to produce Discharges in Gases at Low Pressures. 495 



but this is very small compared with the differences shown 

 for the same pressures in Table I., and is no greater than 

 the variation in the values without regard to current. 



Table II. 



Vi. 



Y, 



v, 



R. 



*• ! 



i 



481 



488 



478 



8 



569 



480 



493 



480 



13 



5-69 



470 



488 



47S 



14 



60 



480 



495 



479 



15 



60 



572 



588 



... 



(16) 



8 25 



330 



320 



323 



-7 



•60 



There seems to be no definite relation between R and the 

 pressure in these experiments, nor much uniformity in the 

 values of R themselves ; but this is not surprising, as the 

 conditions were not the same in all cases. The length of 

 time the current was run was different in different instances, 

 the currents used were of different strength?, and the dis- 

 charge between the plates, though steady, was not always 

 uniform over the surface of the plates. The experiments 

 are of a qualitative rather than a quantitative nature, but 

 qualitatively the results are the same, regardless of variations 

 in the current. 



It is evident from these experiments that the fact that a 

 higher potential is required to maintain a current than to 

 produce a discharge below the critical pressure, is not due 

 to the expulsion of gas from between the electrodes. If in 

 the first method (Table I.) some of the gas is driven out 

 from the spark chamber by the passage of the current, the 

 second value of the sparking potential should be lower than 

 the normal above the critical pressure, and higher below it, 

 since a smaller amount of gas is then occupying the same 

 space that was previously occupied by a larger amount at 

 the same temperature. For the gas cools to its original 

 temperature very rapidly in contact with so much cooling 

 surface, and the diminution of volume due to the rise of the 

 mercury in the bend of the tube is entirely negligible. 

 Table I. shows that, although the results obtained below the 

 critical pressure are in harmony with the supposition tlv.it 

 gas is driven from between the electrode?, those obtained 



