216 FRESNEL. 
that a given ray loses it momentarily, and that another 
given ray, on the contrary, is deprived of it for ever? 
The theory of interferences, considered in this point of 
view, seems more like the reveries of a disordered brain, 
than the exact, inevitable consequence of numberless 
experiments, clear of all possible objection. And fur- 
cuted. But it was not until after lengthened investigation that the 
two philosophers just named succeeded in establishing experimen- 
tally the important law (obvious as it now seems,) that “ polarized rays 
can only interfere when they are polarized in the same plane.” If they 
were polarized in rectangular planes (for example), no interference 
could result, were all other conditions ever so perfectly fulfilled. Now, 
this could only be explained on the supposition of the vibrations 
being performed in planes transverse to the ray. Granting that in a 
ray polarized in one plane all the vibrations take place in one plane, 
(whether iz the same plane or perpendicular to it,) itis then readily 
seen that when the vibrations of two rays are at right angles to each 
other, there can be no mutual destruction, or mutual codperation. It 
is only when they are in the same plane that this can occur. 
This principle was at length found to supply the explanation of the 
polarized tints. Every ray of the light (P) originally polarized in one 
Cc 
a train Se 
Perit iM, -} 
plane, in traversing the crystal plate (c) was divided into two; an 
ordinary (0) and an extraordinary (e); all those of the one kind, 9, o/, 
o//, &c. being polarized in one plane, and all of the other, e, e/, e/”, &c., 
in a plane at right angles to the last. But in each ray o, and e, di- 
verge from each other by a very small angle. The whole pencil also 
diverges at a small angle from Pp; thus, the only rays which can coin- 
cide in direction, will be a ray 0, of one set, with a ray e/ of the next; 
—0o, with e/, &c. &c., and as these are unequally retarded in differ- 
ent degrees according to their inclination, they would be in a con- 
dition to give interference, were it not that being polarized in places 
