Prof. Thomson on the Dynamical Theory of Heat. 109 



(1.) The rate of variation with the temperature, of the pres- 

 sure of saturated steam. 



(2.) The latent heat of a given weight of saturated steam. 



(3.) The volume of a given weight of saturated steam. 



(4.) The volume of a given weight of water. 



The last mentioned of these elements may, on account of the 

 manner in which it enters the formula, be taken as constant, 

 without producing any appreciable effect on the probable accu- 

 racy of the result. 



32. Jttegiiault^s observations have supplied the first of the 

 data with very great accui-acy for all temperatures between — 32° 

 Cent, and 230°. 



33. As regards the second of the data, it must be remarked 

 that all experimenters, from Watt, who first made experiments 

 on the subject, to Regnault, whose determinations are the most 

 accurate and extensive that have yet been made, appear to have 

 either exphcitly or tacitly assumed the same principle as that of 

 Caraot which is overturned by the dynamical theory of heat ; 

 inasmuch as they have defijied the " total heat of steam " as the 

 quantity of heat required, to convert a unit of weight of water 

 at 0°, into steam in the particidar state considered. Thus Reg- 

 nault, setting out with this definition for " the total heat of satu- 

 rated steam," gives experimental determinations of it for the 

 entire range of temperatures from 0° to 230° ; and he deduces 

 the " latent heat of saturated steam " at any temperatm'e, from 

 the "total heat," so determined, by subtracting from it the 

 quantity of heat necessarj^ to raise the liquid to that tempera- 

 ture. Now, according to the dynamical theory, the quantity of 

 heat expressed by the preceding definition depends on the manner 

 (which may be infinitely varied) in which the specified change 

 of state is effected ; differing in different cases by the thermal 

 equivaleuts of the differences of the external mechanical effect 

 produced in the expansion. For instance, the final quantity of 

 heat required to evaporate a quantity of water at 0°, and then, 

 keeping it always in the state of saturated vapour*, bring it to 

 the temperature 100°, cannot be so much as three-fourths of the 

 quantity required, first, to raise the temperature of the liquid to 



* See below (Part III. § 58), where the "negative" specific heat of 

 saturated steam is investigated. If the mean value of this quantity between 

 0° and 100" were — Vb (and it cannot ditfermuch from this) there would be 

 150 units of heat emitted by a pound of saturated vapour in having its tem- 

 perature raised (by compression) from 0° to 100°. The latent heat of the 

 vapour at 0° being f)0()'5, the final (juantity of heat required to convert a 

 pound of «ater at 0° into saturated steam at 100", in the first of the ways 

 mentioned in the te.\t, would consecjuently be 45(!'5, which is only about 

 ^ of the quantity G'.il found as " the total heat " of the satiu-ated vapour at 

 100°, by kegnault. 



