90 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



may be at low pressures, there is absolutely no ground for the assump- 

 tion that the coefficient remains independent of the temperature. 

 This was shown to be the fact in the previous paper on water, but the 

 argument lost force because water is abnormal. However, for none 

 of the twelve liquids of this paper is the relation even approximately 



1.10 L05 1.00 .95 90 .85 .80 .75 

 Volume 

 Carbon Bisulphide^ 



Figure 46. The pressure coefficient of carbon bisulphide plotted against 

 volume. The diagram shows that at high pressures the pressm-e coefficient 

 is a function of temperature as well as of volume. 



satisfied. It is not necessary to show curves for all twelve liquids 

 in order to disprove this one point. The data are at hand so that any 

 one may make a complete test for himself. A single diagram, chosen 

 at random, is sufficient to show that the proposed relation breaks 

 down completely, as indeed one would expect it to in view of the 

 complications of compressibility and dilatation. Figure 46 shows 

 this for carbon bisulphide. It speaks for itself. 



Work of Compression. — The mechanical work of compres- 

 sion is shown in Folder HI; Figures 47 to 58 are for the liquids 

 separately, and Figure 59 shows the average between 20° and 80° of 

 all twelve liquids together. The difference of the work for different 

 temperatures is so nearly the same that the differences could not have 

 been read accurately directly from the curves. The course was 

 adopted, therefore, of plotting the work at 40° only, and in the lower 



