157 
crystallized surfaces upon light. 
I now tried rock crystal and oil of anise seeds , which have 
nearly the same mean refraction ; but on account of the great 
debility of the interior polarising force, it was not able to 
overpower or even to modify that which accompanies partial 
reflection. I could easily have reduced this last force still 
farther till it came under the dominion of the first ; but the 
reflecting power would have been reduced in the same pro- 
portion, and would not have been capable of driving back a 
number of rays sufficient to form a perceptible image. But 
though the polarisation is not changed at the separating sur- 
face of the oil and the rock crystal, yet the character of the 
reflected light is modified in a very remarkable manner. 
When the plane of reflection from one of the sides of the 
prism of rock crystal was in the direction of the axis, or in 
o° of azimuth, the reflected image was a deep blue of very 
little intensity ; whereas in a rectangular direction, where 
the azimuth was 90°, it w-as of a brick red colour, and much 
more luminous. On one of the faces of the pyramid, in azi- 
muth o°, the tint was a. brilliant pink, intermediate between the 
red and the blue ; and on the same face, in 90° of azimuth, it 
was of a brick led colour as before. These variations are 
obviously related to the axis of double refraction, and indicate 
the extension of its force within the sphere of partial reflec- 
tion. The origin of the colours themselves, I shall soon have 
occasion to explain, in a paper on the action of uncrystallized 
surfaces.* 
3. General results deduced from the preceding experiments. 
Had it been in my power to command a series of the most 
* This Paper was read before the Royal Society of Edinburgh, on the 4th January 
*S 19. 
