Observations at the Poles of the Electric Arc. 971 



ions travelling right across the arc and carrying a current 

 of less than a microampere, a reversed pressure of the 

 required magnitude would be set up. If, as is more likely, 

 they travel a shorter distance and then become merged in 

 the body of the arc and balanced there by negatives, the 

 current they must carry would be larger, but still small 

 compared with the total current in the arc. 



At the cathode the pressure effects appear to be roughly 

 similar to those at the anode, though much more erratic. 

 When the luminous patch really covers the hole, the pressure 

 is positive and may be quite large ; but if the patch is off 

 the centre — either playing on the wall of the hole in the 

 case of a small patch, or partly on and partly off in the case 

 of a large one, — the pressure becomes negative. 



This suggests that, as at the anode, there is a space-charge 

 (in this case positive) at the luminous spot itself; while from 

 the hot carbon around it, negative thermions are emitted 

 which travel for an appreciable distance before ceasing to 

 produce suction by their drag on the gas. 



This drag, if it exists, is not to be confounded with the 

 drag which would result from the slowing-down of electrons 

 emitted at' high speeds from the cathode, as postulated by 

 Duffield. The latter, even if it could appreciably affect the 

 manometer, would not give a suction anything like large 

 enough. The drag here referred to is the steady pull of the 

 field upon unbalanced ions, and may be quite large. 



But speculation on this point is the less necessary, as it 

 has no effect on the main conclusions of the work. 



Summary. 



(1) By a wind-pressure method it has been shown to a 

 degree of accuracy unobtainable by potential slope methods 

 that the body of the arc at atmospheric pressure contains 

 no net space-charge, and that therefore the proportion of 

 the currents carried by the two signs of ion is that of their 

 respective mobilities. 



(2) An excess hydrostatic pressure is set up at the anode 

 by the motion of an unbalanced space-charge of negative 

 ions in a very thin layer close to the anode. The results 

 indicate that this pressure is proportional to the current 

 carried by the arc. 



(3) The distribution of pressure in the neighbourhood o( 

 the cathode is much more complicated, being sometimes 

 positive and sometimes negative. The conditions under 

 which given results are obtained are discussed, ami a tenta- 

 tive explanation in terms of space-charge suo-cvosted. 



38 2 



