﻿to maintain a Current in a Gas at Low Pressures. 229 



nomenon duo to a change in the gas, was shown by tho 

 behaviour on reversing the current. Table TIL shows a typical 

 case. 



Table III. 



Effect on Sparling Potential of Reversing the Current. 



2)=io. Jan. 25, 1906. 

 Zn. 



Sp. V. with A+ = 391 volts. 



» >, A- = 403 ,. 

 After running- current 1/4000 amp. 11 mins. 



Sp. V. with A+ = 431 volts. 



„ „ A- = 405 „ 



The sparking potential to start with was in general the 

 same whether the top plate (A) was positive or negative. 

 After running the current with A positive, the sparking 

 potential taken in the same war showed an increased value 

 in this direction, if the plates were zinc ; bat if reversed, 

 i. e. A negative, the sparking potential was still at its original 

 value. This phenomenon must be distinguished from the 

 previous as it is a very temporary one, the effect disappearing 

 almost immediately ; and a further difference is that the 

 effect is in the same direction, whatever the metal. It has 

 been found that the current running only for a few seconds 

 will have this effect, while the effect due to the material of 

 the electrodes, spoken of before, required a large current for 

 a number of minutes, and lasted for hours after the current 

 had been stopped. Unfortunately, it was not thought of in 

 time to take the sparking potential in the reverse way from 

 which the current had been running, in investigating the 

 previous effect ; so this having disappeared, there is no criterion 

 to decide whether the present phenomenon could also have 

 been distinguished then, when much larger changes in the 

 potential due to running the current were available. In 

 general, the determination of the sparking potential, especially 

 with fresh gas or some that had been left standing overnight, 

 did not immediately give its permanent and correct value. 

 Quite a large force could be put on before the spark would 

 pass, but as soon as this had happened, the proper value for 

 the sparking potential could be obtained ; and in the new 

 state of affairs, further sparking would have no appreciable 

 effect on the value. This effect was the same whatever the 

 material or the original direction of the current. On reversing 

 the electrodes, however, the value for the sparking potential 

 was always lower, gradually rising to the same value as for 



