BRIDGMAN. — THERMODYNAMIC PROPERTIES OF LIQUIDS. 5 



mental evidence, has been made the basis of a recent empirical theory 

 of liquids by Tammann,^ who assumes that at high pressures varia- 

 tions in thermal expansion due to changes of temperature ought to 

 become vanishingly small. Such, however, is by no means the case, 

 but, on the contrary, at high pressures the thermal expansion varies 

 with temperature in a more complicated way than at atmospheric 

 pressure. 



The irregularity of the e'ffects at high pressures makes it evident 

 that a complete theory of liquids must be very complicated. The first 

 step, therefore, toward a theory would be to explain only the general 

 features. With this in mind, the average of the various thermo- 

 dynamic properties over the entire temperature range has been com- 

 puted for each liquid. To facilitate comparison, the average of any 

 one property, compressibility for example, is shown on the same 

 diagram for all twelve liquids. One would expect to turn at first 

 to these collected diagrams in seeking light on a theory of liquids. 



In this paper no attempt has been made at the very consideral)le 

 task of developing a quantitative theory to represent the data. 

 The results suggest very strongly in some cases, however, that con- 

 ceptions of the mechanism of a liquid which may be adequate at low 

 pressures can no longer be adequate at high pressures. For instance, 

 we shall probably have to modify our ideas of the mechanism account- 

 ing for pressure and temperature. Some discussion is given of modi- 

 fications that may be necessary, and in several cases it is shown that 

 the proposed modifications are competent to explain, at least qualita- 

 tively, the complicated effects found at high pressures. 



A preceding paper on water^ is very similar to the present one in 

 the scope of its measurements and computations. The hope was 

 expressed in the introduction to that paper that the projected study 

 of twelve liquids (that is, this paper) would show what we might ex- 

 pect of a normal liquid, and that the results for water might then 

 be compared wuth these results, and yield information about the 

 abnormalities of water. This hope now appears to have been un- 

 founded because no liquid is really normal at high pressures; all show 

 individual peculiarities. It is true, nevertheless, that at high pres- 

 sures these twelve liquids do become more nearly alike in a general 

 way, and that water becomes increasingly like them. In many 

 cases it will be found instructive, therefore, to compare the properties 

 of water at high pressures with those of the liquids of this paper. 



2 Tammann, Ann. Phys., 37, 975-1013 (1912) 



3 Bridgman, These Proceedings, 48, 307-362 (1912). . ., 



