108 CONDUCTION AND RADIATION [CH. x. 



which the transfer of an electron from one atom to 

 another is easy, demanding no force as long as the 

 process is not hurried a process of the nature of a 

 diffusion. The transmission of vibrations along a 

 chain of connected molecules may well occur through 

 a not dissimilar kind of connection ; and hence the 

 conduction of electricity and the conduction of heat, 

 though really different processes, may have many 

 points in common. 



A fair approximation to the phenomena of con- 

 duction in metals has been worked out in detail by 

 Riecke, Drude, Thomson, Lorentz, and many others ; 

 in which the electrons are supposed to remain free 

 for periods so long that their mean energy of motion 

 is a function of the temperature, as in gas-theory. 



Most is known about electrolytic and gaseous 

 conduction. In gaseous conduction the negative 

 electrons, when free, fly fast ; whereas the ions 

 generally, and all the positive charges, travel more 

 slowly by reason of their association with matter. 



In liquid conduction charges of either sign are 

 always associated with atoms, and travel only as 

 ions, at a slow diffusion rate which was calculated 

 by Kohlrausch, has been observed directly by myself,* 

 Mr. Whetham and others, and is well known. 



The rate of transmission in solids can only be 

 inferred, and it would appear from the Hall effect (see 

 Larmor, Phil. Trans., Aug. 1894, p. 815), as if in 

 one class of solids the positive were able to travel 

 fastest, whereas in another class negative travelled 

 fastest : a difference which is familiar in liquids. In 

 acids, for instance, the positive charges travel much 

 the quickest, because they are associated with light 



* Lodge, British Association Report, Birmingham Meeting, 1886, 

 pp. 389-413. 



