362 
PHYSICS: P. W. BRIDGMAN 
Proc. N. a. S. 
TABLE I 
SUB- 
STANCE 
— X AT 30° 
XIO' 
SUB- — Xat30° 
STANCE X 107 
X 105 X 104 
Al 
Ca 
Ce 
Co 
Cu 
Ge 
Au 
Fe 
Pb 
Li 
Mg 
Mo 
13.43 
56.97 
35.74 
5.39 
7.19 
13.78 
5.77 
5.87 
23.73 
86.92 
29.60 
3.47 
0.74 
1.66 
1.06 
0.78 
0.72 
0.98 
1.08 
0.72 
1.46 
2.24 
1.36 
0.72 
—2.3 
2.3 
5.6 
7.2 
2.8 
0.6 
5.4 
6.0 
0.4 
3.0 
4.6 
Ni 
Pd 
Pt 
K 
Ag 
Na 
Sr 
Ta 
W 
Tl* 
Urf 
5.29 
5.28 
3.60 
356.5 
9.87 
156.2 
81.87 
4.79 
2.93 
34.2 
9.66 
0.802 
0.80 
1.00 
16.1 
0.90 
3.8 
1.78 
0.10 
1.02 
3.36 
0.52 
—2.5 
2.5 
1.3 
3.7 
35.1 
3.8 
22. 
2.2 
1.0 
1.5 
12.2 
* The crystal system of Tl has not yet been determined. The compressibility listed 
above was determined from a single specimen and is very materially higher than the 
value of other observers. It is therefore probable that the crystal system of Tl is not 
cubic. 
t Crystal system of Ur not yet determined. 
ments were made at two temperatures, 30 ° and 75 °. The relation between 
volume and pressure of all these metals, except Na, K, and Tl, could be 
represented within the accuracy of the experiments by a second degree equa- 
tion in the pressure. For the three exceptional metals, however, the rela- 
tion is more complicated, and the coefficients listed in table I are only the 
initial values at atmospheric pressure. Of the above list of metals, the 
compressibility of Ge, Sr, and Ur seems never to have been measured, 
even at low pressures. 
It is to be noticed that the order of magnitude of the relative change 
of compressibility with pressure or temperature is the same for all the 
metals. The compressibility of every metal decreases numerically with 
rising pressure, and in nearly every case increases with rising tempera- 
ture. There are two apparent exceptions to this last statement in the 
cases of Ge and Ur. It is possible that the result found for Ur may be 
due to experimental error, but this explanation seems less likely in the case 
of Ge. This metal is known to be abnormal in other respects. 
If the metal does not crystallize in a system whose compressibility is the 
same in all directions, then a complete description of the distortion under 
pressure requires that measurements be made of the linear compressibility 
in a number of directions of orientations known with respect to the crys- 
tal structure. This elaborate program I have not as yet been able to carry 
through, but I have been able to establish in several cases that there are 
very considerable differences of compressibility in different directions. 
These differences are shown by the fact that measurements on different 
