490 On the Reflexion of Light of Transparent Media. 
As this integral includes positive as well as negative elements, 
it is evident that this equation is possible for some values of 2. 
And if it be now supposed that for different bodies p is generally 
only approximately the same function of v, when v, increases by 
the positive increment dv,, this integral increases by 
i % af 
[e(1—) -\ pa | 
which is less than 
1 % dv 
[a(1—=) -p.{ 3 |e =0. 
This increase is therefore positive; and it follows that. the 
definite integral in (19), when v, exceeds a certain amount, 
is positive, and consequently A! negative for v, cos? «— cos* B <0 
or fore+ i if the light, as in Jamin’s experiments, passes 
from a less refracting to a more refracting body (e>). But 
we have, however, seen above that this case answers to that of 
bodies with a positive reflexion. 
Calculation therefore, like experiment, proves that positive 
reflexion occurs in the case of bodies with a greater index of 
refraction, negative reflexion in the case of bodies with a smaller 
index, while the difference of phase in the case of bodies which 
lie between these passes suddenly from 0 to +7. 
The intensity of the reflected light polarized perpendicularly 
to the plane of incidence is, according to (17), 
tan? (2—£) 
2 2 Al). 
tan? ere is. 
the intensity of the reflected light polarized in that plane is, 
according to (16), 
“ULL ow (1-+ tan? A), 
sin? (@ + B) 
If the ratio of these two intensities be expressed by k*, we get 
p= 0% (2+) cosA (20) 
~ cos (a—Z£) cos Al” 
These results agree with those of Cauchy. 
As Jamin has in his experiments determined these magnitudes 
directly, it is possible from his experiments, that is, from the 
