

REQUIRED TO PRODUCE DISCHARGE THROUGH VARIOUS GASER 



Nature of gas. 



Atmospheric air . 

 Hydrogen . . . 



Ordinary nitrogen, care- 

 fully dried 



Nitrogen specially freed 

 from all traces of oxy- 

 gen 



Helium 



Cathode fall given by Warburg. 



340-350 volu (' Wied. Ann., 1 vol. 31, p. 559) 

 About 300 volte (' Wied. Ann.,' vol. 31, p. 581) 

 Varies 315-410 ( p. 557) 



230 volt* ('Wied. Ann.,' vol. 40, p. 1) . . . , 



[Values found by myself (see paper on " Dis- 

 charge of Electricity through Argon and 

 Helium "). 226 volu.] 



Minimum spark 

 potential found above 



volte. 

 341 



302, 308 



347, 351 

 369, 388 



251 



Varies 326-261 



In tlic case of air, the agreement is as good as could lie expected, and, indeed, much 

 I >etter. It is also satisfactory in the case of livdrogen. 



In ordinary nitrogen the results are in neither case constant. But my numl>ers all 

 lie l>etween the extremes found by WARBURG. 



In 'the specially purified nitrogen a difference of about 10 per cent, is to lx? 

 "liserved. For measurements of this kind the discrepancy is not enough to establish 

 an essential want of equality. It is possible that my method of removing oxygen, 

 which depended on the use of alkali metals at ordinary temperatures, was not quite so 

 efficient as that of WARBURG, who used a high temperature. 



In the case of helium alone is the value of the cathode fall seriously different from 

 that of the minimum spark potential. 



Although this discrepancy is large, I do not think that, in view of the satisfactory 

 .1 lavement in the >thrr cases, it can Ix- considered to weigh seriously against the 

 equality. It must 1*' remembered that the values found for helium differ much more 

 among themselves than the lowest of them differs from the measured cathode fall. 

 Measurements of the same quantity which do not agree with one another cannot lie 

 expected to agree with other independent measurements. 



I think it may lie considered to l>e established by these experiments that the 

 minimum spark potential i# equal to the cathode fall. 



One point is very strikingly brought out by the curves. I mean the extremely 

 -mall rate of increase of the spark potential with the pressure in the case of helium, 

 and the relatively high pressure at which the minimum occurs. All the features 

 shown by the curves for the common gases are seen in the curve for helium, but the 

 difference in degree is very striking. The enormous length of spark at atmospheric 

 pressure in helium as compared with air might almost suggest that the conduction is 



vol.. : xcin. A. 3 E 



