330 BIOT ON THE EMPLOYMENT OF POLARIZED LIGHT 
same crystal, the deviation of the plane of polarization of one 
and the same simple ray is proportional to the thickness; and, 
for rays of different refrangibilities, this deviation, through one 
and the same thickness, is sensibly reciprocal to the squares 
of the lengths of their fits, and therefore increases with the 
refrangibility. Lastly, the direction of the deviations is con- 
stant for all plates taken from the same mass regularly cry- 
stallized; but it is occasionally different in different crystals, 
being in some directed toward the right of the observer, in 
others towards the left, with the same general laws of disper- 
sion, as well as of absolute intensity. This opposite action is even 
frequently observed in different parts of a plate cut from the same 
crystal. But then these portions are always limited by planes 
parallel to the faces of the hexagonal prism ; which shows that 
these planes are faces of junction, by which the crystals pos- 
sessed of contrary actions have become invisibly joined with one 
another, to form one total solid prism. All these peculiar effects, 
when merely the normal incidence is taken into consideration, 
are similar to those observed through liquids which deviate the 
planes of polarization of luminous rays. There is no difference 
except in the absolute magnitude of the deviations with equal 
thickness, and in their relations for the rays of different refran- 
gibility. Again, with respect to this last circumstance, if we ex- 
cept tartaric acid, which in its solutions by various liquids, and , 
in several combinations into which it may enter, disperses the 
planes of polarization of the various colours according to dif- 
ferent laws from other active bodies, all these act so nearly like 
rock crystal, that, when the absolute deviation has been ren- 
dered equal for a single one of the simple rays, by varying the 
thickness, the tints of the two images developed by the double- 
refracting prism in all its positions around the transmitted ray, 
appear identical in equal azimuths, and we can only ascertain 
differences by very delicate processes of opposition. Neverthe- 
less, from these analogies alone we could not legitimately infer 
that the action exercised thus by the rock crystal on polarized 
light is molecular. To attribute this character to it, we must 
prove that the same phznomena are likewise produced by the 
constituent elements of the crystal, that is to say by silicic 
acid in the state of disaggregation. Now, all the experiments 
tried with this view give negative results. Thus quartz fused 
