THE THERMAL PROPERTIES OF STEAM 13 



the specific heat after reaching its maximum at 87 degrees must diminish 



and thru increase again. It U <iitn< ult to account for such a variation 



on any rational basis, and th< . nrves of Barnes and Dieterict should be 



I. mini's curve. Davis attached no weight whatever t- 



l.n.lin's values and adopted a curve lying between those of Barnes and 



uitli Barnes's values given double weight However, the 



question is again complicated by the experiments of W. R. and W. . 



Bousficldf ^hi< h reproduce I.mlin's results, although tin method em- 



ployed uith a vacuum-jacket calorimeter) was <n; 



in I.u.im'- method of mixtures. Finally Callendar J has 



;liio\v li^ht mi tin- Mibject by a set of experiments in 



h a new and very a. . ur.itr meihud was employed. CaUendar's 



paper contains an MM- and valuable discussion of the whole 



>nl>i< 



I h. in. thods used by Barnes and Callendar, respectively, have the 

 marked advantage of being continuous. In tin- Barnes experiments a 

 steady current of water was heated through a small range of temperature 

 by an electric current, and the result obtained was therefore the a 

 specitii h -at at a pre-determined temperature rather than the mean 

 specific heat over a considerable range. Callendar used a continuous- 

 mixture method in which two steady currents of water at different tem- 

 peraturv- \\.-iv passed through a system of concentric tubes which con- 

 stituted a heat exchanger. The continuous-flow methods have obvious 

 advantages over other nut hods. The water equivalent of the calorimeter 

 is not required, and various corrections that involve uncertain measure- 

 ment- are eliminated. 



The results of Calendar's experiments l>\ the continuous-mixture 

 method completely verify tin- earlier experiments of Barnes by the con- 

 tinuous-elect ric meth>d. As these two independent methods are much 

 superior to the other methods used and give identical results, there can 

 be no question that these result^ should be accept tt I. 



Taking the specific heat of water at 20 C. as unity. Callendar gives 

 the following equation (Off th. \ariation of the specific heat with tempera- 

 lure 



c'- o. 



From the specific heat c' the heat content / of the liquid is derived by the 

 relation 



s*. 



e-f 



Alter changing from C. to F. temperatures and applying a factor to 



Sceam Table, and DUgrmnu. p. 89. 



t PhiL Tram.. Vol. an-A, pp. 199-251 (9"). 



| PhiL Trans.. Vol. 2I2-A. pp. I-J2 (1913)- 



