THE POLAKISCOPE. 
157 
to 0 a' at same angle as o e, are called optic, or ray axes of the 
crystal. Another peculiarity of the surface is that a plane 1 1' 
parallel to o c, touching both the arcs c(,e, e 6, will touch the 
surface at every point of the circle ABC. 
Keturning to the polariscope (fig. 1), we have seen that by 
turniog p Q at right angles to A b no light is seen. If, however, 
a plate of a doubly refracting medium of proper thickness is 
placed on the stage k h, the light will generally reappear, and 
by turning the plate of crystal round, a position will be found in 
general in which the light is brightest. If the light falling on 
A B is white light, the crystal will also be beautifully coloured. 
Then, by turning the plate p q round its axis, the colours of the 
crystal will change. Selenite is a crystal easily obtained, and pro- 
ducing these colours very well. A lamina should be detached 
by a sharp knife. The lamina should not be too thin or too 
thick. A plate about as thick as a sheet of ordinary paper, 200 
leaves to the inch, will give a beautiful rose pink colour in 
one position of p Q, changing to a bright green in the opposite 
position. This plate, if split a second time, will give a bright 
blue colour, changing to gold-yellow. After one or two trials it 
is not difficult to find the best thickness of the plates to produce 
these colours. Tints of brilliancy and beauty equal to those 
produced in this way when a b is illuminated by the light of a 
white cloud are not to be produced, perhaps, in any other way. 
Instead of selenite mica may be employed. 
To explain these results we suppose selenite employed. 
Imagine o a'h' (fig. 5), the stage of the polariscope (fig. 1), of 
parallel to LM (fig. 1), and m' n' parallel to Gi. Two axes of 
elasticity in this crystal are in the plane of the lamina, and one 
axis of elasticity is therefore perpendicular to the plane. oa'E h' 
represents the form of the wave surface. 
The light coming from z (fig. 1) falls perpendicularly on the 
surface oa'e h' (fig. 5), and causes two sets of waves through 
the crystal parallel to the incident wave, one travelling with a 
speed proportional to oc', and the other proportional to oa, one 
polarized in plane oah' and the other in plane oaa'. 
The vibrations coming from z are wholly parallel to o/, and 
these are resolved in the manner described above in speaking of 
the interference of waves ; that is to say, into waves the vibra- 
tions of which are parallel to o h and o h'. Of these the first 
set of waves travels the fastest, and first gets out of crystal ; the 
second coming out after the first, will be a certain number of 
wave lengths behind it. When these sets of waves reach p q 
that portion of the motion only will be preserved which is 
parallel to g-i (fig. 1), or m'n' (fig. 5). We have fallen then on 
the curve (fig. 8) before explained. 'The intensity of the light 
