at Low Pressures through the Passage of Electricity* 173 



It will be seen from the figure that the lower electrode was 

 in metallic connexion with the brass base of the apparatus 

 upon which the bell- jar rested ; and that the upper electrode 

 was in metallic connexion with the brass rod which passed 

 through the ebonite ping fixed in one of the side tubes of the 

 bell-jar. Tims the electrodes could be easily connected by 

 wires to the terminals of the battery which supplied the 

 current. 



The bell-jar was fastened down to the brass base by means, 

 of elastic glue, and the necessary joints in the bell-jar, where 

 it was impossible to fuse the glass, were made in the same 

 way. Every other joint in the glass tubing along which the 

 gases had to pass was made by fusing together the glass. All 

 the stopcocks and drying vessels had mercury cups which 

 prevented air from passing in. By these means the appa- 

 ratus was made satisfactorily air-tight. In fact, joints which 

 are carefullv made with elastic glue, even between laroe 

 perimeters m contact, prove often to be surprisingly airtight, 

 and last for months. 



Phosphorus pentoxide was placed in vessels (not shown in 

 the figure) within the bell-jar, and also in a large tube con- 

 necting the bell-jar to the McLeod gauges. This absorbed 

 the water-vapour as soon as it was formed. 



The bell-jar was about 10 cms. in diameter, and its height 

 up to its shoulders was about 12 cms. The lower plate was 

 protected by an ebonite guard-ring throughout these expe- 

 riments, and when silver electrodes were substituted the upper 

 plate was similarly surrounded. 



To prepare for a series of experiments, the bell-jar was 

 alternately exhausted by the mercury-pump and replenished 

 by small quantities of freshly prepared and dried hydrogen 

 and oxygen. In this way the proportion of the old to the 

 newly introduced gas was reduced to a very small percentage. 

 The stopcock cutting off the bell-jar, the manometer, and the 

 McLeod gauges from the rest of the apparatus was then 

 closed, the total volume thus cut off being nearly 1130 c.c. 



The method of these experiments was throughout the same 

 as the method described in my former paper. The parallel 

 plates AB.CD of the apparatus were connected to the poles 

 of a battery of small lead cells, ranging up to 1080 volts, 

 through a high resistance R, and through a sensitive am- 

 meter V frig. 2, p. 174^. 



Thus with a suitable voltage and resistance a current 

 attended by a glow was passed for an observed time through 

 the mixed gases and was measured by the ammeter. It was, 

 in the case of all the observation:- recorded below, steady 



