SECT, xxv.] TRANSMISSION OF HEAT. 241 



from an argand lamp, but stop eight or nine-tenths of the 

 concomitant heat ; whilst a large piece of brown rock- 

 crystal gives a free passage to the radiant heat, but inter- 

 cepts almost all the light. M. Melloni has established the 

 general law in uncrystallized substances such as glass and 

 liquids, that the property of instantaneously transmitting 

 heat is in proportion to their refractive powers. The law, 

 however, is entirely at fault in bodies of a crystalline tex- 

 ture. Carbonate of lead, for instance, which is colourless, 

 and possesses a very high refractive power with regard to 

 light, transmits less radiant heat than Iceland spar or rock- 

 crystal, which are very inferior to it in the order of refran- 

 gibility ; whilst rock-salt, which has the same transparency 

 and refractive power with alum and citric acid, transmits 

 six or eight times as much caloric. This remarkable dif- 

 ference in the transmissive power of substances having the 

 same appearance is attributed by M. Melloni to their crys- 

 talline form, and not to the chemical composition of their 

 molecules, as the following experiments prove. A block 

 of common salt cut into plates entirely excludes calorific 

 radiation ; yet, when dissolved in water, it increases the 

 transmissive power of that liquid : moreover, the trans- 

 missive power of water is increased in nearly the same 

 degree, whether salt or alum be dissolved in it ; yet these 

 two substances transmit very different quantities of heat 

 in their solid state. Notwithstanding the influence of crys- 

 tallization on the transmissive power of bodies, no relation 

 has been traced between that power and the crystalline 

 form. 



The transmission of radiant heat is analogous to that of 

 light through coloured media. When common white light, 

 consisting of blue, yellow, and red rays, passes through a 

 red liquid, almost all the blue and yellow rays, and a few 

 of the red, are intercepted by the first layer of the fluid ; 

 fewer are intercepted by the second, still less by the third, 

 and so on : till at last the losses become very small and in- 



R 



