MR. J. H. BRINKWORTH ON SPECIFIC HEAT OF STEAM. 385 



where H, is the total heat of the saturated steam at the initial temperature and 

 pressure, and S is the mean specific heat at atmospheric pressure from 100 C. to t", 

 the temperature observed after throttling to atmospheric pressure. The curve of 

 variation of total heat found in this way, instead of being straight like REGXAULT'S, 

 with a constant coefficient dRfdt = 0'305 for saturated steam, showed a decided 

 curvature, the rate of increase falling from 0'40 calorie per 1 C. at 100 C. to 0'30 at 

 160 C. This indicated that REGNATJ L,T'S coefficient 0'305 might be a fair average of 

 the rate of variation over the range 100 C. to 200 C., where his experiments were 

 most concordant, but that the coefficient probably increased at temperatures below 

 100 C., approximating to the value 0'475 (as it should according to RANKIXE'S theory) 

 at low pressures and temperatures, where the vapour should behave as a perfect gas. 

 This view, while disagreeing materially with REGNATJLT'S observations of the total 

 heat at low temperatures, Avhicli were very discordant, gave good agreement with 

 those of DIETERICI at C., and of GRIFFITHS at 30 C. and 40 C. But in order to 

 determine the complete variation of the total heat, H, from the thermodynamical 



formula 



dR = 8 fZT-SC rfP, ......... (3) 



and to deduce the form of the characteristic equation from the relation, 



V, ......... (4) 



it was first necessary to determine more completely the variation of the cooling effect, 

 C, over as wide, a range of T and P as possible. 



The Differential Throttling Calorimeter. 



The single throttling calorimeter previously employed was well adapted for 

 determining the variation of the total heat of saturated steam in terms of the specific 

 heat at atmospheric pressure, but it was not well suited for obtaining values of the 

 cooling effect, C, at a particular point. Values of C at a particular temperature and 

 pressure could be obtained only by successive observations with the single calorimeter, 

 during which the initial state of the steam might vary. Since a change of 1 per cent. 

 in the wetness of the initial steam produced a change of about 10 C. in the observed 

 temperature after throttling, it was most important to employ a differential method 

 in which such effects were automatically eliminated. This was successfully accom- 

 plished by connecting two exactly similar calorimeters to the same steam supply, and 

 adjusting the terminal pressures P' and P" by exit throttles to a suitable difference 

 P' P" read on a differential gauge. The corresponding difference of temperature 

 T' T" could be obtained by a single reading with a pair of differential platinum 

 thermometers, from which the value of the cooling effect at the mean temperature 

 and pressure, C = (T'-T")/(P'-P"), is directly obtained, with almost complete 



3 F 2 



