On Maxwell's theory, metals, in which the electrical conduc tivity is very high (meaning that an actual convection of electric charge, as distinct from a displacement current, is easily set up in the medium), reflect radiation very well, and also absorb very strongly that part of the radiation which is not reflected. The higher the conductivity, the larger the coefficient of absorption.
i It is an experimental fact that in the region of Hertzian waves the absorption by metals is very high: for instance Branly found that tinfoil .008 mm. thick did not let the waves through in measurable intensity. From the same theory it follows that per fectly insulating dielectrics should be perfectly transparent. This has also been to some extent confirmed experimentally, Righi having shown by a direct method that sulphur, ebonite, paraffin and selenite are quite transparent to waves 5 to 20 cms. in length, but glass, marble and wood absorb to some extent. The comparatively good transmission of long electromagnetic waves by substances which are, broadly speaking, non-conducting, is a familiar fact to people living in brick houses who use an inside aerial for the reception of broadcast waves.
According to Maxwell's theory absorption is entirely governed by the conductivity of the medium, and a relation can be worked out which gives the absorption coefficient in terms of the con ductivity and the dielectric constant. A great number of measure ments have been made by Drude and his followers on the absorp tion of feebly conducting substances, such as electrolytes and low pressure gases. Qualitatively the relation holds in all cases. Quantitatively it has been confirmed for water solutions of salts, of which the conductivity is measured directly, while the dielec tric constant is taken as being that of water. With many organic substances, however, especially those containing hydroxyl groups, the measured absorption is greater than that given by the theo retical formula. Drude tried to explain this by the presence of resonators of the type that account for optical absorption, with a large damping coefficient, but there seems no possibility of resonators of sufficiently large free period. A more promising theory is that of Debye, who looks for the explanation in the presence of dipoles in the molecules, which tend to set themselves in the direction of the electric force. When the oscillations are
comparatively slow, as in tht. Hertzian region, the dipoles can follow more or less the alternations of electric force, provided the medium is not too viscous, and dissipate the energy, but when the alternations are very rapid, as in the optical region, they have no time to orient themselves, on account of their great inertia. Some substances show narrow regions of anomalous dispersion in the Hertzian region which do not accord with this explanation. Speak ing broadly the Hertzian waves do experience the strong absorp tion in good conductors and the feeble absorption in dielectrics which can be explained on Maxwell's theory without any consider ation of the structure of the medium, or its chemical nature.
The longer infra-red waves are also well removed from the region where the oscillators of the type considered in optical dis persion make themselves felt, and many of the consequences of Maxwell's theory, based on the assumption of a structureless medium, have been confirmed with such waves by Rubens and others, such as the proportionality of the refractive index to the square root of the dielectric constant for non-conductors. For such experiments the long infra-red waves possess the advantage that they do not necessitate the large-scale apparatus required for Hertzian waves many centimetres in length. Considering the absorption of infra-red and optical radiations a distinction must be made between the absorption associated with metallic reflection and that conditioned by the atomic oscillators. In the case of strong reflection, exhibited by such crystals as those used for isolating the rest-rays, as well as by metals, the absorption is in general so high that the radiations do not penetrate any appreciable distance into the substance. (The effect takes place with crystals only for a radiation whose frequency agrees with that of the free period of the natural vibration of the crystal.) The surface colour, using the term in a general sense, to apply to invisible as well as to visible radiations, is approximately corn plementary to the colour transmitted by an exceedingly thin layer, as exemplified by the green light transmitted by a gold leaf, which appears yellow by reflected light. In the case of the ordi nary absorption associated with anomalous dispersion, the radi ations penetrate further into the body, and the colour by reflected light is the same as the colour by transmitted light. Right down into the optical and ultra-violet region the opacity of metals holds, but it ceases in the X-ray region, where the oscillations are far too rapid to affect the electrons which are responsible for the conduction. It may be noted that there is a certain selectivity in the metallic reflection in the optical region which is not ac counted for by the simple Maxwellian theory: for instance gold reflects red selectively as can be seen by reflecting light many i times between two plane gold surfaces. The issuing light is deep red.