334 BIOT ON THE EMPLOYMENT OF POLARIZED LIGHT 
axis, which is inherent in the general construction of the crystal. 
Let us then leave the normal incidence, and incline progressively 
the plate on the polarized ray, keeping always the plane of its 
faces perpendicular to the plane of primitive polarization. The 
transmitted ray will become oblique to the axis; and if it were 
composed of natural light, it would undergo internally double 
refraction. But being wholly polarized in the plane of incidence, 
which likewise contains the axis, it escapes this effect, and should 
be transmitted undivided, undergoing ordinary refraction, which 
preserves to it its direction of primitive polarization, since it co- 
incides with that which this refraction tends to give it. These 
results of the general theory may be observed through all plates 
of uniaxal crystals cut normally to this line, and they are there 
realized very exactly. But they are quite different in plates of 
rock crystal perpendicular to the axis of the crystals, from the 
intervention of the rotatory power, which is exerted simultane- 
ously with the general double refraction of which it is inde- 
pendent. For first, under the normal incidence, where this 
power acts alone, the whole of the transmitted ray is modified by 
it. In fact, if, after its emergence, it is analysed by a double 
refracting prism, having its principal section parallel to the plane 
of primitive polarization, this ray, instead of passing through 
simple, under the sole influence of ordinary refraction, is broken 
into two distinct portions differently coloured, one of which, O, 
follows the ordinary refraction ; the other, E, the extraordinary 
refraction, forming images similar in colour, as well as in relative 
intensity, to those which would be observed in the same posi- 
tion of the analysing prism if the same ray had traversed a cer- 
tain thickness of a colourless, active liquid. But, under oblique 
incidences, the phenomena become very different for the liquid 
and for the crystal. When the stratum of liquid is progressively 
inclined, preserving it perpendicular to the primitive plane of 
polarization, the two images, O, E, change colours and relative 
intensities, from the increase of length of the course of the ray 
among the active molecules. When the same experiment is made 
with a plate of rock crystal perpendicular to the axis, the image 
KE again varies, as to its tint, conformably to the same rule. But 
its relative intensity decreases more and more in proportion as 
the inclination increases ; and the rays which it abandons, being 
transferred into the image O, recompose there a white more and 
more abundant with the proper portion of that image which is 
complementary to them, and which weakens by so much its 
