In 1845 Faraday discovered that if polarized light is passed through a transparent substance in a magnetic field and in the direction of the field, the plane of polarization is rotated. The amount of rotation for a given substance is found to be proportional to the strength of the magnetic field and to the length of the path in the material. As many substances, such as turpentine, a solution of common sugar, quartz crystals in the direction of their crystalline axes, etc., present us with a similar fact, this would not be so surprising save for a remark able difference in the two cases which may be thus described: When the plane of polarization is rotated by passing through a sugar solution or a similar body, and the transmitted light is reflected back upon its course so as to retrace its path, it is found that the original angle of polarization is perfectly restored by a precisely equal rotation in the opposite direction in the return; but a similar experiment upon the body given the magnetic rotation shows a doubled change of angle. This indicates that, although in the first case we must explain the rotation by the molecular constitution of the material, we are not permitted to suppose that the mag netic field has produced a similar molecular structure in the second case, since the rotation is constant in direction irrespective of the di rection of motion of the light. Of course, from the known nature of magnetism, this is equiva lent to asserting that there must be some rela tion between light and electricity. But this is not the most obvious connection between these two classes of phenomena, for as we now know, the earliest division of materials in accordance with their electrical properties involved a classi fication according to their most characteristic optical properties also. Thus all conductors of electricity, excepting only those liquids which undergo a chemical decomposition when they transmit an electrical current, and therefore be long to an obviously different class, are ex tremely opaque to light; conversely, all sub stances which are good insulators are also trans parent to light, at least to an extent which would make a sheet a few hundred-thousandths of an inch in thickness appearperfectly trans parent, although such a sheet of metal or simi lar conductor would not differ greatly in opacity from a thick plate. An excellent illustration of the generality of this law is furnished by the element carbon, which in the dense opaque form— like graphite, for example —is a very good conductor of electricity, but in the form of the transparent diamond is an insulator.
Before the middle of the 19th century two methods of measuring electrical magnitudes had been developed; one of these is based upon the repulsion which exists between two electrically charged bodies, and the other upon the repul sion which exists between two similar magnet poles. Elaborate and repeated investigations have demonstrated that if a given electrical magnitude is measured according to one of these systems, and the value thus found is compared to a measurement of the same quan tity in the other system, the ratio involves a velocity only. This statement is quite inde pendent of the kind of magnitude chosen for the experiment. Within the limit imposed by unavoidable errors of observation the value of this velocity always appears to be the same as the velocity of light.
Here, therefore, are three distinct relations between light and electricity, which have long been known and to no one of which it is pos sible to attach any a priori reason. It was left to Maxwell to illuminate this obscure field. His long and successful investigations in electricity and magnetism, especially his efforts to reduce Faraday's brilliant discoveries to correlation and to a consistent scientific statement, led him to the conclusion that light itself consists of elec trical vibrations. He attempted to test the va lidity of this hypothesis by every means at his command. For example, according to his theory a non-magnetic substance ought to have a die lectric constant, or what Faraday named its specific inductive capacity, equal to the square of its index of refraction. This indicated relation was found to hold with all expected precision in some cases, but to be re moved from the truth in others. Again, since, according to the theory, only those substances are transparent which will offer a resistance to the motion of electricity within them analogous to elastic reaction, there ought to be a de terminable relation between electrical conductiv ity and opacity. Maxwell attempted to find
this relation in the case of gold-leaf, which is sufficiently thin to transmit a measurable por tion of the light falling upon it. Notwithstand ing that the discrepancy was here found dis appointingly great, the gradual accumulation of knowledge o the more recondite phenomena of the electrical field had led the great majority of physicists to the conclusion that Maxwell's theory was at least a close approximation to the truth, and accordingly one of the most brilliant discoveries of the 19th century. This may be regarded as a fair statement of the attitude of the world of science in when Hertz, a German physicist, made a series of remarkable experiments which have eliminated all possible doubt as to the essential verity of Maxwell's theory of light. Fortunately it is not difficult for us to gain a sufficient knowledge of the character of these experiments to make clear their general bearing.
It had long been known that a Leyden jar suddenly discharged through a thick wire gives rise to an oscillatory current of very brief dura tion, and that in certain simple cases the period of the oscillations can be calculated with con siderable accuracy. Hertz recognized that dur ing the time of discharge such an electrical circuit must be a source of oscillatory changes in the magnetic field, which, if the views of Maxwell are in accordance with fact, should be propagated through space with the velocity of light. Although it is difficult, if not quite im possible, to measure directly this velocity, if one knows the wave-length and the period it is perfectly easy to deduce the velocity from these two elements, since in its period every wave moves a distance equal to its own length. In these experiments the period was calculated from the elements of the electric circuit; it only remained therefore to determine the length of the waves. Hertz accomplished this in the fol lowing simple and ingenious manner: At a con siderable distance from the source of the waves he placed a large sheet of metal perpendicular to its direction from the source. From this sheet the waves were returned upon them selves by reflection. Now, a well-known fact in wave motions is that when two systems of waves of like period are moving in opposite directions, they combine to form a system of standing waves of half the length of the free waves. The regions where motion is destroyed by this kind of interference are called nodes. Hertz demonstrated the existence and position of these nodes by means of an apparatus which possessed the same electrical period as the source. This apparatus he called a resonator. The value of the velocity of these waves de duced from his observations differs no more from the known velocity of light than would be expected from the unavoidable errors of obser vation; thus it complies with the requirements of Maxwell's theory. These waves, therefore, are shown to differ from light waves only in their enormously greater wave-lengths, and that they must be subject to all the established laws of optics which are independent of the length of the waves. The last conclusion was thoroughly tested by Hertz by a series of most interesting and convincing experiments. He found that strictly according to the laws of optics these waves are reflected from the sur faces of all bodies which conduct electricity; that they readily pass through substances which behave as insulators; and that in passing from one insulating medium to another the direction of propagation is altered in accordance with the law of sines. Further than this, he showed that such electrical waves admit of polariza tion, and they are, therefore, characterized by motions at right angles to the direction of propagation. During the time which has elapsed since these investigations a host of experi menters have improved the methods and ap paratus of Hertz, and have largely extended the range of wave-lengths that can be observed. On the other hand, many investigators have been employed in the application of analysis to both the old and the new problems in optics. The difficulties which attach to Fresnel's mode of regarding the optical phenomena of crystal line media are found to disappear, and all the complex phenomena of light admit of explana- tion from a consistent body of premises.