814 Prof. 0. W. Richardson : 



ventures to suggest that the word " Thermionics," which 

 forms the title oil this paper, is very suitable for the purpose. 

 It suggests at once the thermal and electrical nature of the 

 phenomena, while the derivation of the word ion also 

 suggests the kinetic qualities of the ions. 



The lack of such a word is found to be particularly aggra- 

 vating in considering the phenomena attending a hot wire 

 heated to a high temperature by an electric current. Here 

 we have two currents : the current used to heat the w r ire and 

 the thermionic current away from the surface of the latter. 

 It is constantly necessary to distinguish between the two, 

 and for this purpose the thermionic current is usually alluded 

 to as " the leak from the wire " or simply " the leak." 

 There are numerous objections to the use of the word "leak/' 

 It .seems absurd to apply it to cases like that of a very highly 

 refractory substance such as tungsten or carbon or of one 

 of Wehnelt's lime-covered cathodes, in which the so-called 

 " leak " may be as big or bigger than than the rest of the 

 heating current ; there is also the moral objection that the 

 leak analogy rather implies something undesirable and pre- 

 ventable ; while finally we have the most weighty objection 

 of all, that the term leak is not sufficiently adaptable. If 

 we wish to refer to the particles which carry the leak we 

 have to revert to some such phrase as " the ions emitted by 

 the hot body/' whereas the substantive Thermionics furnishes 

 naturally the further substantive Thermion. 



§2. So much by way of definition. The chief object of 

 this paper is to point out a class of theoretical problems 

 which become important in thermionics, to discuss, in as 

 general a manner as possible, methods by which they can be 

 solved, and to deduce the solutions for certain particular cases. 



Suppose that in a region of space otherwise vacuous there 

 are a number of hot surfaces A emitting ions and a number 

 of conducting surfaces B. In general there will be an 

 electric field in the region under consideration, so that any 

 or all of the surfaces may be charged. The ions emitted by 

 the surfaces A will move under the combined influence of 

 their initial velocity and that of the electric field, and will 

 ultimately reach one of the surfaces A or B or go off to an 

 infinite distance. If the temperatures of the surfaces A are 

 maintained constant the number and mode of distribution of 

 velocity of the ions they emit will remain constant, and if 

 in addition the potentials of the various surfaces are main- 

 tained constant it is clear that, whatever may happen at 

 first, a steady state will ultimately be established in which 

 the number and mode of distribution of velocity among the 



