324 



Prof. J. J. Thomson on the Discharge 





its distance from B without the spark ceasing. Bailie *, who 

 observed the same effect in air that presumably was not 

 exceptionally dry, gives the following numbers : — 



Sparking-distance 



Pot.-diff. for 



_ Pot. 



-diff. for spark 



in cm. 



1st spark. 



immediately following. 



•05 



14-52 





12-85 



•10 



23-98 





22-64 



•15 



33-19 





32-49 



•20 



40-86 





39-47 



•25 



49-34 





44-53 



•30 



53-13 





48-58 



•35 



59-43 





50-85 





Thus the effect in the damp gas, though well marked, is 

 not very large. In the dry gas, however, this effect attains 

 quite abnormal proportions, the potential-difference required 

 to produce the first spark being often more than twice that 

 required to maintain it when once started ; indeed, in some 

 cases when the gas had been dried for a very long time the 

 whole electromotive force of the battery was not sufficient to 

 initiate the discharge, and I had to start it by means of an 

 induction-coil: when a spark had been sent through the gas 

 by this means, an electromotive force of about J of that 

 which had previously been applied to it without effect was 

 sufficient to send a second spark through the gas, if the 

 interval between the first and second sparks was not more 

 than a minute or two. If a much longer interval than this 



Fior. 2. 



JZieetrcmcter 



Sparking 



was allowed to elapse the gas recovered its original strength. 

 The experiments were made as follows : — The arm EF (fig. 2) 

 was slowly moved in the direction of increasing potential 

 until a spark was seen to pass through the dry hydrogen ; as 



* Anmiks de Chimk et de Physique, [5] xxix. p. 181 (1883). 



