Heat across Layers of Gas. 431 



tration of heat ; and on referring to the original memoirs, 

 first published more than thirty years ago in the Comptes 

 Rendus, and afterwards with somewhat more detail in the 

 Annales de Chimie*, I had the pleasure of finding the record 

 of two elaborate experimental investigations into what we now 

 know to have been the penetration of heat. At that time the 

 experimental results were regarded as anomalous ; and the only 

 conjecture which De la Provostaye and Desains put forward is 

 that they may in some way depend on the swiftness of convec- 

 tion currents in attenuated gases j. It is, however, easy to see 

 that no such increased swiftness as can exist will account for 

 the observed phenomena. Mr. Fitzgerald was unable to spare 

 the time necessary to follow up the subject, or he would have 

 joined me in working out this part of the present memoir ; 

 but to him is due the whole credit of having perceived the 

 importance of these observations, and to his kindness I owe 

 the advantage of having had my attention directed to them 

 and the permission to make use of them, as I now do. 



9. Dulong and Petit, experimenting with large thermo- 

 meter-bulbs placed at the centre of a hollow copper globe 30 

 centims. in diameter, blackened on the inside and kept at a 

 constant temperature, obserA^ed the rate at which the thermo- 

 meter, after having been warmed, cooled in different gases, at 

 different tensions, and with the bulb naked or coated in various 

 ways. From these experiments they obtained their well-known 

 empirical law for the escape of heat by radiation and convec- 

 tion. The expression which they give consists of two terms ; 

 of which one represents the velocity with which heat escapes 

 by radiation, and the other the velocity with which it escapes 

 by convection, or, as we shall presently see, in some cases by 

 convection and penetration. We have here no concern with 

 the first of these two terms, further than to observe that the 

 escape by radiation is the same at all tensions of the gas and 

 for all dimensions of the receiver, and depends only on the 

 character of the surfaces exposed, on 02 the temperature of the 

 copper globe, and on 6i — 0^ the excess of temperature of the 

 thermometer. For given values of ^i and 62 it was accord- 

 ingly a constant at all the tensions and with all the receivers 



* Comptes Rendus, vol. xx. (1845) p. 1767, and vol. xxii. (1846) p. 77 ; 

 Annales de Chimie, third series, vol. xvi. p. 381, and vol. xxii. p. 362. 



t De la Provostaye and Desains conclude their second memoir in the 

 following words : — "Nous ne chercherons a donner une explication com- 

 plete des difFerents faits cites dans cette communication. Nous ferons 

 remarquer seulement que le pouvoir refroidissant d'un gaz depend de sa 

 densite et de sa mobilite. Oes deux elements varient en sens inverse 

 quand on change la pression, et Ton con9oit que les eftets de ces variations 

 contraires puissent tantot s'equilibrer, tantot se surpasser dans un sens ou 

 dans Tautre.'' 



