1891.] Luminous Discharge of Electricity through a Gas. 89 



stances to within 25 per cent, and it is probable that, in our present 

 state of knowledge about the discharge of electricity through gases, 

 the points of most importance can be settled by a somewhat rough 

 determination of the velocity of propagation. 



Some hundreds of observations were made, and in every case in 

 which the observer declared the images to be very bright, and in 

 every case but one in which the images were declared to be fairly 

 bright, the displacements (if there was not a very large air break in 

 the circuit) corresponded to the luminosity travelling from the 

 positive to the negative electrode. When AB was the negative 

 electrode, the luminous discharge arrived at GH, a place about 

 25 feet from the positive electrode, before it reached BC, which is 

 only a few inches from the cathode, and as the interval between its 

 appearance at these places was about the same as when the current 

 was reversed, we may conclude that, when AB is the cathode, the 

 luminosity, which is found only a few inches from it at BC, has 

 started from the positive electrode, and traversed a path enormously 

 longer than its distance from the cathode. We thus arrive at the 

 conclusion that the positive column, which in a long tube like the one 

 under consideration practically fills the tube, since it extends to 

 within an inch or two of the anode, starts from the positive elec- 

 trode. 



In view of the probability that the passage of the current from the 

 electrodes to the gas might be largely influenced by chemical action 

 between the electrode and the gas, 1 repeated the experiments with 

 electrodes of very different kinds ; the result, however, was the same, 

 whether the electrodes were pointed platinum wires, carbon filaments, 

 flat surfaces of sulphuric acid, or the one electrode a flat liquid 

 surface and the other a sharp pointed wire. Tne positive column 

 starts from the positive electrode, even though this is a flat liquid 

 surface while the negative is a sharp-pointed wire. 



Velocity of Propagation of the Discharge. 



The displacement of the images of the two luminous portions of 

 the tube caused by the rotation of the mirror was equal to the 

 distance between the images of divisions 1'5 mm. apart on a vertical 

 scale placed at GH. Thus, if T is the time the luminosity takes to 

 travel from BC to GH, since the distance of the mirror from the 

 luminous part of the tube is 6 metres, the circular measure of the 

 angle turned through in the time T is 1'5/12,000. If n is the number 

 of revolutions made by the mirror per second, 



I . """- it* r\rif\ . . 



12,000 X 2irn 



