﻿674 



Sir, 



F. J. Th. 



omson 



: Further Studies 



on 







Table II. 











Lead 



, 6 = 95. 



Specific 



Ratio of l/0<r 



0. 



1/0(7. 





e/e. 



heat. 



to specific heat. 



273 



366 





2-88 



•993 



368 



1693 



351 





1-78 



•984 



357 



779 



325 





•82 



•928 



350 



20-18 



150 





•215 



•41 



365 



13-88 



87 



i 



Silver 



•145 



, ©=215. 



•2 



435 



273 



366 





1*27 



•965 



380 



169-3 



343 





•79 



■924 



373 



77-9 



252 





•362 



•691 



365 



20-18 



45 





•095 



■073 



615 



Thus-except at the lowest temperatures the ratio of l/6a 

 to the specific heat is fairly constant ; and inasmuch as 

 Kammerlingh-Onnes and Clay have shown that when a small 

 amount of impurity is present, the resistance at very low 

 temperatures approaches a finite value instead of continually 

 diminishing as the temperature falls, it is evident that at 

 these temperatures a trace of impurity would produce a large 

 increase in the value of \j0a. The higher the value of %, 

 the higher will be the temperature at which an abnormally 

 large increase of the conductivity with fall of temperature 

 sets in. Of all metals, beryllium has the smallest atomic 

 value, and so we should expect it to have the greatest value 

 of v and <8) ; it seems probable that the temperature coefficient 

 of this metal may be abnormal even at room temperatures. 



Thermal Conductivity. 



The motion of the chains of electrons along the lines of 

 the lattices will in an unequally heated conductor tend to 

 equalize the temperature, for much the same reason as on 

 the Kinetic Theory of Gases the conduction of heat is 

 brought about by the motion of the molecules of a gas. 

 There are, however, several points of difference which require 

 discussion before we can proceed to find an expression for 

 the thermal conductivity on the chain electron theory. 

 When the temperature is uniform, there is no ambiguity in 

 the statement that the average kinetic energy of the chain is 

 that corresponding to one degree of freedom. A chain of 

 electrons, however, stretches over a distance large compared 

 with the distance between two atoms, and when the temper- 

 ature is not uniform the temperature at one end of the 

 chain may not be the same as that at the other. As the 



