"Specific Heat of Saturated Ether Vapour." 297 



value of the integral calculated by means of Regnault's 

 formula for c p was supposed to give the heat of vaporization v. 

 On the other hand, the values of v which may be regarded 

 as due to Ramsay and Young and to Battelli are more 

 directly obtained, namely, through measurements of tempe- 

 rature, pressure, and volume. At this point, and nowhere 

 else, the process of calculation followed by Clausius differs 

 from the present, and the values of the temperature-coefficient 



-jm are the sole cause of the discrepancy. 



Of course I do not here assert that the alleged sign of h is 

 false, but simply that the principles of thermodynamics 

 together with certain justifiable assumptions 



H = c 



and e p =f(T), 



and certain experimental results 



c p =O52901 + 0-0005916 .t (Regnault), 

 and 



v= 93-27 -0-256 . t (Ramsay and Young), 

 or 



u=95-70-0'3033.* (Battelli), 



will afford us as much cause for suspicion as for confidence. 

 This is all that an application of thermodynamics can do ; 

 we must appeal to experiments to know what cannot be 

 denied or what can be asserted, and I may remark that there 

 are three ways of experimenting in regard to the question 

 under discussion : — the method of Hirn and Cazin, the method 

 of determination of v (either calorimetrically or by measuring- 

 volumes, pressures, and temperatures), and the method of 

 Regnault. 



The present discussion is incomplete in so far as it leaves 

 Regnault's formula for c p untouched. Regnault's values of 

 c p may possibly turn out too small, and, when corrected, may 

 bring about the positive sign of h at ordinary temperatures 

 even in conjunction with Ramsay and Young's values or 

 Battelli's of v. Unfortunately there are known, besides 

 Regnault's formula, only that which is ascribed to Hirn, and 

 this does not agree well with Regnault's. Hirn's values, 

 though somewhat larger than Regnault's, are not sufficient 

 to cancel the reversal of sign under discussion. 



