the Faraday Dark Space of Vacuum- Tid>es. 5b'l) 



a very small current and higher gas-pressures it may reach a 

 considerable value. Graham's curves show on the whole two 

 minima in this space separated by a small maximum. Those 

 o£ Wilson show this maximum exceptionally marked. With 

 a cathode 2 centim. in diameter, I observed likewise a small 

 maximum between two minima, but with the larger cathode 

 (diam. 3 cm.) the maximum was at no time detectable, thus 

 indicating it as probably dependent on the density of the- 

 discharge from the cathode-surface. 



As to the range over which the observations extended, it 

 may be stated that they were made for gas-pressures inter- 

 mediate between, as lower limit, the pressure at which the 

 luminous column is still unstriated (about 'G mm.), and as 

 upper limit (2 mm.), the power-limit of the battery. The 

 current ranged from the lowest possible to that which produced 

 the dark discharge. The difficulty attached to> the main- 

 tenance of small constant currents lies probably in the unstable 

 condition of the conducting gas, as seen in the variation 

 in the total drop of potential between the electrodes with the 

 current. The total drop of potential throughout the con- 

 ducting gas, being proportional to the area contained between 

 the gradient and the axes of coordinates, is seen with small 

 currents especially to decrease rapidly with increasing current ; 

 and the reverse, increases rapidly with decreasing current.. 

 Thus the action of the gas is to magnify any variation in the 

 rest of the circuit, and to lend itself, with a capacity and high 

 resistance in the circuit, readily to a pulsating discharge. 

 Another difficulty in maintaining constant small currents 

 arises when the discharge, taking place from smaller and 

 smaller surfaces of the cathode as the gas-pressure increases, 

 being reduced to a spot on the cathode, dances rapidly about 

 over its surface. 



Returning to the fact that the curve in fig. 3 for 3 m.a. 

 coincided throughout with that for 1*5 m.a., it was noticed 

 at the same time that, with the current very slightly above 

 1*5 m.a., the discharge began to take place from the back 

 surface of the cathode also. This suggested that the dark 

 space was under the direct influence of a radiation from the 

 discharging-surface of the cathode. To test this point the 

 cathode was turned, as shown in fig. 5, with its plane lying- 

 in the axis of the discharge-tube, so that any disturbance 

 sent out normal to its surface would be directed against the 

 walls of the tube. In addition, any component radiation 

 along the axis of the tube was prevented by a glass cylinder 

 which surrounded the cathode co-axially. 



Under these conditions the luminosity, with an unstriated 

 Phil. May. S. 5. Vol. 50, Noc 307. Dec. 1900. 2 S 



