BRIDGMAN. THERMODYNAMIC PROPERTIES OF LIQUIDS. 49 



tion over the high pressure range, and the second the compressibiHty 

 and dilatation over both the high and low pressure ranges. The use of 

 isobutyl instead of normal butyl alcohol was not intended. In order- 

 ing the chemicals, normal butyl was not specified, and it was not 

 noticed that the substance sent w^as isobutyl until all the preparations 

 had been made for a run. This substance has the disadvantage, 

 of not being one of the same series as the four other alcohols. How- 

 ever, it makes little difference so far as the comparison of the results 

 with those of Amagat is concerned, for Amagat did not work with 

 either normal- or iso-butyl alcohol. Furthermore, the use of this 

 substance has proved very instructive in showing that a change in 

 the structural formula changes the properties even at high pressures. 

 It might be expected that high pressures would wipe out variations 

 due to structural differences, but such has not proved to be the case, 

 at least to 12000 kgm. 



The average discrepancy in the piston displacements of the two 

 determinations of compressibility was 0.0024 inch on a total displace- 

 ment of 2.0 inches. The mean discrepancy of the displacement for the 

 thermal dilatation for 20° was 0.0008 inch on a mean of about 0.070 

 inch. The agreement between the two sets of readings for the highest 

 temperature range, 60°-80°, was virtually perfect. 



The reduction factor for the mathematical formula was 0.9342. 



Landolt and Bornstein's tables do not contain the requisite data 

 for the volume of isobutyl alcohol at atmospheric pressure. The 

 values adopted here were obtained by Naccari and Pagliani, and are 

 apparently the only data which have been published for this liquid. 

 These authors have not expressed their results by a power series, but 

 prefer instead to give the density for a considerable number of tem- 

 peratures. By interpolation from their results the volumes at 20°, 

 40°, 60°, and 80° were found to be: 1.0195, 1.0406, 1.0625, and 1.0880. 

 From these results the value of Kaye and Laby for the density at 18° 

 is reduced to 0.8165 at 0°, against 0.8162 of Naccari and Pagliani, 

 virtual agreement. 



No measurements of the change of volume of isobutyl alcohol had 

 been made beyond a few kgm. previous to these computations, so the 

 value obtained from the low pressure determinations of the present 

 work was adopted. This was 0.0484 at 40° between 1 and 500 kgm. ; 

 not at all an unlikely value, being the same as Amagat's for ethyl 

 alcohol. Compressibility determinations of others at low pressures 

 have shown that isobutyl alcohol has a compressibility considerably 

 higher than that of normal butyl alcohol, so that we are to expect a 



