Chemistry and Physics. 451 



Wherea?, if the latent water exists, it must have existed in the 

 steam used by Regnault, and the steam tables must also be sub- 

 ject to identical corrections ; and, consequently, the percentage of 

 theoretical performance of steam engines would be unchanged. 



It is then pointed out, that, in the reduction of such of these 

 results as have been published, use has been made of Regnault's 

 determination of the specific heat at constant pressure of steam 

 gas (0-48) in a manner which is not consistent with the theory of 

 thermodynamics. Thus, in Rankine's notation, S^ is the weight 

 of steam per lb. of flui<l, and H^ the total heat per lb. from 0° C. 

 to Tj°, Aj the heat required to raise water per lb., and H^, A^, T^, 

 the corresponding values for saturated steam at the pressure 

 after wire-drawing, and T° the observed temperature after wire- 

 drawing. 



The notation assumed for the equation of heat, neglecting 

 incidental losses, is 



S, (H,-AJ+A, = H,-f 0-48 (T3°-T,°) .... 1 

 Whereas, it has been proved by Rankine that the thermody- 

 namic expression for the total heat in superheated steam at T° O., 

 provided it has reached the condition of steam gas, to which the 

 0-48 only applies, is 



C^ + 0-48 (T,°-T„°) 

 Cj, bein^ a constant, depends only on the temperature of the 

 water {T°) from which tlie steam is produced, the value of which 

 from 0** C. is 606-7, approximately, as deduced by Rankine. 



Using Regnault's formula for H^, the right member of equation 

 (1) becomes 



606-5 + -805 T/ + 0-48 (T3°-To°) 



while the value by the thermodynamic formula is 



606-7 + 0-48 T,° 

 which gives as the excess of heat over that assumed 



0-2 + 0-175 T/ 



This excess, if T^ were 100° C, is 17*7 thermal units, and if the 

 initial steam pressure were 200 lbs. above the atmosphere, the 

 latent heat being 467*5 thermal units, the percentage of water it 

 would evaporate, at boiling point, is 



4b7o 

 which is about as much as needs to be accounted for. 



It is also shown that, in order to render Rankine's formula 

 applicable to wire-drawing experiments, it is necessary that the 

 wire-drawing should be continued till the steam is gaseous, 

 whence arises the difficulty of securing that this state has been 

 reached. This, however, may be secured by lowering the pres- 

 sure gradually after wire-drawing, and so increasing the extent of 

 wire-drawing while observing the temperature (Ts°), which, after 

 falling, will gradually become constant as the wire-drawing 



