HIGH-PRESSURE RESEARCH BRIDGMAN 167 



important point of view. The pressure effects are complicated, 

 there bein^ maxima, minima, and complicated temperature effects. 

 We do not in general expect that so simple a thing as a hydro- 

 static pressure will produce complicated effects, unless these effects 

 involve a complicated mechanism. I have therefore drawn the 

 conclusion that the thermal e. m. f. mechanism must be compara- 

 tively complicated, and not the simple thing supposed by the classi- 

 cal theory; this I have considered sufficient justification for not 

 attempting to explain thermal e. m. f. in a first simple theory of 

 the electrical properties of metals. 



We pass now from the electrical properties to the thermal prop- 

 erties of metals. It is well known that there is a close relation 

 between electrical and thermal conductivity, which is expressed in 

 the law of Wiedemann-Franz. The behavior of thermal conduc- 

 tivity under pressure shows that, although the relation may be close, 

 there are other factors which have been insufficiently considered. 



The thermal conductivity of most metals increases under pressure, 

 but that of some decreases, and for the most of them the change of 

 electrical conductivity is greater than that of thermal. The reason 

 for the difference is doubtless to be found in the contribution made 

 to thermal conductivity by the atoms as distinguished from the 

 electrons; but a consideration of numerical values suggests that the 

 atomic part may be greater than has been supposed. 



The thermal conductivity of minerals has been measured under 

 pressure, a matter of importance to geophysicists. It has been found 

 that in general the thermal conductivit}^ of minerals increases under 

 pressure, but the increases are not large enough to demand serious 

 modification in our ordinary geological arguments, except possibly 

 at great depths. 



The effect of pressure on the thermal conductivity of liquids is 

 interesting because it is so large; 12,000 kg. increases the conductivity 

 by from two- to threefold. In speculating as to the reasons for this, 

 it appears that there is a very intimate connection with the velocity 

 of sound. Entirel}^ apart from the pressure effects, a reexamination 

 of the old facts with the new bias given by the pressure measurements 

 has shown that the main mechanism of thermal conduction in a 

 liquid is of surprising simplicity, being a sort of combination of 

 the kinetic mechanism of a gas and the elastic wave mechanism of 

 a crystalline solid. If we conceive of each molecule of the liquid as 

 possessing the kinetic energy of temperature agitation demanded by 

 kinetic theory, and handing this energy on to its neighbors with the 

 velocity of an elastic wave of small dimensions — that is, with the 

 velocity of sound — Ave shall account almost entirely for the thermal 

 conductivity of ordinar}' liquids. 



