Relations of Aif, Heat, and Cold. 455 



of a solid, when broken or reduced to a gross powder, may 

 be determined. 



*' The same regulated mode of cooling will serve to de- 

 tect with precision the expenditure of heat, and to discri- 

 minate its various allotment, in the different gases. For 

 this purpose, a ball of about three inches in diameter, and 

 formed of bright and very thin silver, is preferable ; and it 

 may be succes'sively covered with a pellicle or with cambric, 

 or painted with a coat of ivory-black. Not to multiply un- 

 necessary details, it will perhaps be deemed sufficient to 

 cite the case of hydrogen gas, which is by far the most 

 distinguished. The portion of heat emitted in this ener- 

 getic species of gas by the system of pulsations, whether 

 from a vitreous or a metallic surface, if not exactly, is very 

 nearly the same as in atmospheric air; but that other por- 

 tion which is abstracted by the giadual recession of the 

 nearest heated particles of the fluid, exceeds no less than 

 four times the corresponding discharge in the ordinary 

 medium. Why such a striking difference should arise,can 

 hardly be conjectured. Hydrogen gas, though ten times 

 lighter than air, yet contains, in the same volume, an equal 

 quantity of heat; and it is fitted, by its very superior elasti- 

 city, to' transmit the pulsatory impressions more than three 

 times faster. It must, therefore, as a counterbalance, re- 

 ceive those impressions three times slower from the heated 

 surface. But if such influence be confined, as it would 

 seem most probable, to the mere boundary of the medium 

 or its thinnest conterminous shell, the measure of heat im- 

 bibed at a given rise of temperature from the attenuated ex- 

 panse, would be diminished between two and three times. 

 This mutual compensation of eflTqct nearly agrees with the 

 actual result. With respect to the quadrupled increase of 

 that portion of heat which is abstracted by the slow but con- 

 tinued renewal of the adjacent suatum of the fluid, we must 

 refer it chiefly to the very great mobility of hydrogen gas, ex- 

 ceeding three times that of common air. If these strata were 

 supposed to have in both cases the same thickness, they 

 would each of them carrv off the same share of heat. 



'* The portions of heat transmitted by pulsation through 

 hydrogen gas, from a painted and a metallic surface, being, 

 as belore, expressed respectively by 10 to 1, the other por- 

 tion, which is altogether independent of the nature of the 

 cooling surface, and is disperser! by abduction, or the inces- 

 sant retreat of the strata of the fluid as they come to be 

 successively affected, will amount to 40. Under like cir- 

 cumstances, therefore, the whole expenditure of heat from 

 F f 4 a painted 



