Manchester Memoirs, Vol xlv. (1901), AT'tf- 3. 1 1 



An examination of this table shews that both 

 Calendar's results and those deduced from Ramsey and 

 Young's experiments, give practically identical differences 

 with the results of calculations from Regnault's latent 

 heats, the percentage difference of either from Regnault 

 being given in the last column. Now we have seen that 

 the kind of steam used by Regnault requires but little 

 addition of heat in order to superheat it, and hence we 

 are forced to the conclusion that, if we accept the deduc- 

 tions from the superheated conditions as correct, Regnault's 

 latent heats must be too small by amounts varying from 

 1-33 to 3-4 per cent. It should be observed, however, 

 that the difference,between the specificvolumes of saturated 

 steam given by Callendar, or those obtained by deduction 

 from Ramsey and Young's experiments, and those given 

 by Regnault's latent heats, may be due entirely to the 

 method of deduction adopted in the determination of 

 these volumes in the first two instances, viz., by assuming 

 the data in the superheated condition to hold down to the 

 saturated condition. 



