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

 Gradient near the Anode. 



We have in the above results an explanation of the low 

 gradient near the anode also. The positive ions driven forth 

 with a velocity imparted Irv the anode "drop" render the 

 space in which the higher velocity is maintained more con- 

 ductive thereby in the direction of their motion. That these 

 do not ionize the gas through which they drive, but on the 

 other hand increase the conductivity at right angles, may be 

 seen in the fact that (as given by Wilson) the curve of 

 conductivity drops to a very small value in this space. 



The potential gradient near the anode in figs. 2, 3, and 4 

 reaches a minimum, though not a zero value as found earlier 

 and also corroborated by Wilson. The drop of potential at 

 the anode was normal (18 to 25 volts), so that this dis- 

 crepancy is probably to be attributed to the large discharge- 

 surface of the anode used in this investigation. 



Another observation made in connexion with the anode is 

 of interest, in that it concerns a questionable state of the gas 

 near the anode. As has been noted in the earlier communi- 

 cation by me, aud more thoroughly studied by Wilson, with 

 striated discharge the gradient in the striation on the anode 

 is apparently negative. With the larger anode used in this 

 investigation the same has been observed in an unstriated 

 discharge, though not always. In one case the minimum 

 potential-difference between the anode and the first exploring- 

 wire was obtained with the wire at a distance of one centi- 

 metre from the anode. This increased perceptibly as the 

 wire approached the anode, while at the same time a shadow 

 of the wire was cast on the face of the anode. A negative 

 gradient was thus indicated, yet the gradient indicated by the 

 exploring- wires was a considerable positive quantity. The 

 discrepancy in this case must be attributed to a disturbance 

 of the discharge at the anode by the proximity of the 

 exploring-wires. This strengthens then the view that former 

 similar results were due to the same cause, rather than that 

 there is actually a negative electric intensity in tha space 

 considered. 



Striations. 



In fig. 2 the curve for 4 m.a. represents the gradient for 

 a clearly-marked striated discharge produced by simply 

 increasing the current above 2 m.a.. It was repeatedly 

 observed that with the pressure of the gas near the limit 

 between the striated and unstriated discharge, the latter could 

 be converted into the former by increasing the current — like- 

 wise the reverse. This change is gradual. With a pressure of 



