1891.] Luminous Discharge of Electricity through a Gas. 1'.") 



sadden alterations in the electric or magnetic intensity of the field in 

 which it is placed, unless these are accompanied by the passage of 

 free electricity through the gas. 



In order to get some further information about the laws which govern 

 the propagation of the positive column, some experiments were made 

 in which the discharge had to pass from the gas to mercury and out 

 again from the mercury to the glass several times in its passage from 

 B to G (fig. 1). The arrangement by which this was done is shown 

 in fig. 3. Pieces of glass tubing, bent as in the figure, with baro- 



FIG. 3. 



metric tubes filled with mercury attached to their lowest points, 

 were inserted in the circuit between B and G. By raising or lower- 

 ing the vessels into which the ends of the barometer tubes dipped, 

 mercury could be poured into or taken out of the bends in the tube. 

 There were in all six of these mercury electrodes introduced between 

 B and G. The displacement of the images, as seen through the tele- 

 scopes, was observed (1) when the mercury was below the level of 

 the tops of the barometer tubes, and (2) when the mercury filled the 

 bends of the tube, blocking it up completely in six places. No 

 appreciable difference could be observed between the displacements 

 of the images in the two cases. When, however, the mercury was in 

 the tube, the discharge had very much greater difficulty in getting 

 through than when its path was not interrupted by the columns of 

 mercury ; this was shown by the luminosity in the maiu circuit being 

 very much fainter, and that in a branch circuit leading to the air- 

 pump much brighter, when the mercury was in the tubes than when 

 it was not. 



It seems, I think, pretty clear that what takes place when the 



