﻿224 Mr. J. A. Brown on the Potential required 



fall is exactly the same as the difference of potential between 

 the electrodes. So when the pressure is decreased below 

 this value, the actual cathode fall as such has ceased to exist. 

 In the present investigation with a sparking distance of one 

 centimetre, the critical pressure was about 1/1 mm.; so that 

 in all the cases noted, when the potential rose with the 

 current (fig. 1), the actual cathode fall of potential was no 

 longer anything definite. On the other hand, it is most 

 improbable that there should not be a considerable heating 

 effect under the circumstances, which would of necessity 

 cause the gas to expand and drive some of it out from 

 between the plates. That being so in any case, it is much 

 simpler to find in this the explanation for the phenomenon 

 without calling in further hypotheses. 



Thus, to sum up the explanations given to the observed 

 conditions : — In Case I., where there was a drop from the 

 sparking potential tending toward a constant value, a polari- 

 zation effect is set up in the gas, which tends to maintain the 

 current at a lower potential than is needed to cause the 

 current to pass originally. In Case II., below the minimum 

 sparking potential, when the potential to maintain the current 

 rises with the current, the polarization is small, and the rise 

 in the potential is to a large extent due to a heating effect 

 which drives out some of the gas. 



Part II. 



Some Peculiarities of the Spark-Discharge. 



It has seemed worth while to tabulate a few of the pecu- 

 liarities noticed at various times during the course of the 

 investigations, some of which entered as sources of error 

 which had to be eliminated or which disappeared of them- 

 selves, and others which did not seem to affect the results 

 given, and to which in consequence less attention was paid. 



A. Effect of the E.M.F. of Battery on Value of Potential 

 required to maintain a given Current. 



As was mentioned above, the curves at first obtained were 

 often very complicated, especially at high pressures, when 

 one would expect, from previous work, that the curve would 

 be practically straight. The variations in the current were 

 produced in this case by changing both the resistance and 

 the number of the cells in the battery. On going over the 

 Tables, it was found that whenever there was a small resistance 



