Ma7icJiester Memoirs, Vol. xlv. (1901), No. 3. 9 



nault are not in appreciable error, no objection can be 

 taken to its continued use. Now, for that condition of 

 saturated steam obtained by the evaporation of water in a 

 closed boiler, the steam being relieved of suspended 

 moisture as far as possible by gravitational methods only, 

 if we accept the values of the latent heat given by 

 Regnault as correct, the author's own experiments shew 

 that such steam requires very slight heating to become 

 superheated, an addition of 005 per cent, of the latent 

 heat of steam at atmospheric pressure being sufficient to 

 superheat it. 



Suppose, however, that, instead of using Regnault's 

 latent heats, we take Ramsey and Young's experimental 

 results on the densities of superheated steam, and, taking 

 the law of expansion in the steam under constant pressure 

 given by these results, and assuming this law to hold to 

 the saturation temperature, make u.se of equation (5) to 

 determine the specific volume of steam at the saturation 



temperature. As \-f] ^'^ practically constant and has 

 been already found for various pressures, 



-m. 



will give the specific volume v at saturation. 



Now the value of the specific volume at saturation so 

 obtained is in every case greater than that obtained from 

 Regnault's latent heats, the percentage difference increas- 

 ing as the pressure rises, and the only feasible explanation 

 of the difference, if we do not admit serious error in 

 Regnault's latent heats, appears to be that, near the 

 saturated condition, the law of expansion in the super- 

 heated steam differs considerably from that in highly 

 superheated steam, the alteration being probably due to 

 a change in the ultimate homogeneity in the superheated 

 steam as saturation is approached. 



