If the object under examination has a flaky surface, or is covered with a surface film, it may show iridescent colours such as are seen in a soap bubble. These colours are produced by inter ference between the light reflected by the upper and lower sur faces of the surface film or surface flake. With light falling almost perpendicularly on the surfaces, the reflected light is of feeble intensity unless the refractive index of the film or flake is high. The interference colours are easily masked, therefore, when a vertical illuminator is used, if the flake or film has a comparatively low refractive index, a little stray light being sufficient to prevent the interference colours from being seen. A film of high refrac tive index, however, such as a film of oxide or sulphide on copper, gives quite strong interference colours when illuminated with a vertical illuminator.
With the ring illuminator the light strikes the surfaces much more obliquely, and is more strongly reflected because of this. Any of this light which gets into the object-glass should show stronger interference colours than are seen with the vertical il luminator, and this actually occurs if the film or flake is of low refractive index. With a film or flake of high refractive index, however, the amount of light reflected at the first surface is so large that comparatively little light penetrates into the film or flake, and the amount which can be reflected at the lower surface is small. Moreover, the light which penetrates and is reflected at the lower surface is often much weakened by absorption in the film or flake, and the emerging light is of feeble intensity com pared with the light reflected at the upper surface. As a result, the interference colours are not in general easily seen when an object covered with a surface film of high refractive index is examined under a ring illuminator.
It may be mentioned, in passing, that these differences can be used to identify the cause of the iridescent colours seen on much antique glass and porcelain. If the interference colours show well under the vertical illuminator and are not seen so well, if at all, when the ring illuminator is used, this is strong evidence that the iridescence is due to a film which has a high refractive index, such as a film of tarnish on a thin metal foil burnt on or just into the glass or glaze, e.g., in "lustre" ware. If, on the other hand, the iridescent colours are shown best with the ring illuminator, they are probably produced by surface flakes of comparatively low refractive index such as result from disintegration of the glass or glaze. Evidence as to the refractive index of the iridescent ma terial can also be obtained if any idea of the thickness of this material can be formed by endeavouring to focus to the levels of its top and bottom surfaces down one edge (not through it). If
the refractive index is very high, the layer will be too thin for its thickness to be measured, but if the refractive index is compara tively low it should be possible to distinguish a difference of focus between the levels of the top and bottom surfaces when examined with a high power.
These interference colours, whether due to films or flakes, pre vent the true colour of the object frOm being recognised. They can, however, be got rid of by a method now to be described, in which illumination with polarized light is used. This method of illumination also causes much of the regularly reflected light and stray light to be cut out. It is commonly assumed that only doubly-refracting materials will be visible when examined with polarized light when the axis of the analyser is at right angles to the axis of the polarizer (crossed nicols). This is not so. Re flection at a surface can cause a sufficient change in the state of polarization of the light from the polarizer for an appreciable pro portion of the reflected light to be transmitted through the analy ser. Moreover, if the surface at which the reflection takes place is coloured, this colour will show up strongly in the light which passes through the analyser.
In order that examination by polarized light may be used satis factorily for the recognition of surface colour, it is necessary that the refractive index of the reflecting material should differ as much as possible from that of the material which surrounds it. The in creased reflection resulting from this is of no advantage, of itself, but this increased reflection is associated with a comparatively large change in the state of polarization produced by the reflection. The visibility of a non-doubly-refracting particle when viewed between crossed nicols is thus dependent in a very large measure on the difference between its refractive index and that of the material in which it is embedded. As a result, particles of high refractive index, such as metals or the more highly-coloured metallic oxides, show up quite strongly when surrounded by almost any transparent medium ; whereas particles having a refractive index not greatly different from that of the medium surrounding them, e.g., white "opal" particles in an opal glass, show compara tively feebly, giving but little light compared with that seen on the particles when examined with non-polarized light. To illus trate the ways in which the examination with polarized light can be applied to the identification of surface colours, a few typical examples may be given, though many other applications will occur to the microscopist of experience.