568 Dr. C. A. Skinner on the Potential Gradient in 



While this gives the positive boundary o£ the Faraday dark 

 space (within an error of observation of possibly 5 mm.), it 

 was also noted that the negative boundary of this space could 

 be located as at the second intersection of 11 with the 

 gradient, as it rises rapidly near the cathode. (The negative 

 glow in these experiments was of small extension.) 



Fig. 4. 



/• X« 3 



~ii fi f» 5» 



Millimetres. 



//o /£4 /$e At> 



The full-line curves represent the results obtained with a 

 luminous discharge, while the dotted curves refer to the dark 

 discharge, produced by increasing the current to a value such 

 that the luminous column was diiven back into the form of a 

 disk on the face of the anode. 



Considering first the gradient for the luminous discharge, 

 it may be observed that, with increasing current the Faraday 

 dark space increases in extent, driving the luminous column 

 before it ; this is accompanied by a decrease in the gradient 

 throughout the dark space. To this conclusion the results 

 show a single exception, seen in fig. 3, in which the curves 

 for 1-5 and for 3 milliamperes coincide. This exception, 

 which led to an explanation of the cause of the dark space, is 

 considered in another paragraph. 



It is also seen that the less the extent of the dark space from 

 the cathode, the greater the rate of increase in extent with the 

 current. 



Again, with increasing gas-pressure, the dark space for the 

 same current decreases in extent. 



As found by Graham (he. cit.) and corroborated lately in 

 the results of H. A. Wilson *, the gradient drops to a very 

 low minimum immediately before its rapid rise near the 

 cathode. These curves show, however, that although with the 

 larger currents this minimum approaches a zero value, yet with 

 * H. A. Wilson, Phil. Mag. xlix. p. 505 (June 1900). 



