216 Dr. A. Naumann on the Specific Heat of Gases 



this rule is afterwards carried out, it must apply in general to 

 such gases as are formed from their elementary constituents 

 without condensation, and whose molecules consequently con- 

 sist, according to the usual assumption, of two atoms. For 

 every additional atom in the molecule, however, it must give too 



great a value for the specific heat by - =- * =0*051; for, 



by the rule in question, half the heat of molecular motion (which 

 is a constant quantity and independent of the number of atoms 

 in a molecule) is reckoned for each atom that enters into the 

 composition of the molecule. (In the case of gases which con- 

 tain mercury, arsenic, or any other body of the same kind, the 

 excess of the calculated value would bear a different but easily 

 deducible ratio to the number of elementary atoms contained in 

 one molecule.) It is doubtless to be attributed to this property 

 of BufFs rule (namely, that it gives too great a value for the 

 specific heat, except in the cases already specified, and one which 

 is so much the more in excess in proportion as the number of 

 atoms in a molecule, and therefore also in general the degree of 

 departure from the laws of Mariotte and Gay-Lussac, is greater), 

 that the values calculated by him agree with those observed by 

 Regnault, and sometimes with surprising closeness, exactly for 

 gases which are far from following the laws of Mariotte and 

 Gay-Lussac, whereas this agreement no longer shows itself in the 

 case of gases which follow these laws more exactly and at the 

 same time contain more than two atoms in a molecule ; for such 

 gases the calculated values, in accordance with the above consi- 

 derations, come out too high. 



A method of calculating the specific heat of gases, which agrees 

 in essential points with Buffs method, has also been proposed 

 by Clausius*, to which of course the same remarks apply as have 

 been made in regard to BufPs method. Clausius comes at last 

 to the conclusion that the imperfection of the gaseous state is 

 " certainly not quite sufficient " to account for the difference be- 

 tween Regnault's values and those which he calculated. I believe 

 that I have ascertained the true reason of the discrepancy. The 

 theoretical specific heat ought not, by the nature of the case, 

 ever to be notably higher than the actual specific heat ; and its 



thesis that the specific heats of equal weights at constant volume are in- 

 dependent of alterations in the density of the gases, even when these alte- 

 rations are caused by chemical action. But between mechanical and che- 

 mical condensation there is, however, this essential difference — that the 

 former is not accompanied by any change in the number of molecules, 

 whereas the latter is always accompanied by an alteration, namely a dimi- 

 nution of the number. 



* Ann. der Chem. und Pharm. vol. cxviii. pp. 112 et seq. 



