When light falls upon a thin film such as AEC, _fig. 9_, p. 80, so as to produce the colours of thin plates, it follows, from Sir Isaac Newton’s theory of fits, that a portion of the light is, as usual, reflected at the first surface AE,[22] while the light which forms the coloured image is that which is reflected from the second surface EC, so that all the colours of thin plates are diluted with the white light reflected from the first surface. Now, in the modern theory, which ascribes the colours of thin plates to the interference of the light reflected from the second surface EC, with the light reflected from the first surface AE, the resulting tint arises from the combination of these two pencils, and consequently there is no white light reflected from the surface AE. In like manner, when the thickness of the film is such that the two interfering pencils completely destroy one another, and produce black, there is not a ray of light reflected from the first surface. Here, then, we have a criterion for deciding between the theory of fits and the theory of interference; for if there is no white light reflected from the first surface AE, the theory of fits must be rejected. In a remarkable phenomenon of blackness arising from minute fibres, which I have had occasion to describe, there was no perceptible reflection from the surface of the fibres;[23] and M. Fresnel describes an experiment made to determine the same point, and states the result of it to have been unequivocally in favour of the doctrine of interference. In order to apply this important fact, let us take a piece of coal, one of the blackest and most opaque of all substances, and which does not reflect to the eye a single ray out of those which enter its substance. The size of its particles is so small, that they are incapable of reflecting light. When a number of these particles are placed together, so as to form a surface, and other particles behind them, so as to form a solid, they will not acquire by this process the power of reflection; and consequently, a piece of coal so composed should be destitute of the property of reflecting light from its first surface. But this is not the case,—light is abundantly reflected from the first surface of the coal, and consequently, its elementary particles must possess the same power. Hence the blackness of coal must be ascribed to some other cause than to the minuteness of its transparent atoms. To transparent bodies this argument has a similar application. As their atoms are still less than those of black bodies, their inability to reflect light is still greater, and hence arises their transparency. But the particles forming the surface of such bodies do reflect light, and, therefore, their transparency must have another origin.