Brush Discharges in Gases, 47 



shown) ; two bent glass tubes afford an inlet and outlet for the 

 gas to be experimented on. 



During the experiments a continuous steady flow of gas was 

 maintained through the bell-jar, the clip was connected alter- 

 nately to the two ends of the secondary, and the current in 

 the primary of the induction-coil was reversed in each case. 

 The unused end of the secondary was insulated and entirely 

 immersed in oil. The mercury in the glass dish was electri- 

 cally connected to the plate of a gold-leaf electroscope, the 

 sign of the charge on which was determined in the usual 

 manner. 



In the first case the jar was filled with dry air, of which a 

 steady stream was kept up. The results confirmed those 

 previously mentioned, the charge of the gold leaves being 

 always strong and steady positive. 



Next we passed into the jar a stream of dry hydrogen, and 

 kept this steadily flowing during the experiments, which were 

 of course commenced only after the air was all expelled. 



In this case the leaves were always strong and steady 

 negative. 



This result was confirmed in a remarkable manner by using 

 the ordinary induction-coil only, connected onto the bell-jar. 

 Although there was no visible discharge, yet the electroscope 

 was charged as follows :— 



1. A. Strong steady negative. 



2. A. Leaves go in and out but are always negative. 

 2. B. Strong steady negative. 



1. B. Leaves go in and out but are always negative. 



In the above table A and B represent the secondary ter- 

 minals of the induction-coil, whilst 1 and 2 represent the 

 direction of the current in the primary. "When this was 1 

 the end A was made negative by the break and positive by the 

 make, 



Thus it appears that in the case of our induction-coil the 

 weak E.M.F. due to the make could produce a greater silent 

 discharge than could the strong E.M.F. due to the break, 

 apparently because a negative discharge passes more easily 

 in hydrogen than does a positive. 



Keverting to the previous arrangement of condensers, &c, 

 the jar was next filled with oxygen and the experiments 

 repeated. In this case the electroscope was always charged 

 strongly positive. 



When the induction-coil alone was used, the charge on the 

 leaves changed with the sign of the E.M.F. due to break, as 

 would naturally be expected, but the positive charges were 

 stronger than the negative. 



