A Physical Study of the Thermal Conductivity of Solids

Arthur H. Compton
1916 Physical Review  
OINCE the physical interpretation of the variation of the specific ^ heat of solids with temperature has been found to be of considerable theoretical importance, it is to be expected that a study of the physical nature of the thermal conductivity of solids may also bring interesting results. The present paper may be considered as a start at such a study. In order not to be disturbed by thermal conduction due to electrons, we shall consider only non-metallic substances. For the sake of further
more » ... e sake of further simplicity we shall consider in the first place the conduction of heat through a crystal, in which the atoms are closely packed. Consider two infinite, parallel planes in such a medium, at a distance x apart, and maintained at a constant difference of* temperature, T\ -To. Let dx be the^average distance in the x direction through which the energy of an atomic collision is transmitted. The temperature drop along this distance dx will evidently be 8x 6T = -(r x -To). x This quantity represents the average difference in temperature between the atom which gives out energy at each collision, and the atom or atoms which absorb it. According to ordinary kinetic theory, if the collisions of the atoms are elastic, the average amount of energy transmitted at each collision may be shown to be where Wi and w 2 are the masses of the atoms in collision, and R is the gas constant for a single molecule. This may be written where b is in most cases approximately equal to I. In the kind of a medium we are considering, the number of atomic collisions in unit time on unit area of the surface of a plane separating
doi:10.1103/physrev.7.341 fatcat:a56qclhaqbhwhlxborwmxynx3a