326 Dr G. Olearski, Some experiments on the dielectic [Nov. 23, 



vapour. They determined the quantity of electricity passing at 

 each discharge, and, although it is perhaps not quite evident, it is 

 at least very probable that the quantity of electricity carried at a 

 discharge is proportional to the potential necessary to introduce 

 the discharge. Based on this supposition we should conclude from 

 the diagram 1 given by Wiedemann and Ruhlmann that there is an 

 anomaly as regards the strength of air, which according to them 

 would be intermediate between its constituent gases for pressures 

 larger than 30 mm. of mercury, but weaker than either nitrogen 

 or oxygen for pressures lower than that. This would not be what 

 we could easily explain, and would suggest that for these low pres- 

 sures only the weaker constituent of the mixture is carrying the 

 discharge. 



But as this result does not seem to be conclusively proved 

 by the experiments quoted above, and is not even announced 

 by the investigators, and it was of some interest to make ex- 

 periments on the strength of mixtures of gases, Prof. J. Thomson 

 suggested to me during my stay in Cambridge to investigate the 

 question by the method of Faraday 2 . Accordingly a current given 

 by a Ruhmkorff coil was sent through a multiple conductor, in one 

 branch of which was a glass tube (whose internal diameter was 

 2 cm.), closed air-tight by india-rubber stoppers, having two spheri- 

 cal platinum electrodes at a distance of about 15 cm., while in 

 the other branch there was the usual apparatus for measuring 

 the length of sparks, consisting of two movable brass balls in 

 open air. 



When the distance of movable electrodes is large enough, 

 sparks pass only in the glass tube, when it diminishes they appear 

 between the movable spheres as well. The limit distance measures 

 the strength of the gas enclosed in the tube. There is however 

 a difficulty in making an exact measurement, for, when the distance 

 of electrodes is reached at which the first spark passes across 

 them, the discharge does not always choose the same way. In this 

 manner there is an interval which is larger for higher pressures 

 than for lower ones, in which it is impossible to decide whether 

 the gas enclosed in the tube or the air between the movable 

 electrodes is stronger. But when we only compare the strengths 

 of different gases, faults coming from this incertitude may be 

 possibly eliminated by keeping always the same mode of proceed- 

 ing. For every determination therefore the distance of the movable 

 electrodes was at first made so small that all discharges passed 

 through them. Then the gas was pumped out by means of a 

 Sprengel pump and the pressure was noted at which the first dis- 



1 1. c. tab. iv. fig. 2, or Wied. Elektr. Vol. iv. p. 462, fig. 193. 

 3 Experim. Research. Vol. i. 



