214 On a New Method of Generating Electricity. [Jan. 17, 



and rising from the sides of the crucible, especially at the hottest 

 parts. 



If a platinum tube like that used for the experiments with the 

 suspended plate be somewhat cooled while nearly full of melted glass, 

 so that the glass becomes very viscous, then by applying a flame con- 

 taining free hydrogen to any spot on the lower part of the tube the 

 latter may easily be burst by the bubble of hydrogen which is formed 

 on the inside of the tube. 



Experiments have also been tried in which the hydrogen coming 

 through a cell was removed by means of a Sprengel pump. One 

 experiment may be described : — A platinum tube 2§ inches long and 

 § inch diameter, closed at one end, was soldered to a strong iron tube 

 and fixed vertically. The platinum tube was immersed for 2^ inches 

 in fused glass contained in a platinum cell. This latter was 2 J inches 

 deep and 1 inch diameter. 



The two platinum tubes were connected by platinum wires with the 

 galvanometer, and the iron tube was connected with a Sprengel 

 pump. 



The cell being heated to bright redness in an oxidising flame, a 

 good vacuum was produced by the Sprengel pump. Then, while no 

 bubbles of gas came down the fall tube of the pump, the galvano- 

 meter showed no deflection. 



The cell was then heated by a reducing flame. The galvanometer 

 soon gave a steady deflection of 15°, and bubbles of hydrogen came 

 down the fall tube of the pump. The experiment was continued for 

 half an hour, and during fifteen minutes the hydrogen coming down 

 the fall tube was collected. It measured 1*33 cub. centims. 



From occasional experiments with several vitreous mixtures the 

 conclusion formerly arrived at by the author regarding their electric 

 conductivity is confirmed, viz., that these fused vitreous matters do 

 not conduct electricity of low tension unless hydrogen be present. 

 When working with large cells it is, of course, difficult to avoid the 

 presence of hydrogen if the cell be heated by flame. 



It appeared desirable to try whether any hydrogen could be made 

 to pass through the walls of a porcelain tube either under the influence 

 of oxygen or by means of a vacuum. 



A glazed Berlin porcelain tube 20 inches long and -J- inch diameter 

 was sealed up at one end and connected with a Sprengel pump. The 

 closed end of the tube exposing a surface of 4 square inches was 

 heated to whiteness in the gas furnace, but no hydrogen could be 

 drawn by the Sprengel pump when the porcelain tube was heated in 

 a reducing flame. 



After this the porcelain tube was filled with hydrogen, and the 

 same part as before was heated in an oxidising flame, but no loss of 

 hydrogen from the tube could be perceived during half an hour. 



