38 -PROCEEDINGS OF THE AMERICAN ACADEMY. 



accuracy to be expected for each liquid (because for some liquids the 

 accuracy is greater than for others), and also the sources and the 

 numerical values of the results of other observers which have been 

 used in computing the results given here. The results taken from 

 other work are the density at atmospheric pressure, the thermal 

 dilatation, the initial compressibility for low pressure ranges, and the 

 specific heats at atmospheric pressure. Unless otherwise specified, 

 the values for the density at atmospheric pressure have been taken 

 from the recent tables of Kaye and Laby, and the values for the 

 thermal dilatation have been deduced from the tables of Landolt 

 and Bornstein. In these tables the volume at any temperature, is 

 given in terms of the volume at 0° by a power series of the form 

 Vt = Vq {1 + at + bf + cf). In reproducing this expression it will 

 not be necessary to repeat the formula each time, but merely to give 

 the values of the three constants a, b, and c. 



It has been mentioned on page 22 that in computing tlie changes 

 of volume with pressure at 40°, it was found that beyond 500 kgm. 

 the shape of the curves was nearly the same for all twelve liquids, 

 the only difference being in the numerical magnitudes. The con- 

 stants used in the general pressure-volume formula of page 22 for the 

 average of the twelve liquids were as follows; a = — 0.0029, 

 iS = -0.0546,7 = +0.2969, and 5 = -0.1804. To pass from this 

 general formula to any one of the twelve liquids each of these four 

 constants is to be multiplied by the same factor. This factor will 

 be given in the following under the name of the "reduction factor." 



The discussion is to be one of merely the numerical details of the 

 measurements and the computations. The discussion of the general 

 character of the results and their significance will be reserved until 

 the data have all been presented. 



Methyl Alcohol. — Three sets of measurements were made 

 on this substance with three different fillings of the apparatus, the 

 last being separated by nearly three months from the earlier two. 

 The first measurement was of the thermal dilatation and compres- 

 sibility at low pressures with the larger bulb adapted for low pressure 

 work. The next set of measurements, made immediately afterwards, 

 was of the compressibility and dilatation at the higher pressures 

 with the smaller bulb for the high pressure work. The third measure- 

 ment was with the high pressure bulb, and included the compres- 

 sibility and dilatation over the entire pressure range, both high and 

 low pressures. The measurements at low pressures were made, as 

 already explained, before the piston had been upset by the higher 



