Proceedings of the Royal Institution. 379 



experiment is illustrated by the following diagram, where a b and 





..^ju — V'u . h \ ^i .tMl..-.'d 



c d are sections of two parabolic mirrors, whose foci respectively are 

 e and/, and whose common axis is x y. It had been observed, that 

 if a hot body were placed in e, it immediately raised the temperature 

 of a thermometer at/, and it was proved that this effect was not due 

 to any direct radiation from e to /, but to the reflection of the cone 

 of heated rays egh from the mirror ab, and to their concentration by 

 a second reflection, forming the cone fc d, at whose apex is the bulb 

 of the thermometer. When this experiment was repeated, with the 

 substitution of ice at e, and when it was found that the thermometer 

 at /was depressed, much more than was due to the direct efi*ect of 

 the ice upon it along the line e/, it was inferred, that since cold is 

 reflected and concentrated by mirrors exactly like heat, cold must be an 

 independent principle or substance, and not merely a privation of heat. 

 The hypothesis of M. PrevoSt removed all difficulty from this ex- 

 periment. He supposed that all bodies in nature radiate heat in pro- 

 portion to their temperatures, just as all luminous bodies radiate light. 

 To understand, therefore, the effect produced upon the thermometer, 

 it was necessary only to observe, that when the ice was placed at e it 

 intercepted all those rays of heat from surrounding bodies comprised 

 within the cone e m n, which have previously passed through the focus, 

 and had been concentrated on the thermometer. The effiect of the 

 ice is therefore perfectly intelligible, without supposing any radiation 

 of positive cold, but merely by considering that it intercepts rays of 

 a higher, and substitutes its own of a lower temperature. 



Thus far is the theory of radiant heat generally explained, though 

 frequently with this defect, that the temperature of the body at e is 



