M. MELLONI ON THE POLARIZATION OF HEAT. 337. 
We have found by transmission that the rays emerging from the green 
and opake-black glasses may be said to possess properties diametrically 
opposite to those of the rays issuing from alum*. The same antagonism 
of properties is manifested with respect to the apparent polarization 
which these two species of heat undergo in their passage through the 
tourmalines ; for in the one the index of polarization increases three- or 
four-fold, while in the other it suffers a diminution of eight or nine 
tenths. 
In fine, experiment has shown that the calorific rays immediately trans- 
mitted by alum approximate closely to the luminous rays, both in their 
abundant transmission through all uncoloured diaphanous substances 
jorm of the calorific spectrum preserved with great regularity, as we have seen 
elsewhere (Ann. de Chim. et de Phys., tom. 1x. p. 426. ; Scientific Memoirs, vol. i. 
part i.), that is to say, possessing one maximum, and lower temperatures regu- 
larly decreasing on each side of it, but we see that the distances from this maxi- 
mum and the surrounding bands toa given zone of the luminous normal spectrum 
remain sensibly invariable. As to the absolute quantity of heat, it varies con- 
siderably with the tint and nature of the glass; but this variation is always 
proportional to the value of the ordinates which represent the temperatures of 
the different zones for any one of the coloured plates; so that the intensities of 
the maximum and the adjacent bands are more or Jess-affected in a constant 
ratio through the whole extent of each new spectrum produced by changing 
the glass. From these two facts it clearly follows that the quality of the calo- 
rific stream transmitted by the different plates of coloured glass does not vary 
in its passage from one plate to another. In this however, as in the other ana- 
lyses that we have made of this phenomenon, the green glass possessing the 
qualities already mentioned presents a very striking exception: for this species 
of glass displaces the calorific spectrum and throws it, in the direction of the in- 
ferior refraction, almost totally beyond those limits that are common to the 
spectra produced by all the other species of coloured glass. 
W1.en several different methods (and the processes of absorption, polarization, 
and refraction described here and elsewhere are really such,) lead to one and the 
same conclusion, it seems to me that the conclusion is sufficiently secure to be 
ranked among truths firmly established by experiment. 
Thus the colouring matters of the coloured glasses, while they so power- 
fully affect the relations of quantity which the different rays of ordinary light 
bear to each other, exercise no elective action on the concomitant calorific 
rays. This curious phenomenon is the more remarkable, as the same colouring 
matters absorb, almost always, a very considerable portion of the heat naturally 
transmitted by the glass. ‘he following are, in fact, the calorific transmissions 
of the seven coloured glasses referred to the transmission of the colourless glass 
which is represented by 100: Red glass 82:5, Orange 72-5, Yellow 55, Bluish 
green 57:5, Blue 52:5, Indigo 30, Violet 85. The quantity of heat absorbed 
through the action of the colouring substances is therefore 17-5 in the red glass, 
27-5 in the orange, 45 in the yellow, 42-5 in the green, 47:5 in the blue, 70 in 
the indigo, and 15 in the violet. Now, as these absorptions extinguish a pro- 
portional part of each of the rays which constitute the calorific stream trans- 
mitted by common glass, they may be compared, as we said before, with the 
absorbent action exercised on light by matters more or less deeply brown or 
dark when they are immersed in water or some other colourless liquid which 
dissolves but does not affect them chemically. 
* Ann. de Chim. et de Phys., tom. lv. p. 382. 
