BRIDGMAN. — THERMODYNAMIC PROPERTIES OF LIQUIDS. 11 



and the coil, mercury cup, and the liquid under investigation were 

 inserted together into the cylinder. This naturally had to be done 

 from below, and for this purpose the cylinder was held vertical during 

 this operation in a special vise, projecting at right angles from the wall. 

 The packing and the retaining screw were also inserted while the 

 cylinder was in this position. The cylinder was now filled with a 

 weighed quantity of kerosene through the open upper end. This 

 filling with kerosene was accomplished in two operations; first, 

 half of the kerosene was poured in, and all air extracted from the 

 lower part of the cylinder by gently exliausting it, and then the re- 

 mainder of the kerosene was poured in. The moveable plug was then 

 inserted, with special care not to spill any kerosene, and finalh' the 

 cylinder, kept always upright, was put in position in the lower part of 

 the press. After a run the bulb was cleaned for the next run by 

 heating it nearly to redness. That the cleaning was thorough was 

 shown by the constancy of weight of the bulb, which seldom varied 

 more than one milligram from experiment to experiment. 



The essential idea of the experimental method remains the same as 

 in the work on water, but there are slight modifications. The purpose 

 is to obtain by direct measurements the change of volume with pres- 

 sure over the entire pressure range at some one constant temperature, 

 and then by another independent set of measurements to obtain the 

 change of volume with temperature over the entire temperature range 

 at a sufficient number of pressures to completely cover the field. 

 This method has a distinct advantage over that employed by Amagat, 

 for example. Amagat ol)tained the volume as a function of pressure 

 and temperature by measuring the change of volume with pressure 

 isothermally at a number of temperatures. This has the disadvantage 

 that the thermal dilatation can be found only by taking the difference 

 of compressibility determinations at different temperatures. Obvi- 

 ously this makes it necessary to measure the compressibility with a 

 much greater degree of accuracy than the desired final accuracy of 

 the dilatation, and this is a difficult matter to accomplish. But the 

 present method gives both compressibility and dilatation with the 

 accuracy to be expected of direct measurements. 



The procedure in measuring the change of volume with pressure 

 at constant temperature was the same as in the work on water. The 

 temperature chosen for this determination was 40°, because it is 

 the same as was used for water. Reference is made to the former 

 paper for the details of manipulation. In order to cover completely 

 the pressure range, it was necessary to make the readings in two series, 



