Fiar. 1. 



564 Dr. C. A. Skinner on the Potential Gradient in 



placed at any desired position along the axis of the discharge- 

 tube, the same being noted on a scale attached to the 

 barometer - tube. The connexion 

 between the exploring-wires and 

 the electrometer passes, air-tight, 

 through the U-tube. The dis- 

 charge-tube is jointed in the middle, 

 thus allowing the exp'oring- wires 

 to be adjusted, and the electrodes to 

 be exchanged. Its diameter is 

 about 40 mm., and the distance be- 

 tween the electrodes 145 mm. 



Observations were made in nitro- 

 gen only. This was prepared from 

 air and purified after the method 

 of Hittorf *, freed from the last 

 trace of oxygen by nascent sodium f, 

 and thoroughly dried in a chamber 

 containing phosphorus pentoxide. 

 Spec al effort was made to re- 

 move all trace of moisture, which 

 seems to be the cause of much un- 

 certainty in vacuum-tube discharge. 

 For this purpose a tube of phos- 

 phorus pentoxide was placed between 

 the Rapps automatic air-pump and 

 the rest of the system of apparatus, 

 and another of calcium chloride 

 between the pump and the aspi- 

 rator. In addition, only dry air was 

 allowed in contact with the mercury 

 used in pumping. Leakage was 

 guarded against by securing all 

 ground joints with mercury (except- 

 ing the discharge-tube, the joint of 

 which proved absolutely air-tight). 

 It is believed that these precautions 

 insured a degree of constancy, with 

 a rate of operation, which in vacuum 

 discharge is exceptional. The electric current was produced 

 by a battery of 600 secondary cells, and measured by a 

 Weston milliammeter (scale-divisions in tenths). All potential 

 measurements were made by a Thomson quadrant electro- 

 meter carefully insulated, while sealing-wax insulators and 



* Loc. cit. p. 725. 



t See E. Warburg, Wied. Ann. xl. p. 1 (1890). 



