354 PROF. C. FEE WEN JENKIN AND MR. D. R. PYE ON THE 



INTRODUCTION. 



L\ a former paper (l) the authors described a series of measurements on the thermal 

 properties of C0 3 and the construction of a #</> chart embodying those results. It 

 was pointed out that the superheated area of the chart was incomplete, the 

 constant pressure lines being only approximately accurate and the total heat (I) lines 

 omitted. It was also pointed out that throttling experiments on the superheated 

 gas would form a valuable check on the accuracy of the chart, and the construc- 

 tion of an L/> chart was postponed till more accurate measurements should have 

 been made on the superheated gas and the whole had been checked by throttling 

 experiments. 



The present paper describes the additional experiments required to complete and 

 check the 9</> chart, including a re-measurement of the total heat of the liquid for 

 which some extrapolated values had been used before, and finally the construction of 

 the L/> chart. This chart lias been constructed graphically, as the 6(j> chart was, 

 directly from the observed data and its accuracy checked in various ways by tliermo- 

 dynamic equations. These equations apply quite generally to all Irf> charts and are 

 independent of the particular properties of carbonic acid. 



The value of the !</> chart in all calculations connected with refrigeration was 

 originally pointed out by MOLLIKR, and has recently been emphasized in the Report of 

 the Research Committee of the Institution of Mechanical Engineers (1914). 



The authors hope that the new chart may be of practical use, as it extends and 

 corrects the original one prepared by MOLUKK. 



The experiments were made, as before, in the Engineering Laboratory at Oxford. 

 The apparatus employed was generally similar to that used in the former experiments, 

 but all the heat measurements were made in anew calorimeter, no;. 1, which eliminates 



' O J 



radiation and conduction losses. 



The measurements of the total heat of the gas for each pressure were made at 

 several temperatures, so that the variation of specific heat with temperature was 

 determined. The values of the specific heat found by these experiments were used to 

 plot the new constant-pressure lines on the 9<p chart. The method used for plotting 

 these lines is explained in Appendix I. The re-measurement of the total heat of 

 liquid C0 2 showed, as was expected, that the values previously used required 

 correction ; the liquid-limit curve of the 6<f> chart was corrected accordingly. The 

 throttle experiments, which provide an independent check on both the gas-limit curve 

 and constant-pressure curves, confirmed the accuracy of the pressure curves and of 

 the limit curve above -8 C., but showed that a small correction was required 

 below that temperature. This correction having been made, the chart was completed 

 by drawing the I lines in the superheated area, (See Appendix II., fig. 6.) The 

 combined corrections of the two limit curves leave the values of the latent heat 

 practically unaltered ; the alteration at -50 C. is only 0'3 per cent. 



