102 PROFESSOR FORBES'S RESEARCHES ON HEAT. 



actly as stated above, for incandescent platinum and darl^: heat, and somewhat 

 smaller for that of the Argand-lamp. 



46. I desire it to be recollected, that, in speaking of these somewhat starthng 

 lengths of waves of heat, I am using the language of only one of the two hypothe- 

 ses which serve to interpret the results of this section ; for, if the variation be in 

 — e, or the difference of the velocities of the doubly reflected rays in mica, the 

 result would be the same. The experiments in a subsequent part of this paper 

 may serve to guide us in our choice. Meanwhile, I would observe, that, supposing 

 the above results to be explained on the supposition that o — e is smaller, instead 

 of >. greater for heat than for light, it is equivalent to supposing the doubly refract- 

 ing energy weaker, or a greater thickness of a crystal required to produce a given 

 effect. Our suggestion respecting the existence of sensible vibrations normal to 

 the wave surface (art. 28) will not avail us here. For, by the mode of reducing 

 the experiments on Depolarization, the unpolarized part of the heat does not 

 enter into consideration at all ; * consequently those parts of the total eifect which 

 are due to transverse vibrations alone, are not modified by double refraction as so 

 much light would be. 



^3. On the Refrangibility of Heat 



47. Since the admirable discovery by M. Melloni of the power of rock-salt to 

 transmit and refract heat of every kind, one of the most obvious and important 

 questions (fonnerly intractable) of which it seemed to offer the means of solution, 

 was the accurate determination of the refrangibility of heat fi'om various sources, 

 luminous or non-luminous. Such a determination is of the first consequence to 

 the formation of a just theory of heat, and a detection of the subtle bond by 

 which it is connected with the comparatively familiar modifications of light. 



48. Such experiments have not been awanting. M. Melloni, in his Second 

 Memoir on Radiant Heat, in the Annales de Chimie for April 1834, has described 

 the apparatus which he employed, and which is figured in Plate III. of that vo- 

 lume. It consists of a thermo-electric pile, constructed of a single vertical row 

 of elements, so as to be exposed to a very narrow beam of heat. It was made to 

 move on a sector of a circle, at whose centre was placed a prism, by which the 

 beam of heat was refracted from its primitive direction a h into that c d, (see 

 next page), and therefore produced a maximum effect on the galvanometer when 



* I do Dot mean to offer any opinion on the nature of light in a partially polarized ray generally ; 

 but, as in the present case, the angle of incidence is that of complete polarization nearly, I presume that 

 the transmitted ray is undoubtedly composed partly of light polarized perpendicularly to the plane of in- 

 cidence, and partly of common light. 



