198 



PROFESSOR FORBES'S RESEARCHES ON HEAT. 



and farther, if it be admitted that such heat passes most difficultly through such 

 substances as glass, it follows, that after total reflection has proceeded a certain 

 way, so that the more refrangible, and therefore more transmissible, rays have 

 suffered total reflection, whilst the remaining rays constituting the primitive beam 

 continue to be refracted, the heat thus reflected will be more copiously transmit- 

 ted by glass, than when it came direct from the source. This conjecture was 

 precisely verified. 



62. Subsequent experiment stiU more fully confirmed this result, and by 

 shewing that, during the whole progress from partial to total reflection, the specific 

 quality of the heat changes, gave countenance to the view that the gradation is 

 in a great measm-e owing to the want of homogeneity of the heat, and that the 

 figure of the curve becomes (as we have said) a real test of the composition of a 

 ray. 



63. At the inferior limit of the curve, or when partial reflection takes place, 

 all kinds of heat are equally reflected (in the case of light, the light is white), just 

 as at the superior limit, or after total reflection is complete, the beam has exactly 

 the same relative composition as before. In the intennediate stages the compo- 

 sition is perpetually var3dng. The first rays totally reflected (and combining with 

 the scattered and partially reflected rays) are the more refrangible, or those more 

 easily transmitted by glass. At a certain point a maximum proportion of these 

 enter into the reflected beam. As the angle of incidence becomes greater, more 

 and more of the less refrangible rays enter into the composition of the reflected 

 heat, which at last possesses the same qualities as at first. This is weU illustra- 

 ted by the following early experiment which I made on the proportion of the re- 

 flected rays transmitted by a plate of glass .06 inch thick, at different stages of 

 reflection (7th February 1838). 



Diagonal ab, 

 Fig. 1, in 

 Inches. 



Deviations of Galvanometer. 



Ratio. 



REMARKS. 



Glass. 



No Glass. 



14.5 



15.0 



16.26 



15.5 



15.75 



16.0 



16.5 



o 



8.3 



7.85 



7.1 



5.5 



3.4 



2.3 



1.45 



13.75 

 12.65 

 10.9 

 7.85 

 5.1 

 3.75 

 2.3 



60 : 100 

 62 : 100 

 65 : 100 

 70 : 100 

 67 : 100 

 61 : 100 

 63 : 100 



Total reflection complete. 

 Partial reflection. 



64. The experiments of which I am now to state briefly the results, were 

 made with heat from various sources, and modified by transmission through 

 different media. Considering them of great importance, I have spared no pains 

 in verifying the results, and ascertaining the limits of error. My latest experi- 

 ments, in which I availed myself of the experience which earlier ones had afford- 

 ed, are of com-se most to be depended upon, and to them I shall chiefly refer 



