. Joos in 1930 increased the accuracy of measurements so that he could have detected an ether wind of velocity ~1 km/s. All experiments yielded negative results. Therefore, the conclusion that there is no ether stands fast.
Curiously, both the Fitzgerald-Lorentz contraction and the time dilation did not revive the ether concept, for what they were invented. On the contrary, they were later used by A. Einstein in his relativity theory, helping to deny the existence of ether. As we shall prove later (Section 9.6), these effects are unreal, because in the electron-positron lattice-space atomic bodies would disintegrate at velocities a hundred times lower than the velocity of light.