366 Professor Forses’s Researches on Heat. 
sion (17th December 1835). The fact is simply mentioned here, 
as we do not at present resume the subject of depolarization. 
35. I made some experiments, with a view to the determina- 
tion of the maximum polarizing angle for heat, with a more con- 
venient apparatus than the one above described. Heat was sim- 
ply reflected from the first surface of a thick mica-plate, and its 
state of polarization examined by means of a refracting bundle 
of mica, fixed in a tube (art. 26). Even then it is difficult to 
arrive at a direct conclusion ; for whilst, in the case of light, the 
variation in the quantity of polarized light reflected or transmit- 
ted at the polarizing angle varies with extreme rapidity, the 
same does not seem to hold true of heat,* the angle of minimum 
reflection at a second plate depending sensibly on the increased 
quantity of heat reflected at great angles of incidence, and there- 
fore making the apparent polarizing angle too small; a source of 
error which is not perceptible in the case of light, because of the 
abruptness of the polarizing action near its maximum. Thus, 
we must not simply incline the incident rays of heat differently 
to the reflecting surface until the intensity of the analyzed pen- 
cil reaching the pile is a minimum, because that minimum is on- 
ly apparent, being due to two sets of effects varying according 
to different laws ; one, the effect of polarization increasing up to 
a certain unknown angle of incidence and then diminishing ; the 
other, the intensity of a reflected pencil, constantly increasing up 
to an incidence of 90°. To make the experiment correctly, we 
must measure accurately, for a great number of angles, the pro- 
portion of the reflected pencil polarized in the plane of reflection, 
and take that which is a maximum. Such an investigation I 
have only very partially performed. 
36. I have, however, been enabled to determine approximate- 
ly the polarizing angle in a way which might appear at first sight 
much more complicated than the other, but which, at the same 
* Probably owing to its still more heterogeneous character. 
