Colour is merely an impression produced upon the retina, and therefore on the brain, by various surfaces or media when light falls upon them or passes through them.  Remove the light, and colour ceases to exist.  The colour of a substance does not depend so much on the chemical character of that substance, but rather and more directly upon the physical condition of the surface or medium upon which the light falls or through which it passes.  I can illustrate this easily.  For example, there is a bright-red paint known as Crooke’s heat-indicating paint.  If a piece of iron coated with this paint be heated to about 150 deg.  F., the paint at once turns chocolate brown, but it is the same chemical substance, for on cooling we get the colour back again, and this can be repeated any number of times.  Thus we see that it is the peculiar physical structure of bodies which appear coloured that has a certain effect upon the light, and hence it must be from the light itself that colour really emanates.  Originally all colour proceeds from the source of light, though it seems to come to the eye from the apparently coloured objects.  But without some elucidation this statement would appear as an enigma, since it might be urged that the light of the sun as well as that of artificial light is white, and not coloured.  I hope, however, to show you that that light is white, because it is so much coloured, so variously and evenly coloured, though I admit the term “coloured” here is used in a special sense.  White light contains and is made up of all the differently coloured rainbow rays, which are continually vibrating, and whose wave-lengths and number of vibrations distinguish them from each other.  We will take some white light from an electric lantern and throw it on a screen.  In a prism of glass we have a simple instrument for unravelling those rays, and instead of letting them all fall on the same spot and illumine it with a white light, it causes them to fall side by side; in fact they all fall apart, and the prism has actually analysed that light.  We get now a coloured band, similar to that of the rainbow, and this band is called the spectrum (see Fig. 16).  If we could now run all these coloured rays together again, we should simply reproduce white light.  We can do this by catching the coloured band in another prism, when the light now emerging will be found to be white.  Every part of that spectrum consists of homogeneous light, i.e. light that cannot be further split up.  The way in which the white light is so unravelled by the prism is this:  As the light passes through the prism its different component coloured rays are variously deflected from their normal course, so that on emerging we have each of these coloured rays travelling in its own direction, vibrating in its own plane.  It is well to remember that the bending off, or deflection, or refraction, is towards the thick end of the prism always, and that those of the coloured rays in that