14 Conduction of Heat through Rarefied Gases. 



The quantity of heat lost by the lower plate is 



Q=— j~r [0 2 n 2 ' €2 + 0^^ — 2 n 2 c 2 — 0in l c 1 ']= ~j— (# 2 — #i)/Vi. 



V 07T ^/ b7T 



Now equations (1) and (4) give 



n c<> 



2 



M-i-V 



so we have 



_ 2 f mm c lCs 

 H v /6tt(2-/ , )c 1 + ( 



9 



If we put c= ~~ , and /=1 — /?, we get finally 



Q V^r^ (0) 



This is the exact value for the conduction of heat in highly 

 rarefied gases ; we see that it is greater than the value calcu- 

 lated before, and all the numbers given by Messrs. Soddy 



■XT 



ancT Berry for ^ ought to be multiplied by the factor 

 7t =0*7236, which gives the series : 



A. Ne. N 2 . 2 , CO. N 2 0. 2 H 2 . 00 2 . CH 4 . He. H 2 . 

 ^0-79 075 0-G8 62 059 OoG 0*52 0'52 0-49 0-37 0-18 



It proves that the coefficient ft is never to be neglected, 

 that is to say, that the heat interchange of the molecules on 

 impact is always imperfect. The order of gases suggests the 

 rule that the interchange of energy is worse for the lighter 

 molecules and for the diatomic and polyatomic ones in com- 

 parison with the heavier and monatomic ones. The first part 

 of this rule is to be explained by a simple mechanical reason- 

 ing, showing the interchange of energy between colliding 

 spheres to be the more imperfect, the greater the difference 

 of their masses (here we have the molecules of the gas 

 colliding with the heavy platinum molecules). Also the 

 second part of this rule seems to be in accordance with other 

 phenomena of conduction of heat, showing intramolecular 

 energy being comparatively less disposed to equalization by 

 impacts than energy of progressive motion. 



Leniberg University. 



