120 Professor Forbes's Researches on Heat. 



equally opake, or, rather perhaps, I should say, equally in- 

 different to the kind of incident heat (i. e. colourless in optics). 



47. So far as the eye could judge of the proportion of ob- 

 stacles in surfaces strewed with different kinds of powders, 

 there did not seem any very marked peculiarity in their trans- 

 parency for heat. A surface dusted with alum or citric acid 

 appeared to transmit nearly as much as one strewed with 

 powdered rock-salt. Nor could this arise merely from the 

 minute thickness of the substance, which is well known to pro- 

 duce in heat, as in light, an approximation to a colourless 

 character ; for the proportion stopped by the powder was al- 

 ways a large fraction (usually from f ths to T % ths) of the inci- 

 dent heat. The opacity, then, is the result of the innume- 

 rable reflexions and interferences which scatter and stifle the 

 transmitted heat; and this is almost equally effectually done, 

 whatever be the nature of the substance. On reflection, 

 therefore, this general result does not appear surprising. I 

 will quote one experiment, in particular, in illustration of it. 



48. When I was at a loss to procure fine metallic fibres, 

 I thought of employing a diaphragm irregularly covered with 

 fine threads of spun glass, with a view (just as in the case of 

 the alum powder) of ascertaining how far the mechanical 

 condition of the glass might modify its well-known qualities 

 with respect to the transmission of heat. When Locatelli 

 lamp-heat, having been transmitted by thick plate-glass, fell 

 upon the spun-glass fibres, forming an irregularly reticulated 

 diaphragm, no more than 47*5 percent, of the incident heat 

 was transmitted. Now, we know perfectly from the experi- 

 ments of De la Roche and Melloni, that, after passing through 

 such a thickness of plate glass, an additional film, the thick- 

 ness of the glass fibres used would produce no sensible resist- 

 ance to the further passage of the heat, excepting only its su- 

 perficial reflexion. The loss of 52*5 per cent, of the heat 

 was therefore due to the scattering and stifling of heat by re- 

 flexion at the surfaces of the fibres, refraction through their 

 cylindric surfaces, and interference. We cannot, therefore, 

 be surprised, if the refracted part of the heat reaching the 

 pile (the only portion very materially affected by the nature 

 of the medium) should not greatly alter the quantity of differ- 

 ent sorts of heat indicated by the galvanometer. Accordingly, 

 we find, that heat from a dark surface of brass warmed by an 

 alcohol lamp, had 44 per cent, transmitted under the same cir- 

 cumstances ; and even hot water had 42 per cent., although a 

 small thickness of glass is sensibly opake for that kind of heat. 



49. If this be the case, — if the differences be so trifling — 

 for a reticulation of regularly- formed, transparent, and po- 

 lished threads of glass, much more must it hold with impal- 

 pable crystalline or other powders, presenting (no doubt) mi- 



