Suppose that a plate of tourmaline cut parallel to its crystallographic axis is fastened to the cap and the latter rotated. We should notice, on looking through the instrument, that in the course of a complete revolution there are two positions, orientated at right angles to one another, in which the tints of the two images are identical, the positions of greatest contrast of tint being midway between. If we examine a uniaxial stone in a direction at right angles to its optic axis we obtain the colours corresponding to the ordinary and the extraordinary rays. In any direction less inclined to the axis we still have the colour for the ordinary ray, but the other colour is intermediate in tint between it and that for the extraordinary ray. The phenomenon presented by a biaxial stone is more complex. There are three principal colours which are visible in differing pairs in the three principal optical directions; in other directions the tints seen are intermediate between the principal colours. Since biaxial stones have three principal colours, they are sometimes said to be trichroic or pleochroic, but in any single direction they have two twin colours and show dichroism. No difference at all will be shown in directions in which a stone is singly refractive, and it is therefore always advisable to examine a stone in more than one direction lest the first happens to be one of single refraction. For determinative purposes it is not necessary to note the exact shades of tint of the twin colours, because they vary with the inherent colour of the stone, and are therefore not constant for the same species; we need only observe, when the stone is tested with the dichroscope, whether there is any variation of colour, and, if so, its strength. Dichroism is a result of double refraction, and cannot exist in a singly refractive stone. The converse, however, is not true and it by no means follows that, because no dichroism can be detected in a stone, it is singly refractive. A colourless stone, for instance, cannot possibly be dichroic, and many coloured, doubly refractive stones—for example, zircon—exhibit no dichroism, or so little that it is imperceptible. The character is always the better displayed, the deeper the inherent colour of the stone. The deep-green alexandrite, for instance, is far more dichroic than the lighter coloured varieties of chrysoberyl.