796 



TRUEBLOOD. 



ing the difference in the average apparent Joule-Thomson coefficients 

 over identical ranges of flow to the effect of pressure. The three 

 values of {dix/dp)i obtained in this way are: 



+0.03G °C. om.Vkgm.^ for the overlapping of the 5.84 



and the 4.09 kgm./em.- lines 

 +0.028 " " " of the 4.09 



and the 2.30 kgni./cni.- lines 

 +0.024 " " " of the 5.84 



and the 2.30 kgm./cm.^ lines 



Taking the mean of the first two and the mean of this mean with the 

 third, there results + 0.02S °C. cm.Vkgm.- as the value of {dn/dp)t. 

 This is of course only a very rough approximation, and is probably an 

 upper limit. 



|>3.5 



E 





Figure 17. 



0.02 



1 



0.04 



0.06 



0.08 0.10 



Reciprocal of now. hn/kgm 



IX. VS. plots for three sets of runs at three different mean 



Numbers 



pressures, plug VI. Mean temperature, 165° C. in each case 

 attached to lines indicate average mean pressure, kgm./cm.^ 



In a paper on the properties of saturated and superheated ammonia 

 vapor, Goodenough and Mosher ^^ have re-examined the data dis- 

 cussed by Davis, in an effort to verify an empirical equation of state for 

 ammonia vapor by comparing reduced values of /x as calculated from 

 it with the reduced observed values used in Davis' paper. For this 

 purpose they recomputed Davis' reduced results, using later critical 

 data for water than DaA'is employed. In this way they were able to 

 show that their calculated /x curves for reduced pressures correspond- 

 ing to ammonia pressures of 4, 5, 7 and 9 atmospheres agreed well 

 with the observed results of which Davis made use, when these results 

 were grouped in such a way that the mean reduced pressure for each 

 group was approximately the same as that of the curve with which 

 comparison was made. Inferior agreement was obtained in the case 



12 Goodenough and Mosher, Bulletin No. 66, Univ. of Illinois Engineering 

 Experiment Station, January 27, 1913. 



