On a Suggestion as to the Constitution of Chlorine. 35 



3*17 feet less than the experimental value 1089*42 feet deduced 

 by the same author from a large number of observations. 

 For these reasons I consider the theoretical value of the velo- 

 city of sound to be 



as deduced exclusively from hydrodynamical principles, such 

 as I have denned them to be in the foregoing communication. 



J. Challis. 



IV . On a Suggestion as to the Constitution of Chlorine, offered by 

 the Dynamical Theory of Gases. By A. W. RtJCKER, M.A., 

 Professor of Physics in the Yorkshire College, Leeds*. 



IF a gas of density 8 consists of molecules each of which 

 possesses m degrees of freedom, and if also the intermo- 

 lecular forces are negligible, the specific heats at constant 

 pressure (c p ) and at constant volume (c v ) are connected by 

 the two well-known equations 



(c p -c v )S = -0694:, (1) 



^-=1+J-, (2) 



c v m + e v y 



where e is a quantity which depends upon the potential energy 

 of a molecule. Hence, if c p is given by experiment, c v can be 

 calculated from the first of these equations ; and then m + e is 

 known from the second. 



The accuracy of the value of w. + e thus deduced will depend 

 upon that of c p , and on the legitimacy of the application of the 

 two equations to the gas or vapour under consideration. 



With respect to the first of these points, it may be remarked 

 that E. Wiedemann has recently (Fogg. Ann. Bd. clvii. p. 1, 

 1876, and Wied. Ann. Bd. ii. p. 195, 1877) determined the 

 specific heats at constant pressure of 14 out of the 35 gases 

 and vapours studied by Regnault. The difference between 

 the results of the two investigators amounts in two cases only 

 (ethylene and ammonia) to 6 per cent. ; in three cases it is 

 about 5 per cent., and in all the others less. Thus even on 

 the assumption that the later experiments are absolutely cor- 

 rect, it follows that Begnault's numbers may be trusted to 6 

 per cent. His results, however, can only be taken as true for 

 the particular temperatures at which the experiments were 

 made, as Wiedemann shows that in all but the most perfect 



* Communicated by the Physical Society, having been read on No- 

 vember 22nd. 



D2f 



