ia PROPERTIES OF STEAM AND AMMONIA 



but the agreement is satisfactory and is probably well within the limits 

 of accuracy of tin* throttling experiments. Beyond the last point the 

 HUM- begins to bend d wnward rather sharply and thus diverge from 

 tlu prolonged Davis curve. The maximum value of i" is reached at 

 about 440 degrees, whiK- the Davis equation gives the maximum at 

 about 550 degrees. 



For the lower range 32-212 degrees, values of i" calculated from 

 equation (D) show excellent agreement with the available experimental 

 values. In tin- following section on latent heat a comparison will In- 

 shown 



In the region of superheat formula (D) may be checked by the throt- 

 tling experiments of Grindley, Griessmann, and Peake. According to 

 the principles of thermodynamics a throttling process is also a constant-*' 

 process; that i>. the points obtained in any particular throttling ex- 

 ix-rinu-nt wlu-n plotted on the />/-plane should lie on a curve *'= const. 

 When the curves are superposed on the experimental points good agree- 

 nu nt i> >ho\vn. (See Fig. 9, Bulletin No. 75, Eng. Exper. Station Univ. 

 of 111.) 



Essential ly the same test may be applied in another way. The slope 



of a curve i - const, on the Tp-plane is given by the derivative [ m jr] 



and thU is the Joule-Thomson coefficient /* From equation (D) the fol- 

 lowing expression for M is readily obtained : 



I >.i\ is has computed values of M from the various throttling experiments, 

 and these may be compared with values calculated from the preceding 

 equation. Reasonably good agreement is shown. (See Bulletin No. 75, 

 Fig. 15.) 



Specific Heat of Water. Heat of Liquid. For the temperature 

 range 32-2i2F. (o-ioo C.) there are available five sets of experi- 

 ments on the variation of the specific heat of water with the temper- 

 ature. The curves that represent the results of these experiments are 

 separable into two groups having quite different characteristics. Ludin * 

 working with the method of mixtures obtained a curve which shows a 

 minimum value of c' at about 20 C., then a rapid rise to a maximum, 

 at 87 C. The curves obtained by Dieterici f and Barnes J are similar 

 in character; each shows a decrease of c' to a well defined minimum, 

 then a steady rise without any suggestion of a maximum. The experi- 

 ments of Regnault and Dieterici above 100 C. show a steady rise of the 

 specific heat with the temperature; hence, if Ludin's curve be accepted, 



* Inaug. DIM. Zurich, 1895. 



* Annalen dcr Physik (4), Vol. 16, pp. 593-620 (1905). 

 I Phfl. Trans., Vol. I99-A, pp. 55-148, 149-263 (1902). 



