146 Professor Forzes on the Refraction 
strontia (the muriate is better); the yellow, with muriate of 
soda; the green, boracic acid; the blue, pure alcohol. The un- 
steadiness of intensity of an alcohol flame prevents great numeri- 
eal accuracy. 
Number of Rays of Heat out of 100 transmitted by 
Colour of Flame. ; Alum. Glass. Rock Salt. 
Red, : : : : 2 11 26 85 
Yellow, . : : : ? 113 28 87 
Green, . : : . - 11 26 84 
Blue, : ; F ; ; 10 30 83 
The differences are certainly within the limits of errors of obser- 
vation. 
30. I am disposed to believe, however, that in these experi- 
ments, as well as Meuuont’s, some effect is probably due to the 
simple presence of light of a particular quality, though its heat- 
ing power may be small. This my experiments with tourmalines 
countenance. We can hardly, however, look for a solution of 
these difficulties, until some of the most stubborn difficulties in 
the theory of light, the laws of dispersion and absorption (and es- 
pecially that peculiar absorptive power which permits the tour- 
maline only to transmit one polarized pencil) are completely over- 
come. Meanwhile, we pass with pleasure to the consideration of 
some of those properties of heat which serve to connect it with 
the best determined and best explained departments of optics. 
§ 3. On the Polarization of Heat by Refraction and Reflection. 
31. Soon after the discoveries connected with the polariza- 
tion of light, which illustrated the earlier part of this century, 
the question of the polarization of heat was taken up by Matus 
and Berarp.* In the case of heat accompanying solar light, it 
was decisively proved, as might have been anticipated ; but in the 
case of heat from terrestrial, and especially non-luminous sources, 
though M. Berarp considered that he had proved it, he gives 
* Memoires d’Arceuil, tom. iii. 
