(The best way of obtaining a knowledge of these phenomena is to construct a model of thin wood or pasteboard, representing the plate of gypsum, its planes of vibration, and also those of the polarizer and analyzer.  Two parallel pieces of the board are to be separated by an interval which shall represent the thickness of the film of gypsum.  Between them two other pieces, intersecting each other at a right angle, are to represent the planes of vibration within the film; while attached to the two parallel surfaces outside are two other pieces of board, which represent the planes of vibration of the polarizer and analyzer.  On the two intersecting planes the waves are to be drawn, showing the resolution of the first polarized beam into two others, and then the subsequent reduction of the two systems of vibrations to a common plane by the analyzer.  Following out rigidly the interaction of the two systems of waves, we are taught by such a model that all the phenomena of colour obtained by the combination of the waves, when the planes of vibration of the two Nicols are parallel, are displaced by the complementary phenomena, when the planes of vibration are perpendicular to each other.)

In considering the next point, we will operate, for the sake of simplicity, with monochromatic light—­with red light, for example, which is easily obtained pure by red glass.  Supposing a certain thickness of the gypsum produces a retardation of half a wave-length, twice this thickness will produce a retardation of two half wave-lengths, three times this thickness a retardation of three half wave-lengths, and so on.  Now, when the Nicols are parallel, the retardation of half a wave-length, or of any odd number of half wave-lengths, produces extinction; at all thicknesses, on the other hand, which correspond to a retardation of an even number of half wave-lengths, the two beams support each other, when they are brought to a common plane by the analyzer.  Supposing, then, that we take a plate of a wedge form, which grows gradually thicker from edge to back, we ought to expect, in red light, a series of recurrent bands of light and darkness; the dark bands occurring at thicknesses which produce retardations of one, three, five, etc., half wave-lengths, while the bright bands occur between the dark ones.  Experiment proves the wedge-shaped film to show these bands.  They are also beautifully shown by a circular film, so worked as to be thinnest at the centre, and gradually increasing in thickness from the centre outwards.  A splendid series of rings of light and darkness is thus produced.

When, instead of employing red light, we employ blue, the rings are also seen:  but as they occur at thinner portions of the film, they are smaller than the rings obtained with the red light.  The consequence of employing white light may be now inferred; inasmuch as the red and the blue fall in different places, we have iris-coloured rings produced by the white light.