158 
POPULAR SCIENCE REVIEW. 
reflected by P Q will depend on tbe thickness of the crystal, on 
the direction o/, and on the angle between of and m'n\ 
Eeturning to figure 1, the light in zu is polarized in the plane 
of incidence ; that is, each particle moves backwards and forwards 
in a straight line parallel to lm or de. In like manner all the 
light reflected from pq must consist of vibrations parallel to 
IG. It is easy to see that a motion parallel to lm cannot give 
rise to a motion at right angles to it, parallel to iG ; and so no 
light is reflected by pq when lm, gi are at right angles. The 
reason why the light reflected from AB at the particular angle 
stated before must consist of vibrations perpendicular to the 
plane of incidence is not easy to explain in an elementary way. 
Suffice it to say, that the experimental truth has been satisfac- 
torily explained, first by Fresnel, since more completely by Grreen 
and others, to whose works we refer our readers. 
The reason of the colouring is that the red light being made 
up of waves of greater length than the blue or yellow light, of 
the two waves into which red light is split up in the crystal, the 
one travelling slowest will be behind the other on coming out of 
the crystal, by an amount of space depending on the thickness 
of the crystal and on the difference of speed. This space will 
be a smaller fraction of the wave length in the case of red light 
than in that of blue light, and therefore when these waves have 
been reflected by the analyzer, and that portion of the motion of 
each is preserved which is parallel to Gi (fig. 1), the two waves of 
red light to be compounded (as in fig. 8) will differ in phase 
by a different amount, in the cases of red light and blue light ; 
and the intensity of the resultant wave will also be different in 
the two cases, or the light will be coloured. These phenomena 
are complicated by the fact, that the speed of light on one 
colour is not the same as that of another colour in the same 
crystal. 
We have in this case supposed the light passing through the 
crystal to consist wholly of parallel rays. If instead of these we 
use a pencil of very divergent rays, a new and very beautiful set 
of colours and rings are produced. To produce these we must get 
a plate of crystal cut perpendiculax to its axis, if it is uniaxal; or 
perpendicular to the plane containing the optic axes, if it is a 
biaxal crystal. The plate should be placed on the stage kh, and 
a lens of short focal distance under it to receive the light coming 
from z. Or if the crystal is small, a little arrangement used by 
Sir John Herschel, and described in The Philosophical Transac- 
tions ” for 1820, may be employed. A b (fig. 6) is a lens on which 
light falls in parallel rays, and converges to the point t. AB is 
mounted at end of a brass tube acdb : in this tube fits closely 
a^smaller piece of tube carrying the small plate of tourmaline t 
This can be turned round through an angle of some 120° by 
