636 ' LECTUBE LI. 



between cliiFerent portions of the same fluid, almost entirely by the mixture 

 of their particles: hence a fluid heated on its surface transmits the lieat very 

 slowly downwards, since the parts which are first lieated, being rendered spe- 

 cifically lighter, retain their situation above the colder and heavier parts; 

 while, on the contrary, any cause of heat, applied at the bottom of a vessel, 

 very soon reduces all its contents to a uniform temperature. It appears also, 

 from some late experiments, that the immediate transmission of heat within 

 the internal parts of solids is much slower than has commonly been supposed; 

 and it has been found almost impossible to keep a thermometer, at the centre 

 of a large and solid globe of metal, at the same temperature with that of its 

 superficial parts. 



Besides the communication of heat by contact, it is usually, if not always, 

 emitted from the surfaces of bodies in the form of radiant heat, which is 

 thrown off in all directions, wherever it meets no obstacle from a substance 

 impervious to it, and is propagated nearly in the same manner as light, and 

 probably with the same velocity, without producing any permanent effect on 

 the temperature of the medium transmitting it. Thus, a thermometer, sus- 

 pended by a fine thread under the receiver of an air pump, or in the Torricel- 

 lian vacuum, will continue to vary its temperature with that of the surround- 

 ing bodies: and in this case the whole of the heat must be communicated by 

 radiation. Mr. Leslie has discovered that the quantity of heat thus emitted 

 depends not only on the temperature, but also on the nature of the surface 

 concerned, a polished surface of tin emitting only -r^, or less than one 

 eighth part as much, as the same surface blackened. A surface of tin scraped 

 with a file in one direction has its powers of radiation more than doubled; 

 but by crossing the scratches, they are reduced nearly to their original state; 

 and a coating of isinglass, resin, or writing papers, or a glossy surface of any 

 kind, produces an eft'ect nearly approaching to that of black paint. This radia- 

 tion from a heated surface, like that of light, takes place in almost equal degrees 

 ■in every direction; and its magnitude is nearly independent of the nature of 

 the fluid in contact with the surface, provided however that it be an elastic 

 fluid ; for water does not seem to transmit every kind of radiant heat with 

 freedom. It appears that the radiant heat emitted by a surface of glass, or of 

 black paint, is about one third greater than that which is at the same time 

 carried off by the atmospheric air; but that the radiation from a metallic sur- 



