the Faraday Bark Space of Vacuum-Tabes. 571 



column, where this convexity is very slight, yet from the 

 fact that it disappears when the cathode rays are turned 

 aside, it must be considered as under the control of cathode 

 radiation. 



In connexion with these considerations, the curves of 

 H. A. Wilson (loc. cit.) showing the electric conductivity of 

 the gas normal to the direction of the current are o£ special 

 interest. According to these the conductivity of the Faraday 

 dark space increases in passing towards the cathode, which 

 agrees with the decreasing gradient, and with an increasing 

 ionization as the source of the cathode rays is approached. 

 Passing in the opposite direction through the dark space 

 into the luminous column, while the gradient shows that the 

 conductivity decreases gradually to a constant minimum in the 

 luminous column, a certain region in the dark space is reached 

 where, on the other hand, the conductivity in a direction 

 normal to the current becomes a minimum, then rises to a 

 constant value in the luminous column. Were the conduc- 

 tivity dependent on ihe degree of ionization alone, it should 

 be independent of the direction in which it is measured. 

 Some other factor evidently enters to which this difference 

 must be ascribed. 



While Wilson suggests it to be due to the luminosity 

 ionizing the gas at the surface of the " cross-current " 

 electrodes,' and thus, for the same difference of potential 

 between them, allowing a larger " cross-current " to pass in 

 the luminous portions than in the dark spaces ; it may also 

 be explained from the conclusion that cathode rays are active 

 in producing the dark space. This explanation rests upon 

 the assumption that the negative ions driving through the gas 

 with a decreasing velocity, cease to ionize it while their 

 velocity is still above that maintained by them in the luminous 

 column. Since the conductivity in any direction (other 

 conditions remaining the same) must increase with increased 

 velocity of the ions in that direction, the conductivity along 

 the discharge, in the higher velocity non-ionizing region of 

 the negative ions, would be greater than in the luminous 

 column, and also greater than in a direction at right angles 

 to the discharge in the same region. Furthermore, with the 

 same degree of ionization in the luminous column as in this 

 region, the conductivity at right angles to the direction of 

 motion of the ions would be lower in the latter, due to the 

 stream tending to carry them past the limited " cross-field." 

 It seems possible that, with a sufficient velocity of the ions 

 along the direction of the discharge, the conductivity at right 

 angles might be reduced without limit, 



2 S2 



